JOM - April 2021

376 Dubey et al.: Piloting a longitudinal curriculum

like ours could be essential to encouraging and maintain- 2. Hensel KL, Buchanan S, Brown SK, Rodriguez M, Cruser d.
ing OMT’s adoption in residency programs with OR. Pregnancy research on osteopathic manipulation optimizing
treatment effects: the PROMOTE study. Am J Obstet Gynecol 2015;
Conclusions 212:108.e1–9.

The multitude of applications for OMT, its growing evi- 3. Noll DR, Degenhardt BF, Morley TF, Blais FX, Hortos KA, Hensel K,
dence base, and the novel landscape of the ACGME single- et al. Efficacy of osteopathic manipulation as an adjunctive
accreditation system herald a new chapter in postgraduate treatment for hospitalized patients with pneumonia: a
osteopathic training. As more residency programs attain randomized controlled trial. Osteopath Med Prim Care 2010;4:2.
OR, an inclusive osteopathic curriculum is needed to
ensure that OMT has broad respect and use. Results from 4. Baltazar GA, Betler MP, Akella K, Khatri R, Asaro R,
the quality improvement study of our pilot curriculum Chendrasekhar A. Effect of osteopathic manipulative
showed that a longitudinal residency-based OMT elective treatment on incidence of postoperative ileus and hospital length
for MDs can improve their attitudes and confidence of stay in general surgical patients. J Am Osteopath Assoc 2013;
regarding osteopathic principles and practices. Areas for 113:204–9.
future research include competency assessment and
impact on practice patterns. 5. Mills MV, Henley CE, Barnes LL, Carreiro JE, Degenhardt BF. The
use of osteopathic manipulative treatment as adjuvant therapy in
Acknowledgments: The authors thank Mindy Smith, MD, children with recurrent acute otitis media. Arch Pediatr Adolesc
and David Rabago, MD, for review of the manuscript for Med 2003;157:861–6.
clarity.
Research funding: None reported. 6. McReynolds TM, Sheridan BJ. Intramuscular ketorolac versus
Author contributions: All authors provided substantial osteopathic manipulative treatment in the management of acute
contributions to conception and design, acquisition of neck pain in the emergency department: a randomized clinical
data, or analysis and interpretation of data; all authors trial. J Am Osteopath Assoc 2005;105:57–68.
drafted the article or revised it critically for important
intellectual content; all authors gave final approval of the 7. Attali TV, Bouchoucha M, Benamouzig R. Treatment of refractory
version of the article to be published; and all authors agree irritable bowel syndrome with visceral osteopathy: short-term
to be accountable for all aspects of the work in ensuring that and long-term results of a randomized trial. J Dig Dis 2013;14:
questions related to the accuracy or integrity of any part of 654–61.
the work are appropriately investigated and resolved.
Competing interests: Authors state no conflict of interest. 8. American Osteopathic Association. Osteopathic recognition:
Ethical approval: The Institutional Review Board at the American Osteopathic Association. Available from: https://
University of Wisconsin in Madison determined this project osteopathic.org/graduate-medical-educators/single-gme/
to be exempt. osteopathic-recognition/.

References 9. Hempstead LK. Single accreditation system: opportunity and duty
to promote osteopathic training for all interested residency
1. Qaseem A, Wilt TJ, McLean RM, Forciea MA, Clinical Guidelines programs. J Am Osteopath Assoc 2015;115:193–5.
Committee of the American College of Physicians. Noninvasive
treatments for acute, subacute, and chronic low back pain: a 10. James S. Allopathic faculty supervision of osteopathic education:
clinical practice guideline from the American College of what support is needed? Osteopathic Family Physician 2019;11:
Physicians. Ann Intern Med 2017;166:514–30. 10–3.

11. Rubeor A, Nothnagle M, Taylor JS. Introducing osteopathic
medical education in an allopathic residency. J Am Osteopath
Assoc 2008;108:404–8.

12. Chila A. Foundations of osteopathic medicine, 3rd ed. Lippincott
Williams & Wilkins; 2011.

13. Rowane M. Basic musculoskeletal manipulation skills – the 15
minute office encounter. American Academy of Osteopathy;
2012.

14. Wu P, Siu J. Brief guide to osteopathic medicine, for students, by
students [Online]. American Association of Colleges of
Osteopathic Medicine; 2015.

15. Gustowski S, Budner-Gentry M, Seals R. Osteopathic techniques:
the learner’s guide. New York: Thieme; 2017.

Supplementary Material: The online version of this article offers
supplementary material (https://doi.org/10.1515/jom-2020-0038).

J Osteopath Med 2021; 121(4): 377–383

Medical Education Original Article

Michael A. Downing, BS, Michael O. Bazzi, BS, Mark E. Vinicky, MS,
Nicholas V. Lampasona, BS, Oleg Tsvyetayev, BS and Harvey N. Mayrovitz*, PhD

Dietary views and habits of students in health
professional vs. non-health professional graduate
programs in a single university

https://doi.org/10.1515/jom-2020-0178 participated in less exercise compared to men. Women
Received July 11, 2020; accepted August 21, 2020; also consumed more sweets compared to men.
published online February 12, 2021 Conclusions: Results suggest that NSU students enrolled
in HP and NHP programs have similar nutritional concepts
Abstract and eating habits. This may indicate a need to strengthen
nutritional education in dietary health and wellness for HP
Context: Students enrolled in health professional (HP) students.
programs receive varying amounts of credit hours
dedicated to nutritional education, and obesity remains Keywords: diet; graduate school; health professional;
an issue in the United States among healthcare providers. non-health professional; nutrition; obesity.
Objectives: To assess whether HP students differ in
nutrition and exercise habits from non-health professional The incidence of obesity, defined as a body mass index
(NHP) students at a single university, and whether any (BMI) greater than 30 kg/m2, has become an epidemic in
gender-related differences existed in those habits. the United States [1, 2]. A systematic review of controlled
Methods: From September 25, 2018 to October 10, 2019, a trials [3] showed that nutritional plans and physical ac-
16-question multiple-choice survey was distributed via tivity leading to weight loss of ≥5% yielded improvements
e-mail or in person to HP and NHP students enrolled at in patients’ lipid panels, including triglycerides, total
Nova Southeastern University (NSU) in Fort Lauderdale, cholesterol, low-density lipoprotein (LDL), and high-
Florida. Questions targeted participant dietary and exer- density lipoprotein (HDL) [3]. As of 2016, the Centers for
cise habits. Each question had five multiple-choice answer Disease Control and Prevention (CDC) estimated that
options, each of which was assigned a coded value to roughly 40% of adults in the United States were obese [4].
compare similarities and differences between the HP and Meanwhile, obesity among college graduates was 22.7% in
NHP groups. 2016 [4]. Even in the “healthiest” age demographic in the
Results: Of 732 responses (569 HP, 163 NHP), results United States (adults aged 18–24), the prevalence of
showed no statistically significant difference between obesity is 16.5% [4], which is still higher than the world-
enrollment groups (p>0.05) in any response parameter wide prevalence of 13.5% [5].
including consumption of sweets, fast food, red meat,
caffeine, water, fruit, and vegetables. Comparisons among Studies have also demonstrated increasing rates of
sexes demonstrated significant differences. Women obesity among healthcare providers. According to a study
consumed less red meat, water, and protein, and women published in the American Journal of Preventative Medicine
[6], US healthcare workers had a 22% obesity prevalence in
*Corresponding author: Harvey N. Mayrovitz, PhD, Dr. Kiran C. Patel 2010. Obesity among nurses is especially elevated, ac-
College of Allopathic Medicine, Nova Southeastern University, 3200 S. cording to a survey distributed to American nursing
University Drive, Davie, 33328-5326, Fort Lauderdale, FL, USA, professionals, which revealed that 54% of the 760 re-
E-mail: [email protected] spondents were overweight or obese (mean BMI, 27.2 kg/
Michael A. Downing, BS, Michael O. Bazzi, BS, Mark E. Vinicky, MS, m2) [7]. Furthermore, certain studies have indicated that
Nicholas V. Lampasona, BS and Oleg Tsvyetayev, BS, Dr. Kiran C. unhealthy habits practiced among healthcare providers may
Patel College of Osteopathic Medicine, Nova Southeastern University, have originated during their time in health professional (HP)
Fort Lauderdale, FL, USA school [8, 9]. Unhealthy habits among medical students
have been shown to impact their counseling of patients. For

Open Access. © 2020 Michael A. Downing et al., published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0
International License.

378 Downing et al.: Dietary views and habits of students

example, a “Healthy Doctor = Healthy Patient” survey [8] in Surveillance System (YRBSS), specifically pertain to nutri-
2015 among medical students at a Colombian university tion [1]. The YRBSS utilizes a survey for data collection
assessed attitudes about counseling, personal attitudes, regarding nutritional habits in a specific population.
physical activity, personal nutrition, risky alcohol con-
sumption, smoking, inadequate nutrition, and noncom- We hypothesized that students in HP programs would
pliance; survey results were also compared against the report different diet and exercise habits compared with
overall university environment where the medical stu- students in NHP programs.
dents were studying and showed that despite the uni-
versity’s effort to encourage better nutrition among Methods
medical students, inadequate nutrition and sedentary
habits continued to increase throughout their medical This study was reviewed and determined exempt by the Nova
education [8]. Students who reported more physical ac- Southeastern University institutional review board (IRB No.
tivity were more likely to counsel patients on preventive 2018-455-NSU).
health. Further, these students also reported healthier
lifestyle choices, defined by increased fruit and vegetable A 16-question survey was created and distributed to NSU students
consumption, less tobacco use, and less alcohol con- enrolled in both HP and NHP programs. NHP programs included
sumption [8]. graduate-level programs in nonmedical fields. Survey invitations
were distributed to students of any year throughout their respective
The National Research Council (NRC), which provides program. The survey was distributed to HP programs utilizing an all-
objective advice to the U.S. government relating to science inclusive university email server, and the survey was distributed
and technology, has established a minimum requirement to students in NHP programs via an email from the dean of each
of 25 hours of classroom nutrition education during the respective program. The survey was distributed to programs in
preclinical years of medical school to inform future phy- September of 2018, with one reminder email sent to each program
sicians on healthy eating habits. A previous study sug- three months after initial distribution. The survey was open for re-
gested some inadequacy in such programs [10]. However, sponses until September of 2019. Google Survey (Google, Inc.) was
another study [9] from 2008 reported that personal diet used to create the survey and store response data.
improved in 60% of first-year medical students after a
health educational program was implemented. Nova The full survey is available as Supplementary Material. Ques-
Southeastern University Kiran C. Patel College of Osteo- tions were designed to cover a broad range of dietary and exercise
pathic Medicine (NSU-KPCOM) and other professional HP information while remaining simple and concise. The first six
programs currently include nutritional education in their questions focused on demographic information including the re-
respective first-year curriculums. Because HP students spondent’s sex, age, ethnicity, living situation, dietary preference
tend to receive more nutritional education than non-health (omnivore, vegetarian, vegan), and the specific professional pro-
professional (NHP) students, the primary goal of this study gram in which the respondent was enrolled. The remaining 10
was to evaluate whether this education creates differences questions focused on how often the student consumed vegetables,
between HP and NHP students’ everyday diet and exercise fruit, protein, junk food, sweets (candy, ice cream, cookies, etc.),
practices. red meat, water, and caffeine per week, as well as how often the
respondent exercised and considered making healthy food choices
To evaluate these potential differences, we elected to per week.
utilize a survey-based methodology because this
approach to collecting vast obesity trend data is utilized Each of the final 10 questions had five answer choices, with
by other organizations. Currently, the World Health Or- answer choice “A” designated as the minimum value of the question
ganization (WHO) uses self-reported surveys to collect topic being examined, and answer choice “E” as the maximum value.
much of its data pertaining to obesity [11]. Self-reporting For example, choice “A” was correlated with the least amount of ex-
is known to lead to errors in data collection, yet is the ercise time/week (0–30 min/week), and choice “E” with the most
mainstay in understanding the prevalence of obesity (120–150 min/week). Answer choices A–E were then assigned a “code
worldwide. The CDC uses a variety of survey systems to value” that was used to assess results; these “code values” were
collect their data regarding obesity. The Behavior Risk assigned in either ascending or descending order by 2s (2, 4, 6, 8, 10 or
Factor Surveillance System (BRFSS) is the largest ongoing 10, 8, 6, 4, 2), with the highest value being assigned to the preferred
telephone survey system and tracks current health con- choice. For example, answer choice A was assigned a “code value” of
ditions, risky behavior, and preventative behavior. Other two, answer choice B was assigned a “code value” of four, answer
surveys, such as the National Health and Nutrition Ex- choice C was assigned a “code value” of six, answer choice D was
amination Survey (NHANES) and the Youth Risk Behavior assigned a “code value” of eight, and answer choice E was assigned a
code value of 10 when participants were asked how many vegetables
they consumed per day, but the converse “code values” were assigned
when participants were asked how much red meat they consumed per
week, with answer choice A assigned the “code value” of 10. Of the
10 coded questions, six were devoted to different food categories
(high-protein food, red meat, sweets, fast food, fruits, and vegetables),
one to fluids (water), one to caffeine intake, one to exercise, and one to
self-assessment of diet.

Downing et al.: Dietary views and habits of students 379

Analysis of survey data Results showed no significant HP-NHP difference
(p>0.01) in any parameter including the consumption of
Comparisons between groups (HP vs. NHP) were based on the average sweets, fast food, red meat, caffeine, water, fruit, and vege-
coded values for each of the 10 coded questions. The analysis of tables, as well as healthy choices and exercise. Mean “code
variance statistical method was utilized to analyze survey data. Tests values” for each survey question are shown in Table 2. While
for normality of values were based on the Shapiro–Wilk criteria for there were no statistically-significant differences found be-
each question. Results showed that the normality criteria were not tween HP and NHP students, there were statistically-significant
satisfied for any of the questions (p<0.001). Thus, all comparisons
between groups were based on the nonparametric Mann–Whitney Table : Mean “code values” for HP and NHP respondents by survey
U test. Responses for each of the 10 questions were evaluated sepa- topic.
rately, and a p<0.01 was considered significant. A similar approach
was used for the comparison of responses for men and women; all Program comparison Sex comparison
responses were included in the analysis by sex independent of the HP NHP Men Women
program in which the respondent was enrolled.

Results Sweets . ± . . ± . . ± . **. ± .
Fast food . ± . . ± . . ± . . ± .
Of 732 total respondents, 569 (77.7%) were HP students Red meat . ± . . ± . . ± . **. ± .
and 163 (22.3%) were NHP students; respondents’ area- Caffeine . ± . . ± . . ± . . ± .
s of study are summarized in Table 1. The majority Protein . ± . . ± . . ± . **. ± .
(527; 72%) were women (205 [28%] men). Self-reported Water . ± . . ± . . ± . **. ± .
racial and ethnic distribution analysis showed that Fruit . ± . . ± . . ± . . ± .
371.9 (50.8%) respondents were White, 167 (22.8%) were Vegetables . ± . . ± . . ± . . ± .
Hispanic, 56 (7.7%) were Black or African-American, Healthy . ± . . ± . . ± . . ± .
56 (7.6%) were Asian, 44 (6.0%) selected Asian-Indian, choices
and 37.3 (5.1%) selected the “other” category from the Exercise . ± . . ± . . ± . *. ± .
survey.
*p<., statistically significant difference in values between sexes.
**p<., statistically significant difference in values between
sexes. HP, healthcare professional; NHP, non-health professional.

Table : Response distribution for students in health professions and non-health professions programs (n=).

Professional programs Respondents

n (%) Men, n (%) Women, n (%)

Health professions  (.)  (.)  (.)
Anesthesiology assistant  (.)  ()  ()
Dental medicine  (.)
Medical sciences  (.)  (.)  (.)
Nursing  (.)  (.)  (.)
Occupational therapy  (.)  (.)
Osteopathic medicine  (.)  (.)  (.)
Pharmacy  (.)  (.)  (.)
Physical therapy  (.)  (.)  (.)
Physician assistant  (.)  (.)  (.)
 (.)  (.)
Health professions total  (.)  (.)
Non-health professions  ()
 (.)  ()
College of Arts, Humanities &  (.)  (.)
Social Science  (.)  (.)  (.)
College of Business and Entrepreneurship  (.)  (.)
College of Law  ()
College of Natural Sciences and Oceanography  (.)  ()  ()
College of Psychologya   ()  (.)
Non-health professions total  (.)
Grand total  (.)
 (.)

aCollege of Psychology was deemed a non-health professions due to the universities specific categorization of programs listed under the Health
Professions Divisions title.

380 Downing et al.: Dietary views and habits of students

differences between sexes irrespective of the HP and NHP during their clerkship, because these experiences vary
programs in which they were enrolled (Table 2). Women had depending on each student’s preceptor and clinic location.
higher mean “code values” than men in weekly red meat
consumption (7.34 ± 1.9 vs. 5.96 ± 2.03; p<0.001), indicating It has been reported that only 27% of medical schools
that they consumed it fewer times per week. Women had teach the recommended 25 hours of nutritional education
lower mean “code values” than men in weekly sweet con- [12]. An average of 19.6 hours of nutritional education was
sumption (6.19 ± 1.74 vs. 6.82 ± 1.66; p<0.001), indicating incorporated throughout the four years of medical school,
that they consumed it more frequently. Women also had which accounts for less than 1% of total medical lecture
lower mean “code values” for daily water intake (6.23 ± 1.96 hours. It was also found that the majority of this nutritional
vs. 6.85 ± 1.90; p<0.001) and daily protein intake (4.41 ± 1.03 education involved medical biochemistry, rather than spe-
vs. 4.90 ± 1.23; p<0.001), indicating that they consumed both cific nutritional counseling [12]. Data suggest that 71% of
less frequently. Women also had a lower mean “code value” incoming medical students reported that they believe
for minutes of weekly exercise value than men (5.62 ± 3.0 vs. nutrition is clinically important; however, upon graduation,
6.36 ± 3.00; p<0.01), indicating that they did not exercise less than half of them maintained this opinion [12]. Beyond
as much. medical school, fewer than 14% of physicians believe that
they received appropriate training in nutritional counseling
Discussion for patients [12]. The lack of emphasis on nutritional edu-
cation in medical school may be explained by the compe-
Our hypothesis prior to administration of the survey was tition for time with other subjects in the medical curricula
that HP students would demonstrate different nutritional [11]. The introduction of new courses, such as those focused
and lifestyle choices than their NHP counterparts. The re- on nutritional education, is met with resistance not only
sults showed that for all 10 coded questions regarding diet, because there is limited time available, but also because the
lifestyle choices, and exercise, there was no statistically appropriate faculty to lead such a course are lacking clinical
significant difference between HP and NHP students. This experience [13]. Another reason for the lack of nutritional
indicates that despite differences in program curricula, HP education is that there is a greater focus on disease di-
and NHP students demonstrate similar dietary, lifestyle, agnostics and intervention rather than disease prevention
and exercise habits. [11]. Finally, even if a medical school curriculum incorporates
the 25 hours recommended by the NRC, there is no proposed
One question worthy of discussion is whether the standardized curriculum [11].
quantity of nutrition education in HP curricula is suffi-
cient. Among the four HP programs in our study with the The Association of American Medical Colleges (AAMC)
highest number of respondents, it is difficult to determine released an updated report with updated curriculum
exactly how much nutritional education each student recommendations for medical schools in 2017 [14]. These
receives. For example, students in the osteopathic recommendations were classified into three subjects of
medicine program attend a newly implemented 1-credit medical education: basic sciences, clinical sciences, and
nutrition course. The physician assistant program offers population health sciences. Basic science pertains to class-
two credit hours on Complementary Medicine & Nutrition, room education on the fundamentals of obesity. This in-
as well as three credit hours in Health Promotion and cludes the physiology of diet and hunger, energy balance,
Disease Prevention. The occupational therapy curriculum physical activity, and energy consumption. Clinical science
includes multiple courses regarding mental health and involves the practical application of nutritional and dietary
wellness, as well as environment and lifestyle, pertaining education, including topics such as calculating BMI,
to occupational medicine. The physical therapy program measuring waist circumference, and understanding the
contains three credit hours on the Essentials of Exercise secondary causes of obesity and its comorbidities. Finally,
Physiology, Health Promotion, and Wellness, along with population health science refers to the epidemiology of
a Healthcare Educator course for one credit hour. While obesity and its effect on public policy, social factors, and
all these curricula contain dedicated nutritional courses, environmental factors that influence a population’s life-
nutritional education might also be incorporated into style and nutritional choices [15, 16].
other coursework, such as biochemistry. Furthermore,
clinical rotations remain a crucial aspect of each pro- Continuity of nutritional education and its incorpora-
gram’s curriculum. It is difficult to assess the precise tion into clinical practice are two specific challenges facing
amount of nutritional education that HP students receive medical students. A systematic review published in 2012
by Vitolins et al. [16] compared separate studies regarding
medical student education on obesity intervention training.
They found that none of the studies in their review integrated

Downing et al.: Dietary views and habits of students 381

obesity education continuously throughout all four years of constructed, potential solutions should be explored. One
medical school. They also discovered a lack of published logical proposal may be to augment existing medical
evidence pertaining to medical obesity education, which curricula with an interactive, comprehensive, and free online
they believe coincides with physician reports of inadequate nutritional education system such as “Nutrition in Medicine”
training in patient weight management [16–22]. Addition- [11]. This online system provides 15-minute modules
ally, they reviewed five different studies that assessed the involving basic nutritional education as well as assessment
various methods in which students were educated on the and intervention learning, thus providing students with
application of nutritional and lifestyle information [16]. One preventative and therapeutic aspects of nutrition. This so-
of the studies, a systematic review by Vitolins et al. [16], lution would bypass the aforementioned challenges such as
included a previous study [23] examining 115 first-year the lack of appropriate faculty, resistance to incorporation
medical students and evaluating their response to two due to scheduling conflicts and time commitment, and the
interactive lectures and two standardized patient activities need for designed objectives [11]. However, even if health-
that were focused on nutrition and exercise counseling. care curricula were altered to sufficiently deliver nutritional
Of the 57 medical students who completed the pre- and and exercise education, the issue of nutrition and its clinical
post-intervention questionnaire, there was a reported in- application would remain. The failure to incorporate nutri-
crease in confidence of nutritional counseling and exercise tional recommendations into clinical practice may not be
prescription [16, 22, 23]. Carson [24] examined fourth-year due to inadequate education during medical school but
medical students who were provided a tape measure, pocket rather due to the lack of continued nutritional training
reference card with nutritional and exercise information, throughout residency and fellowship. Through their
and two computer-based patient cases. A multiple-choice research, Eisenberg and Burgess [12] discovered that the
quiz was administered to the students prior to the distribu- word “nutrition” was never mentioned on Internal Medicine
tion of the items and then again after their use in a clinical licensing board examinations. Even in cardiology fellow-
setting. Correct responses regarding the interpretation of ships, nutritional counseling was omitted from the required
waist circumference rose from 82% (pre-clerkship quiz) training [12]. Correction of America’s obesity pandemic will
to 98.5% (post-clerkship quiz) [16, 24]. require an all-encompassing team effort involving changes
made through governmental policy, healthcare reform,
Regarding the results of our study, it is likely that improved physician training, and individual patient efforts.
dietary and lifestyle habits may have already been
developed prior to the onset of graduate-level education, There are potential solutions to the ongoing issues
irrespective of HP or NHP programs. This may explain the illustrated in our research. In his book Upstream, Dan Heath
reason why no statistically significant difference was discusses how healthcare experts (namely primary care
found between HP and NHP students. After all, each physicians) are best equipped to reduce the amount of pa-
student who responded to the survey obtained at least an tient hospitalizations, emergency room visits, and
undergraduate degree, with the exception of nursing re-admissions [27]. However, primary care physicians —
students who were included in this study because nursing overcome with insurmountable pressure to meet patient
is considered a HP program. As previously mentioned, quotas and abide by insurance regulations — understandably
obesity is less prevalent in college graduates compared to struggle in preventing patient hospitalizations, ER visits, and
those with a high-school degree or equivalent [4]. A study re-admissions. The Accountable Care Organization (ACO)
comparing the prevalence of obesity between graduate was implemented in the 2010 Affordable Care Act in an
and undergraduate students would be of interest and attempt to solve the aforementioned struggles [27–29]. ACOs
may support this finding, if no statistically significant are created by coordinated, physician-formed groups. Using
differences are found between the two groups. estimates of expected hospital visits, Medicare calculates the
potential annual costs of each patient under an ACO. Thus,
The prevalence of obesity among healthcare pro- through this plan, if physicians were to reduce the amount of
fessionals is estimated at 22% [8]. It has been found that hospitalizations among their patients, the money that
physicians who were considered obese provided far less Medicare saved would be shared with the participating
nutritional and exercise information to their patients physicians. This created an incentive for physicians, who
[19, 25]. As Dietz et al. [26] explained, just as physicians started spending more time with each patient and focused
who smoke are less likely to counsel their patients about more on patients’ blood sugar levels, blood pressure, diet,
smoking, a physician’s own BMI is indicative of the like- and exercise [27]. Their focus shifted from reacting to a
lihood that they will counsel their patients who are obese. patient who was already hospital-bound to prevention
through adequate nutritional and lifestyle education [27].
If the nutritional education-related curriculum for future
healthcare professionals is indeed insufficient as currently

382 Downing et al.: Dietary views and habits of students

The ACO was estimated to reduce combined patient costs 2. Katherine M, Flegal PD. Trends in obesity among Adults in the
$470 million from 2012 to 2015 [28]. It should be noted that United States, 2005 to 2014. Available from: https://
seven of the 32 pioneer ACOs withdrew from the program jamanetwork.com/journals/jama/fullarticle/2526639 [Accessed
because no savings were demonstrated in 2012 [29]. 25 Jul 2020].
Although the perfect plan does not exist, this innovative
thinking provides promise for possible future solutions. 3. Menezes MCD, Duarte CK, Costa DVDP, Lopes MS, de Freitas PP,
Campos SF, et al. A systematic review of effects, potentialities,
There are limitations to this study. Survey results are often and limitations of nutritional interventions aimed at managing
inaccurate and experience response bias. Also, considering obesity in primary and secondary health care. Nutrition 2020;75–
that women accounted for 72.6% of survey responses, this 76:110784.
justly could be regarded as a limitation when comparing the
results of this study between men and women. 4. Centers for Disease Control and Prevention. Adult obesity
prevalence maps. Available from: https://www.cdc.gov/obesity/
Conclusions data/prevalence-maps.html [Accessed 25 Jul 2020].

The main result of this survey-based research demonstrated 5. World Health Organization. Obesity and overweight. Available
that there is no difference in dietary and exercise practices from: https://www.who.int/news-room/fact-sheets/detail/
between HP and NHP students. A lack of emphasis on obesity-and-overweight [Accessed 25 Jul 2020].
nutritional education in HP settings, especially medical
school, is a possible reason for the lack of difference in 6. Information for CME Credit. Prevalence of obesity among U.S.
dietary habits among HP and NHP students. Low physician Workers and associations with occupational factors. Am J Prev
competency regarding nutritional education results in Med. 2014;46:237–48.
abandonment of nutritional assessment and counseling for
their patients [11]. It is therefore important to provide future 7. Miller SK, Alpert PT, Cross CL. Overweight and obesity in nurses,
medical professionals adequate nutritional training. Future advanced practice nurses, and nurse educators. J Am Acad Nurse
studies should investigate not only how to improve nutri- Pract 2008;20:259–65.
tional education among potential medical professionals,
but also how to develop methods that will allow them to 8. Alba LH, Badoui N, Gil F. Attitude toward preventive counseling
translate this understanding into their future practice. and healthy practices among medical students at a Colombian
university. Front Med 2015;9:251–9.
Research funding: None reported.
Author contributions: All authors provided substantial 9. Barss P, Grivna M, Al-Maskari F, Kershaw G. Strengthening public
contributions to conception and design, acquisition of data, health medicine training for medical students: development and
and analysis and interpretation of data; all authors drafted evaluation of a lifestyle curriculum. Med Teach 2008;30:e196–
the article or revised it critically for important intellectual 218.
content; all authors gave final approval of the version of
the article to be published; and all authors agree to be 10. Adams KM, Kohlmeier M, Powell M, Zeisel SH. Nutrition in
accountable for all aspects of the work in ensuring that medicine. Nutr Clin Pract 2010;25:471–80.
questions related to the accuracy or integrity of any part of
the work are appropriately investigated and resolved. 11. Bauman A, Rutter H, Baur L. Too little, too slowly: international
Competing interests: Authors state no conflict of interest. perspectives on childhood obesity. Publ Health Res Pract 2019;
Informed consent: Informed consent was not obtained. 29:2911901.
Ethical approval: This study was reviewed and determined
exempt by the Institutional Review Board at Nova 12. Eisenberg DM, Burgess JD. Nutrition education in an era of global
Southeastern University (IRB No. 2018-455-NSU). obesity and diabetes. Acad Med 2015;90:854–60.

References 13. Kris-Etherton PM, Akabas SR, Bales CW, Bistrian B, Braun L,
Edwards MS, et al. The need to advance nutrition education in the
1. Centers for Disease Control and prevention. Surveillance training of health care professionals and recommended research
systems. Available from: https://www.cdc.gov/obesity/data/ to evaluate implementation and effectiveness. Am J Clin Nutr
surveillance.html [Accessed 25 Jul 2020]. 2014;99:1153S–66S.

14. Maeshiro R. Contemporary issues in medicine: the prevention
and treatment of overweight and obesity. Association of
American Medical Colleges; 2007. Available from: https://store.
aamc.org/downloadable/download/sample/sample_id/57/.

15. BSE. Expert committee recommendations regarding the
prevention, assessment, and treatment of child and adolescent
overweight and obesity: summary report. Available from: https://
pubmed.ncbi.nlm.nih.gov/18055651/ [Accessed 25 Jul 2020].

16. Vitolins MZ, Crandall S, Miller D, Ip E, Marion G, Spangler JG. Obesity
educational interventions in U.S. Medical schools: a systematic
review and identified gaps. Teach Learn Med 2012;24:267–72.

17. Kolasa K. Developments and challenges in family practice
nutrition education for residents and practicing physicians: an
overview of the North American experience. Eur J Clin Nutr 1999;
53:s89–96.

18. Beno L, Hinchman J, Kibbe D, Trowbridge F. Design and
implementation of training to improve management of pediatric
overweight. J Continuing Educ Health Prof 2005;25:248–58.

19. Kopelman P. Developing an action plan for obesity education. Int J
Obes 2001;25:S16–9.

Downing et al.: Dietary views and habits of students 383

20. Bruer RA, Schmidt RE, Davis H. Commentary: nutrition counseling cessation practices, and beliefs. Nicotine Tob Res 2010;12:
—should physicians guide their patients? Am J Prev Med 1994;10: 724–33.
308–11. 26. Dietz WH, Baur LA, Hall K, Puhl RM, Taveras EM, Uauy R, et al.
Management of obesity: improvement of health-care training
21. Walsh CO, Ziniel SI, Delichatsios HK, Ludwig DS. Nutrition and systems for prevention and care. Lancet 2015;385:
attitudes and knowledge in medical students after completion of 2521–33.
an integrated nutrition curriculum compared to a dedicated 27. Heath D. Upstream: the quest to solve problems before they
nutrition curriculum: a quasi-experimental study. BMC Med Educ happen. New York: Avid Reader Press; 2020.
2011;11:58. 28. Federal Register. Medicare Program; Medicare shared savings
program; accountable care organizations-pathways to success.
22. Block JP, Desalvo KB, Fisher WP. Are physicians equipped to Available from: https://www.federalregister.gov/documents/
address the obesity epidemic? Knowledge and attitudes of 2018/08/17/2018-17101/medicare-program-medicare-shared-
internal medicine residentsਮ. Prev Med 2003;36:669–75. savings-program-accountable-care-organizations-pathways-to-
success [Accessed 25 Jul 2020].
23. Iii PFB, Stetson BA, Rising W, Wesley GC, Ritchie CS. Development 29. Mcwilliams JM. Changes in Medicare shared savings program
and evaluation of a nutrition and physical activity counseling savings from 2013 to 2014. Jama 2016;316:1711.
module for first-year medical students. Med Educ Online 2004;9:
4359. Supplementary Material: The online version of this article offers sup-
plementary material (https://doi.org/10.1515/jom-2020-0178).
24. Carson JAS. Pocket tape measure for waist circumference:
training medical students and residents on a simple assessment
of body composition. J Nutr 2003;133:547S–9S.

25. Tong EK, Strouse R, Hall J, Kovac M, Schroeder SA. National
survey of U.S. health professionals’ smoking prevalence,



J Osteopath Med 2021; 121(4): 385–390

Medical Education Original Article

Tyler Hamby*, PhD, W. Paul Bowman, MD, Don P. Wilson, MD and Riyaz Basha, PhD

Mentors’ experiences in an osteopathic medical
student research program

https://doi.org/10.1515/jom-2020-0251 productivity was a motivation, and hospital-based mentors
Received September 21, 2020; accepted November 24, 2020; were statistically significantly more likely to endorse this
published online February 23, 2021 source of motivation (OR=2.02; 95% CI=1.18–3.45; p=0.01).
Most respondents were satisfied with the quality of the
Abstract students' work (59 [84.3%]) and with the program (59
[85.5%]). However, 46 (65.7%) suggested the program
Context: Medical students, especially at osteopathic medi- could be enhanced by requiring medical students to be
cal schools, have limited research exposure. Systematic in- physically present in the clinic or laboratory for a minimum
struction in research, supervised by qualified mentors, amount of time. Importantly, most (58 [84.1%]) mentors
could motivate osteopathic medical students to pursue reported that they would be interested in participating in
research in their careers, thereby increasing the number of future mentored research programs.
future clinician-scientists. Recruiting and retaining suitable Conclusions: Mentors were motivated to participate in
research mentors are crucial to sustaining such programs, the voluntary research program for both altruistic and
but this task is also particularly challenging for osteopathic professional reasons. Since most mentors reported being
medical schools. satisfied with the program, it is likely they would partici-
Objectives: To assess mentors' experiences in a voluntary pate in future mentored research programs. Our results
student-mentor medical research program. suggest that mentors viewed this voluntary research pro-
Methods: An online survey was sent to 76 university- or gram as mutually beneficial.
hospital-based participants who previously mentored 219
medical students between 2014 and 2019. The questionnaire Keywords: medical education; medical student; mentoring;
consisted of 13 items with responses in checklist, five-point research.
Likert scale, and categorical multiple-choice formats,
assessing motivation for participation, satisfaction with the Educating medical students in the art and science of medical
program, and interest in future participation. Data were research and providing research opportunities are potential
analyzed descriptively, and responses from mentors at the solutions to the declining number of clinician-scientists.
university and hospital were compared using univariate Medical students are frequently interested in medical
logistic and ordinal regression analyses. research [1, 2]. In a previous meta-analysis [1], three or four
Results: Among 70 (92.1%) mentors who responded to the studies showed each of the following results: students who
survey, 61 (87.1%) reported being motivated by a desire to conducted research in medical school were more likely to
help medical students learn research. Forty-nine (70.0%) be interested in conducting research in their careers and
mentors indicated that furthering their own research pursue academic careers; later in their careers, they were
more likely to attain faculty rank, conduct research, and
*Corresponding author: Tyler Hamby, PhD, Department of Research publish research [1]. However, in 2019, 43% of college
Operations, Cook Children's Health Care System, 801 7th Ave, Fort osteopathic medicine (COM) graduates reported receiving
Worth, 76104-2796, TX, USA; and Texas College of Osteopathic inadequate training in research techniques [3]. Indeed,
Medicine, University of North Texas Health Science Center, Fort Worth, compared with their peers in allopathic medical schools,
TX, USA. E-mail: [email protected] COM students graduate with fewer scholarly works,
W. Paul Bowman, MD and Riyaz Basha, PhD, Texas College of including peer-reviewed publications and presentations,
Osteopathic Medicine, University of North Texas Health Science which may adversely affect their ability to compete for res-
Center, Fort Worth, TX, USA idency training programs [4].
Don P. Wilson, MD, Department of Pediatric Endocrinology, Cook
Children's Health Care System, Fort Worth, TX, USA The systematic training of medical students in research
is costly and such programs are difficult to sustain. Among

Open Access. © 2020 Tyler Hamby et al., published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0
International License.

386 Hamby et al.: Mentors' experiences in OMS research program

other barriers, there is a shortage of available mentors to students also work with other researchers and/or
provide appropriate guidance and supervision [5]. Yet, healthcare staff associated with the study.
focus groups of medical students have indicated that
they view mentors as being critical in influencing them to Students begin their research experience during the
pursue careers as clinician-scientists [6]. This shortage of summer break between their first and second years of med-
mentors may be more common among COMs. For instance, ical school. For one month, they work closely with a mentor
in surveys of all 2019 allopathic medical school [7] and COM on study activities (e.g., literature review, data collection,
[3] graduates in the US, respondents from COMs were protocol development, and manuscript writing). All students
less likely to report feeling satisfied or very satisfied with give an oral presentation of their findings in the fall and are
faculty mentoring in general (52 vs. 81%). Furthermore, required to submit their abstract to at least one conference.
enthusiastic, qualified physicians who wish to serve as Many students present their work at local, regional, and
research mentors at COMs may lack the protected time and national professional conferences, and several students
resources necessary for research activities [8]. Compared have had their work published in peer-reviewed journals.
with allopathic medical schools, COMs have fewer faculty Long-term mentor-mentee relationships are encouraged and
members [9, 10]; faculty members have fewer articles often continue until students graduate or beyond.
published and fewer citations per article [11, 12]; and they
receive less grant funding from the National Institutes of The purpose of the present study was to examine the
Health [13]. Thus, understanding the expectations and experiences of research mentors.
perceptions of research mentors is critical to designing
sustainable mentored student research programs at COMs Methods
[14]. Physicians who voluntarily participate but do not view
the experience positively are unlikely to continue in the An online survey (Supplemental Material) was developed by PRP
program [14]. Ideally, the program should be mutually managers (T.H. and R.B.) in consultation with mentors; research staff,
beneficial for both the student and the mentor, particularly who support mentors' research, served as subject-matter experts to
if mentors have limited time or resources to conduct help validate the survey's content. The questionnaire contained 13
research. A mutually beneficial program can simulta- items with responses in checklist, five-point Likert scale, and cate-
neously address the clinician-scientist shortage on two gorical multiple-choice formats that asked for mentors' level of
fronts: training future and assisting current investigators. satisfaction with the program, feedback on the structure of the PRP,
and motivations for participating. All mentors from the 2014 through
Since 2014, medical students at the Texas College of 2017 cohorts were surveyed in September 2017. Mentors from the 2018
Osteopathic Medicine (TCOM), University of North Texas and 2019 cohorts were surveyed in August 2018 and July 2019,
Health Sciences Center (UNTHSC) have been participating respectively. To improve the response rate, the survey was emailed to
in the Pediatric Research Program (PRP) guided by men- the mentors up to four times at two week intervals, and those who did
tors located at two participating training sites: the main not respond were subsequently given printed surveys. Because part
campus of UNTHSC and Cook Children's Medical Center of the purpose of the survey was to identify interested mentors for
(CCMC), which is a large pediatric healthcare system. future cohorts, it was not anonymous. The survey was developed in
Students who apply to participate in the PRP are inter- Qualtrics (Qualtrics, LLC).
viewed and successful candidates are selected each year,
depending on the availability of mentors. Research Respondent characteristics and responses to quantitative items
mentors include university faculty, community-based were described with frequencies and percentages. The impact of
practicing physicians, and other healthcare pro- location (CCMC vs. UNTHSC) on responses was analyzed using uni-
fessionals who voluntarily participate. Following an variate logistic regression analyses. For Likert-type items, ordinal
orientation, mentors are coached by qualified and expe- regression with a proportional odds model was used. Inferential an-
rienced PRP faculty (R.B., W.P.B., D.W., and T.H.) and alyses were checked for test assumptions, and a p-value <0.05 was
staff who oversee the program. Students assigned to used as a cutoff for statistical significance. Data were analyzed in SAS
CCMC work with a fulltime project manager (T.H.) who Enterprise (version 6.1; SAS Institute Inc.). Graphs were developed
helps coordinate meetings and assists mentors and stu- using RStudio (RStudio, Inc.). For mentors who responded to surveys
dents with study design, data collection, statistical anal- in multiple years, the most recent survey was used. This study was
ysis, publication, and presentation. Every student has one approved by the CCMC institutional review board (IRB) as a quality
mentor—the principal investigator of the study —who improvement project and determined exempt by the UNTHSC IRB.
takes primary responsibility for guiding the student dur-
ing his or her participation in research study. Most Results

Surveys were sent to 76 mentors (CCMC, 49 [70.0%];
UNTHSC, 27 [30.0%]) who had mentored 219 (CCMC, 139
[63.5%]; UNTHSC, 80 [36.5%]) students. Responses were

Hamby et al.: Mentors' experiences in OMS research program 387

obtained from 70 mentors (31 [44.3%] men; 39 [55.7%] Table : Mentors' reported reasons for working with PRP student(s).
women), yielding a 92.1% response rate (RR): 43 (61.4%;
RR=87.8%) from CCMC and 27 (38.6%; RR=100.0%) from Reasons CCMC UNTHSC OR (% CI) p-value
UNTHSC. Among respondents, 38 (54.3%) were physicians,
19 (27.1%) held a PhD, and 13 (18.6%) were other pro- To help student(s) learn research . (.–.) .
fessionals (five [7.1%] nurses, three [4.3%] pharmacists, three Yes  (.%)  (.%) . (.–.) .
[4.3%] DrPH, one [1.4%] dietitian, and one [1.4%] research No  (.%)  (.%) . (.–.) .
associate). . (.–.) .
For help starting research . (.–.) .
With regard to mentors' motivations for working with Yes  (.%)  (.%) . (.–.) .
medical students, 61 (87.1%) reported a desire to help stu- No  (.%)  (.%)
dents learn how to conduct research. A minority of mentors
endorsed any specific reason involving wanting research For help with data entry
support—starting research (30 [42.9%]), data entry (25 Yes  (.%)  (.%)
[35.7%]), writing a manuscript(s) (21 [30.0%]), and research No  (.%)  (.%)
staff support (13 [18.6%])—yet, the majority of respondents
(49 [70.0%]) selected at least one of these four professional For help with writing manuscript(s)
motivations. Yes  (.%)  (.%)
No  (.%)  (.%)
Most mentors were “very satisfied” or “somewhat
satisfied” with the quality of the work produced by the For research support staff
student(s) (59 [84.3%]) and with the structure of the program Yes  (.%)  (.%)
(59 [85.5%]). However, when asked whether students should No  (.%)  (.%)
be required to be present in the clinic or laboratory a mini-
mum number of hours each week during the summer, Professional motivationa
the majority answered “yes” (46 [65.7%]) or “don't know” Yes  (.%)  (.%)
(10 [14.3%]). Most respondents reported that they were “very No  (.%)  (.%)
interested” or “somewhat interested” in mentoring future
medical students for the PRP (58 [84.1%]) and, to a lesser aThese include help starting research, with data entry, with writing
degree, for research rotations (46 [65.7%]). manuscript(s), and/or research support staff. CI, confidence interval;
CCMC, Cook Children's Medical Center; OR, odds ratio; PRP, Pediatric
Using logistic regression analyses, mentors from Research Program; UNTHSC, University of North Texas Health
CCMC and UNTHSC were compared in their responses to Sciences Center.
questions about motivation, satisfaction, and interest.
Responses did not statistically significantly differ by research support staff. Mentors at CCMC were statistically
location in reporting motivations to help students, for significantly more likely to report wanting help in writing
help in starting research, for help in data entry, or for manuscript(s) (OR=1.94; 95% CI=1.05–3.58; p=0.03), and
they were more likely to select at least one professional
motivation involving research support (OR=2.02; 95%
CI=1.18–3.45; p=0.01; Table 1). There were no statistically
significant differences between locations in satisfaction
with students' work or the structure of the program
(Figure 1), and there were no significant differences

Figure 1: Mentors' satisfaction with the
pediatric research program (PRP) stratified
by location.
CCMC, Cook Children's Medical Center;
PRP, Pediatric Research Program; UNTHSC,
University of North Texas Health Sciences
Center.

388 Hamby et al.: Mentors' experiences in OMS research program

Figure 2: Mentors' future interest in the
pediatric research program (PRP) stratified
by location.
CCMC, Cook Children's Medical Center;
PRP, Pediatric Research Program; UNTHSC,
University of North Texas Health Sciences
Center.

between locations in interest for working with students in methodological assistance). University-based clinical fac-
the PRP or research rotations (Figure 2). ulty may lack protected time for research and many lack
research staff support, but community clinicians typically
Discussion have even less time and support for research than their ac-
ademic colleagues [19]. Therefore, it is not surprising that,
The PRP provides an opportunity for osteopathic medical compared with university faculty, hospital clinicians were
students to conduct extracurricular research concurrently significantly more likely to indicate that they participated in
with their medical school training. Students are mentored the PRP to receive assistance with their research. Mentors
by qualified faculty, community-based physicians, or will be more likely to participate in the future if the program
other healthcare professionals who voluntarily partici- genuinely provides assistance with their research. Fortu-
pate. The primary objective of the program is to introduce nately, most mentors in our study—whether at the university
medical students to the principles of research. For such a or at the hospital—were satisfied with their students' work.
program to be sustainable, volunteering mentors must Finally, in line with the above results, most mentors in our
view it as being worthwhile and, ideally, beneficial to study expressed a willingness to continue their participa-
themselves and to participating students [14, 15]. Our tion, which will likely help sustain the PRP for future medical
study explored mentors' perceptions of their PRP experi- students.
ences by examining their responses to a self-report online
survey. While most mentors reported being satisfied with the
structure of the program, most mentors also thought that
For students to learn and to have a positive experience, students should be required to be on campus a minimum
mentors ideally should offer guidance [5] and encourage number of hours per week. Because this program is
students to take an active role rather than simply collecting voluntary for students, having a universal attendance
data [16]. Consistent with prior research on mentoring stu- policy may result in students becoming less interested in
dents in clinical practice [17], most mentors in our survey participating, resulting in attrition. Nonetheless, students
reported that they participated to help students. In our study, are informed that some mentors and projects will demand
most mentors also indicated that they chose to work with more time than others. In our subjective experience, many
students as a means of obtaining assistance with their own of the students selected to participate in the PRP exceed the
research (i.e., starting research, data entry, writing, staff recommended time guidelines and take advantage of the
support), though the specific reason or reasons given program's flexible work schedule. These considerations
differed by mentor. It is critically important to understand underscore the importance of matching students with
what motivates a research mentor, to increase the likelihood suitable mentors [20]. It is important to match students'
of future participation [14, 18]. The PRP helps mentors strengths and availabilities with mentors' needs. Perhaps
by having medical students and research staff complete these factors should be weighted more heavily than the
tasks that busy clinicians may have insufficient time to students' interests and the mentors' fields; these consid-
complete (e.g., literature reviews, statistical analysis, or erations have received greater focus in selecting and
matching students in recent cohorts.

Hamby et al.: Mentors' experiences in OMS research program 389

Websites for some COMs describe mentored student achieving success and sustainability. Our mentor re-
research programs, but few have published the results spondents were motivated to participate in the voluntary
of their programs; the present data is a step toward filling research program for both altruistic and professional
this gap in the literature. Our results provide evidence reasons; since most mentors reported being satisfied
of success in providing a mentored student research with the program, it is likely they would participate in
program. Although limited to a single academic institu- future mentored research programs. Our results suggest
tion, these findings will hopefully assist other COMs in that mentors viewed this voluntary research program as
achieving similar success. The PRP provides a unique and mutually beneficial, so use of qualified volunteer men-
cost-effective solution to the problem of limited qualified tors is feasible, although more research is clearly
mentors in COMs, as it relies on a close and mutually needed.
beneficial collaboration with a local hospital. Although
information about COM-sponsored research programs Acknowledgments: This study was conducted as part of the
for medical students is limited, there is a program at University of North Texas Health Science Center and Cook
Ohio University Heritage College of Osteopathic Medi- Children's Pediatric Research Program (PRP).
cine (OU-HCOM) that collaborates with several outside Research funding: None reported.
institutions [21]. In addition to medical students, the Author contributions: All authors provided substantial
OU-HCOM program includes residents and fellows from contributions to conception and design, acquisition of data,
numerous institutions. Both the OU-HCOM program and or analysis and interpretation of data; all authors drafted the
the PRP utilize research personnel to support the stu- article or revised it critically for important intellectual content;
dents and physicians, who often lack time for or expertise all authors gave final approval of the version of the article to be
in research. published; and all authors agree to be accountable for all
aspects of the work in ensuring that questions related to the
Limitations accuracy or integrity of any part of the work are appropriately
investigated and resolved.
While our results are promising, they must be interpreted Competing interests: Authors state no conflict of interest.
with an appropriate understanding of the limitations. Most Informed consent: Participants in this survey study
importantly, the present results reflect a mentored student provided informed consent.
research program from only one COM. During the six years Ethical approval: This study was approved by the
covered in this study, the PRP has been a very successful Institutional Review Board at Cook Children's Health Care
program for 219 students who worked under 76 mentors. System as a quality improvement project and deemed
Still, the extent to which the present results are generaliz- exempt by the Institutional Review Board at University of
able to other institutions is unclear. Owing to a dearth of North Texas Health Science Center.
relevant publications, it was not possible to compare our
results to that of other COMs. It is important that more COMs References
publish information that describes their program as well as
share helpful insights and “lessons learned” with other 1. Amgad M, Man Kin Tsui M, Liptrott SJ, Shash E. Medical student
osteopathic academic medical centers. Further research is research: an integrated mixed-methods systematic review and
needed to facilitate a better understanding of the scope and meta-analysis. PloS One 2015;10:e0127470.
success of all COMs mentored student research programs.
Such an endeavor would provide valuable data and help to 2. Nguyen V, Kaneshiro K, Nallamala H, Kirby C, Cho T, Messer K, et al.
improve existing programs. Assessment of the research interests and perceptions of first-year
medical students at 4 colleges of osteopathic medicine. J Am
Conclusions Osteopath Assoc 2020;120:236–44.

A structured research program utilizing voluntary men- 3. American Association of Colleges of Osteopathic Medicine.
tors is an inexpensive yet productive means of teaching AACOM 2018-2019 academic year survey of graduating seniors
medical students the value of medical research. Under- summary report. Available from: https://www.aacom.org/docs/
standing the motivation and satisfaction of mentors who default-source/data-and-trends/aacom-2018-2019-academic-
participate in such a program is critically important in year-graduating-seniors-survey-summary-report.pdf?
sfvrsn=b5310e97_6 [Accessed 16 Oct 2020].

4. Matthews CN, Estrada DC, George-Weinstein M, Claeson KM,
Roberts MB. Evaluating the influence of research on match success
for osteopathic and allopathic applicants to residency programs.
J Am Osteopath Assoc 2019;119:588–96.

390 Hamby et al.: Mentors' experiences in OMS research program

5. Siemens DR, Punnen S, Wong J, Kanji N. A survey on the attitudes 14. Stenfors-Hayes T, Kalen S, Hult H, Dahlgren LO, Hindbeck H,
towards research in medical school. BMC Med Educ 2010;10:4. Ponzer S. Being a mentor for undergraduate medical students
enhances personal and professional development. Med Teach
6. National Institutes of Health. Physician-scientist workforce working 2010;32:148–53.
group report. Available from: https://acd.od.nih.gov/documents/
reports/PSW_Report_ACD_06042014.pdf [Accessed 16 Oct 2020]. 15. Straus SE, Johnson MO, Marquez C, Feldman MD. Characteristics
of successful and failed mentoring relationships: a qualitative
7. Association of American Medical Colleges. Medical school study across two academic health centers. Acad Med 2013;88:
graduation questionnaire: 2019 All schools summary report. 82–9.
Available from: https://www.aamc.org/system/files/2019-08/
2019-gq-all-schools-summary-report.pdf [Accessed 16 Oct 2020]. 16. Murdoch-Eaton D, Drewery S, Elton S, Emmerson C, Marshall M,
Smith JA, et al. What do medical students understand by
8. Degenhardt BF, Standley PR. Strategic plan for research: a role for research and research skills? Identifying research
everyone in promoting research in the osteopathic medical opportunities within undergraduate projects. Med Teach 2010;
profession. J Am Osteopath Assoc 2013;113:654–9. 32:e152–160.

9. 2019 U.S. Medical School Faculty. American Association of 17. May M, Mand P, Biertz F, Hummers-Pradier E, Kruschinski C. A
Medical Colleges website. Available from: https://www.aamc. survey to assess family physicians’ motivation to teach
org/data-reports/faculty-institutions/interactive-data/2019-us- undergraduates in their practices. PloS One 2012;7:e45846.
medical-school-faculty [Accessed 19 Jan 2021].
18. Ramani S, Gruppen L, Kachur EK. Twelve tips for developing
10. AACOM Reports on FacultyAcademic Administrative Staff. effective mentors. Med Teach 2006;28:404–8.
American Association of Colleges of Osteopathic Medicine
website. Available from: https://www.aacom.org/docs/default- 19. Young RA, Dehaven MJ, Passmore C, Baumer JG. Research
source/data-and-trends/2018-19-osteopathic-medical-college- participation, protected time, and research output by family
basic-science-faculty-by-rank.xlsx?sfvrsn=1d30e97_10 physicians in family medicine residencies. Fam Med 2006;38:
[Accessed 19 Jan 2021]. 341–8.

11. Suminski RR, Hendrix D, May LE, Wasserman JA, Guillory VJ. 20. Straus SE, Chatur F, Taylor M. Issues in the mentor-mentee
Bibliometric measures and National Institutes of Health funding relationship in academic medicine: a qualitative study. Acad Med
at colleges of osteopathic medicine, 2006–2010. J Am Osteopath 2009;84:135–9.
Assoc 2012;112:716–24.
21. Brannan GD. Growing research among osteopathic residents
12. Hendrix D. An analysis of bibliometric indicators, National and medical students: a consortium-based research
Institutes of Health funding, and faculty size at Association of education continuum model. J Am Osteopath Assoc 2016;116:
American Medical Colleges medical schools, 1997–2007. J Med 310–5.
Libr Assoc 2008;96:324.
Supplementary Material: The online version of this article offers
13. Clark BC, Blazyk J. Research in the osteopathic medical supplementary material (https://doi.org/10.1515/jom-2020-0251).
profession: roadmap to recovery. J Am Osteopath Assoc 2014;114:
608–14.

J Osteopath Med 2021; 121(4): 391–400

Musculoskeletal Medicine and Pain Original Article

Musa S. Danazumi*, PT, MSc, Bashir Bello, PT, PhD, Abdulsalam M. Yakasai, PT, PhD
and Bashir Kaka, PT, PhD

Two manual therapy techniques for management
of lumbar radiculopathy: a randomized clinical
trial

https://doi.org/10.1515/jom-2020-0261 all outcomes (p=0.001). Participants receiving combined
Received October 5, 2020; accepted December 10, 2020; SMWLM + PINS treatment experienced greater improve-
published online February 26, 2021 ment in leg pain, back pain, disability, and sciatica at all
timelines (immediately posttreatment, and three, six, and
Abstract nine months follow-up) than the participants receiving
SMWLM or PINS alone (p<0.05). However, participants
Context: Evidence has shown that spinal mobilization receiving SMWLM alone showed better improvement than
with leg movement (SMWLM) and progressive inhibition of the participants receiving PINS alone at all timelines
neuromuscular structures (PINS) are individually effective (p<0.05).
in the management of lumbar radiculopathy. However, Conclusions: A combined SMWLM + PINS treatment pro-
previous evidence reported data for only a short-term study tocol showed greater improvement than the individual
period and did not investigate the effect of the combined techniques alone in the management of individuals with
manual therapy techniques. LR in this study.
Objectives: To compare the combined effects of two
manual therapy techniques (SMWLM and PINS) with the Keywords: lumbar disc herniation; lumbar radiculopathy;
individual techniques alone (SMWLM or PINS) in the neuromuscular inhibition; spinal mobilization.
management of individuals with lumbar radiculopathy.
Methods: A total of 60 patients diagnosed with unilateral Lumbar radiculopathy is characterized by radiating pain
lumbar radiculopathy secondary to disc herniation were in an area of the leg typically served by one lumbar or
randomly allocated into three groups: 20 participants each sacral spinal nerve root in combination with dermatomal
in the SMWLM, PINS, and combined SMWLM + PINS and/or tendon reflex abnormalities [1]. The problem is
groups. Each group attended two treatments per week for commonly caused by degenerative conditions, such as a
30 min each, for three months. Participants were assessed herniated nucleus pulposus or lumbar spinal stenosis,
at baseline, immediately posttreatment, and then at three, and occurs in approximately 3–5% of the population,
six, and nine months follow-up using the Visual Analog with about 10–25% of those affected having a recurrence
Scale (VAS), Rolland-Morris Disability Questionnaire of symptoms that usually persist for more than three
(RMDQ), and Sciatica Bothersomeness Index (SBI). months [2].
Results: Between-groups analyses using a two-way
repeated-measures analysis of variance indicated signifi- Studies have indicated that conservative manage-
cant interactions between groups and follow-up times for ment of lumbar radiculopathy should be attempted in the
absence of worsening neurological signs or cauda equina
*Corresponding author: Musa S. Danazumi, PT, MSc, Department of syndrome before surgical interventions [3, 4]. Based on
Physiotherapy, Federal Medical Center, Nguru, 630101, Yobe State, this, many forms of conservative care, including manual
Nigeria, E-mail: [email protected] therapy, have been developed and investigated in the
Bashir Bello, PT, PhD and Bashir Kaka, PT, PhD, Department of management of this condition [5–7]. Moreover, evidence
Physiotherapy, Faculty of Allied Health Sciences, College of Health from international guidelines [8–10] and systematic re-
Sciences, Bayero University, Kano, Nigeria views [11–13] indicate that manual therapy techniques are
Abdulsalam M. Yakasai, PT, PhD, Medical Rehabilitation Therapists effective for both short and long-term pain and disability
(Reg.) Board of Nigeria, North-West Zonal Office, Kano, Nigeria; and management in patients with back-related lower extrem-
Department of Physiotherapy, College of Health Sciences, University ity symptoms. Despite this, there are no standardized
of KwaZulu-Natal, Durban, South Africa

Open Access. © 2020 Musa S. Danazumi et al., published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0
International License.

392 Danazumi et al.: Randomized trial of manual therapy for lumbar radiculopathy

guidelines for appropriate manual therapy techniques factors) analysis of variance (ANOVA). According to a previous study
comprising the most effective interventions for in- [23], a 20% attrition rate should be accounted for a clinical trial of
dividuals with lumbar radiculopathy, suggesting that equal allocation to groups, so our the sample size was adjusted to 60
more treatment options are needed [8–17]. participants (20 per group) to anticipate this potential loss due to
attrition.
A previous study [18] investigated the effect of two
manual therapy techniques, spinal mobilization with Eligibility criteria
leg movement (SMWLM) and progressive inhibition of
neuromuscular structures (PINS), in the management Participants diagnosed as having a chronic (lasting three months or
of lumbar radiculopathy; results showed that both longer) [2–4] lumbar radiculopathy secondary to L4/L5 and L5/S1
SMWLM and PINS were individually effective on pain, lumbar disc herniation (confirmed on magnetic resonance imaging)
disability, and sciatica, but no significant differences were included in the study. Patients with chronic lumbar radiculopathy
were observed between the two manual therapy tech- were selected because of the prior evidence that acute cases may resolve
niques postintervention. That study [18] reported data for spontaneously within a few weeks of onset [2–4]. Adults with an age
a short-term period (eight weeks postintervention) and range of 18–65 years having unilateral radiculopathy and pain in the
did not investigate combined manual therapy techniques. distribution of sciatic nerve (leg-dominant symptoms) were included.
Given that SMWLM has been hypothesized to relieve nerve Pain was accepted as evidence of L5/L4 root compression when
compression [19] and PINS has been hypothesized to distributed to the anterolateral aspect of the leg and the dorsum of the
normalize reflex activity in the neuromuscular structures foot, and as evidence of L5/S1 root compression when distributed to the
[20], we hypothesized that a combination of the two posterior aspect of the leg extending to the heel and lateral aspect of the
techniques could produce a better outcome than the foot. Radiating pain was evoked by specific clinical tests, including
individual techniques alone. Accordingly, this study was slump, straight leg raising [24–26], and Lasègue’s sign [27]. The exclu-
conducted to examine the combined effect of SMWLM and sion criteria were: diagnosis of lumbar disc herniation with bilateral
PINS in the management of individuals with lumbar radiculopathy, diagnosis of other back conditions (e.g., failed back
radiculopathy secondary to disc herniation. syndrome, spondylolysis/spondylolisthesis, spinal stenosis, mechani-
cal low back pain or neoplasmic and infectious processes), and par-
Methods ticipants currently receiving treatments. Participants with severe nerve
root compression (non-ambulant/wheelchair-bound) or having cauda
Research design and ethics equine syndrome were also excluded because it was believed that these
participants might better benefit from surgery [28].
This was a single-blind, parallel, randomized, clinical trial; the research
participants were blinded to randomization. Because of the nature of Data collection procedure
the study, investigators were not blinded, but outcomes assessors and
data analysts were blinded. Ethical approval to conduct this study was Participant recruitment: Participants were referred from the general
obtained from the Health Research Ethics Committee of the Federal outpatient department (GOPD) and surgical outpatient department
Medical Centre (FMC), Nguru, Yobe State, Nigeria (Registration number: (SOPD) to the department of physiotherapy where the study was con-
FMC/N/CL.SERV/355/VOL.IV/139). Written informed consent was ob- ducted. Participants were verbally invited to participate in the study and
tained from the participants, they were assured of their rights to with- then screened for eligibility by a trained orthopedic physiotherapist
draw from the study at any point in time, and their full anonymity was (M.S.D.). Participants who met the eligibility criteria and were willing to
maintained. The trial was registered with the Pan African Clinical Trial participate were recruited, and written informed consents were obtained.
Registry (PACTR201907782710996) on July 11, 2019. The study was Participants were recruited consecutively throughout the study period
begun on July 15, 2019 and ended on September 11, 2020. until the sample size needed was achieved.

Sample size estimation Outcomes assessment: Outcomes were assessed at baseline and
posttreatment (after three months of intervention). Treatment ended
The sample size was calculated using the G*Power version 3.1 software after three months of intervention and then patients were given lum-
[21]. The effect size used for calculating the sample size was obtained bar stabilization and stretching exercises [18] as a home regimen.
from previous research by Das et al. [22] using disability as the primary Outcomes were further assessed through physical visits and phone
outcome. The probability level (α), the power (p), the effect size (ES), calls at three, six, and nine months of follow-up.
the number of groups, the number of measurements, and correlation (1) Visual Analog Scale (VAS): This scale was used to measure leg and
among repeated measures used for the calculation were set at 0.05,
0.8, 0.38, 3, 6, and 0.5, respectively, which yielded a total sample size back pain intensities experienced by the participants. The assess-
of 45 participants (15 per group), using repeated measures (between ment was based on a horizontal 10 cm scale varying from 0 (no pain
in the leg) to 10 cm (the worst pains ever). VAS has been shown to
have high interobserver reliability coefficient (r=0.88) and a 30%
change is considered clinically significant [29].
(2) Roland Morris Disability Questionnaire (RMDQ): This questionnaire
was used to measure functional limitation among participants.
Scores range from 0 to 24, with higher scores reflecting more severe

Danazumi et al.: Randomized trial of manual therapy for lumbar radiculopathy 393

disability. The RMDQ has been shown to have a high level of in- least sensitivity (endpoint) was found, a moderate ischemic
ternal consistency reliability (Cronbach’s alpha of 0.87), and test- compression was applied with the index finger of one hand without
retest reliability (ICC of 0.9); 2–3 (8–12%) point change is consid- relieving the pressure until the treatment is complete. When an area of
ered clinically significant [29]. most sensitivity was found (primary point), a moderate ischemic
(3) Sciatica Bothersomeness Index (SBI): This scale was used to assess compression was also applied with the index finger of the other hand
the level of sciatica among participants. The scale’s scores range for 30 s, after which another sensitive point was identified with the
from 0 to 24, with higher scores reflecting more severe sciatica middle finger of the same hand proximal to the endpoint without
bothersomeness. This scale has a high level of internal consis- relieving the pressure of the index finger. When the pressure on the
tency reliability (Cronbach’s alpha of 0.70) and test–retest reli- middle finger was noted as more sensitive than on the index finger,
ability (ICC of 0.90); a change of 6.5 points is considered clinically then the index finger pressure was relieved and that of the middle
significant [30]. finger was maintained without relieving the endpoint pressure. This
was maintained for 30 s before the third point was identified. Similar
Randomization of participants: Eligible participants who provided patterns for all identified sensitive points were followed continuously
informed consent were randomized into one of three treatment along the neuromuscular continuum until the last point, approxi-
groups: SMWLM, PINS, or combined SMWLM + PINS. A randomization mately 2 cm proximal to the endpoint, was found. The pressure was
timeline was prepared by a research assistant who had no commu- maintained for 30 s concurrently on the two points (the last and the
nication with the participants throughout the trial and was unaware of endpoint) and then relieved (Figure 3).
the study protocol. The randomization series (using block randomi-
zation) was created using SAS 9.4 statistical software with the par- Compliance with home program: Therapeutic exercises (lumbar sta-
ticipants likely to be assigned to a group with an equal chance of bilization and stretching exercises) were initiated at low speed
allocation (Figure 1). (beginner stage) during the treatment sessions to ensure that each
participant received appropriate exercise instructions and progression
Intervention procedure: There were three intervention groups in this modes before discharge. After treatment ended, participants were
study, with the first group receiving SMWLM, the second group given a handout detailing appropriate exercise dosages and pro-
receiving PINS, and the third group receiving combined gressions. We provided the same exercise guidelines reported in our
SMWLM + PINS treatment techniques. Treatments sessions for both previous study [18]. Additionally, exercise compliance was assessed
the single and combined treatment groups were carried out on the and reinforced every four weeks via phone calls and every 3 months at
same day. However, participants in the single treatment groups spent in-person visits. Further, participants were also asked to refrain from
additional time receiving lectures on lumbar spine anatomy, to all other rehabilitation techniques or exercise not prescribed by the
compensate for not receiving dual therapies. Each group (both single trial physiotherapists; this was also monitored periodically via phone
and combined) attended two treatments per week of for 30 min each, calls and physical visits.
for three months. Additionally, each group also received neural tissue
mobilization [31] as a baseline treatment. Statistical analysis: All data were analyzed using SPSS 23.0 (SPSS,
Inc.). Descriptive statistics were used to summarize the demographics
SMWLM technique: Participants in the SMWLM group received and clinical parameters of the participants. Shapiro–Wilk test was
SMWLM as per Mulligan’s guidelines [32]. The participant was laid on used to assess the normality of the data, while Levene’s test was used
the side, facing the treating physiotherapist (PT; M.S.D.), with the leg to assess the homogeneity of variances among groups. One-way
to be treated on top. An assistant PT supported the uppermost leg. The ANOVA was used to compare outcomes at baseline. The primary
treating PT then bent over the participant and put one thumb, sup- aim of the study (effects of interventions on pain, disability, and
ported by the other, on the spinous process of the affected vertebra as sciatica) was examined by the use of two-way repeated-measures
determined with reference to the posterior superior iliac crest. The ANOVA with the intervention group (SMWLM, PINS, and combined
treating PT then pressed down on the palpated spinous process. This SMWLM + PINS) as the between-subject variable and time (baseline,
pressure was maintained while the participant actively performed the posttreatment, and three, six, and nine months follow-up) as the
straight leg raise (SLR) for the leg supported by the assistant PT, in as within-subject variable. Separate ANOVAs were performed with the
much as this did not cause pain. This position was retained for 30 s, outcomes (back pain, leg pain, disability, and sciatica) as the
after which the treating PT released the pressure on the spinous pro- dependent variables and for each ANOVA, the analysis of interest was
cess and the participant lowered the supported leg to the treatment the two-way interaction (Group × Time). The data of the participants
table. Three repetitions were applied during the first visit. However, as lost to follow-up were treated by intention-to-treat analysis (ITT) by
the participant improved in subsequent visits, the assistant PT applied removing the lost data in the analysis (on protocol/per protocol/
overpressure on the supported leg as the participant performed the completer analysis procedure) [33]. Differences between the means
SLR. This was also maintained for 30 s, after which the leg was lowered were considered at 5% probability level (p<0.05) and the value of
to the treatment table (Figure 2). This procedure was repeated six times confidence interval (CI) was set at 95%.
on subsequent visits.

PINS technique: Participants in the PINS group received the PINS Results
program as per Dowling’s guidelines [20]. Two connected points,
named primary and endpoints, were palpated using the index fingers A total of 60 participants (age range, 36–59 years; mean
of both hands. The points were areas of most and least sensitivity, age, 47.8 ± 4.32 years) were enrolled in the study, with 20
respectively, found along a neuromuscular structure. When an area of

394 Danazumi et al.: Randomized trial of manual therapy for lumbar radiculopathy

Figure 1: Chart demonstrating the flow of
patients through this randomized clinical
study.

Figure 2: Spinal mobilization with leg movement. The patient is participants each in the SMWLM, PINS, and combined
lying on the side with the affected lower limb uppermost, and the SMWLM + PINS treatment groups. Thirty-seven (61.7%)
treating physiotherapist (PT) is facing and stooping over the patient. participants were men and 23 (38.3%) participants were
The treating PT pushes down the chosen spinous process using women. One participant in the SMWLM + PINS treatment
reinforced thumbs and the assistant PT supports the uppermost leg group was lost at the nine-month follow-up period and his
of the patient when straight leg raising is performed. data was adjusted through intention-to-treat (ITT) analysis
by removing the nine-month follow-up data from the
analysis. No significant differences were observed in the
baseline characteristics of the study participants among
groups (Table 1).

Between-groups analyses using a two-way repeated-
measures ANOVA indicated significant interactions be-
tween group and time (Group × Time) for all outcomes:
RMDQ [F(4, 116)=15.994; p=0.001)], VAS in the leg [F(4, 116)
=14.10,7; p=0.001)], VAS in the back [F(4, 116)=7.469,
p=0.001)], SBI [F(4, 116)=8.073; p=0.001)] (Table 2).

Danazumi et al.: Randomized trial of manual therapy for lumbar radiculopathy 395

Figure 3: Progressive inhibition of
neuromuscular structures. The patient is
lying prone and the treating
physiotherapist (PT) is standing at the same
side with the patient’s ipsilateral lower
limb. The treating PT’s proximal index finger
is placed at the primary end point and the
distal index finger is placed at the
secondary end point. The proximal finger of
the PT applies ischemic compression at
primary end points, moving progressively
towards the distal index finger to
approximately the 2 cm mark, and then both
pressures are relieved at once.

Post-hoc analyses (of Group × Time effects) indicated that radiculopathy secondary to disc herniation. The findings of
participants who received combined SMWLM + PINS)= this study showed that there were significant within-group
treatment experienced greater improvement in leg pain differences for all outcomes at all timelines (immediately
(p<0.05), back pain (p<0.05), disability (p<0.05) and posttreatment and three, six, and nine months follow-up
sciatica bothersomeness (p<0.05), immediately posttreat- periods), indicating that all measures significantly
ment (Table 3), and at three-month (Table 4), six-month improved over time from baseline to 12 months. This
(Table 5), and nine-month (Table 6) follow-up periods than finding is congruent with those of a previous study [18], in
the participants receiving SMWLM or PINS alone. However, which a significant improvement in all outcomes at all
participants receiving SMWLM showed better improve- timelines was reported. Similar findings were also reported
ment than the participants receiving PINS at all timelines in other previous studies [22, 31, 34, 35] that examined the
(Tables 3–6). effect of SMWLM compared with other interventions.
However, in contrast with a previous study [18] and other
Discussion studies [22, 34, 35], ours examined the combined effects of

This study investigated the combined effects of SMWLM Table : Two-way repeated measures ANOVA for between group and
and PINS compared with the individual techniques alone time interactions.
in the management of individuals with lumbar

Variable F(df) p-value

Table : Baseline demographics and clinical characteristics. RMDQ . (, ) <.
Group . (, ) <.
Variable SMWLM PINS Combined p-value Time . (, ) <.
Mean (SD) Mean (SD) Mean (SD) Group × Time
. (, ) <.
(n=) (n=) (n=) VASL . (, ) <.
Group <.
Age, years . (.) . (.) . (.) . Time . (, )
BMI, kg/m . (.) . (.) . (.) . Group × Time <.
DOS, months . (.) . (.) . . (, ) <.
RMDQ . (.) . (.) . (.) . VASB . (, ) <.
VASL . (.) . (.) . (.) . Group
VASB . (.) . (.) . Time . (, ) <.
SBI . (.) . (.) . (.) . Group × Time <.
. (.) . (, ) <.
. (.) SBI . (, )
Group
BMI, body mass index; DOS, duration of symptoms; PINS, progressive Time . (, )
inhibition of neuromuscular structures; RMDQ, Rolland–Morris Group × Time
disability questionnaire; SBI, sciatica bothersome index; SD,
standard deviation; SMWLM, spinal mobilization with leg movement; RMDQ, Rolland–Morris Disability Questionnaire; SBI, sciatica
VASL, Visual Analog Scale for leg pain; VASB, Visual Analog Scale for bothersome index; VASL, Visual Analog Scale for leg pain; VASB,
back pain. Visual Analog Scale for back pain.

396 Danazumi et al.: Randomized trial of manual therapy for lumbar radiculopathy

Table : Baseline to posttreatment change scores for clinical outcomes.

Variable Group (n) Baseline Posttreatment Within-group Between-group change score p-value
Mean (SD) Mean (SD) change score Mean (% CI)

Mean (SD)

RMDQ SMWLM () . (.) . (.) . (.) . (.–.) .
VASL PINS () . (.) . (.) . (.)
VASB COMBINED () . (.) . (.) . (.–.) .
SBI SMWLM () . (.) . (.)
PINS () . (.) . (.) . (.) . (.–.) .
COMBINED () . (.) . (.) . (.)
SMWLM () . (.) . (.) . (.) . (.–.) .
PINS () . (.) . (.) . (.)
COMBINED () . (.) . (.) . (.)
SMWLM () . (.) . (.) . (.)
PINS () . (.) . (.) . (.)
COMBINED () . (.) . (.) . (.)
. (.) . (.)

CI, confidence interval; RMDQ, Rolland–Morris Disability Questionnaire; SBI, Sciatica Bothersome Index; SD, standard deviation; VASL, Visual
Analog Scale for leg pain; VASB, Visual Analog Scale for back pain.

Table : Posttreatment to three-month follow-up change scores for clinical outcomes.

Variable Group (n) Immediate Three-month Within-group Between-group change p-value
posttreatment follow-up change score score mean (% CI)
Mean (SD)
Mean (SD) Mean (SD)

RMDQ SMWLM () . (,) . (.) . (.) . (.–.) .
VAS PINS () . (.) . (.) . (.) . (.–.) .
VASB COMBINED () . (.) . (.) . (.–.) .
SBI SMWLM () . (.) . (.) . (.) . (.–.) .
PINS () . (.) . (.) . (.)
COMBINED () . (.) . (.) . (.)
SMWLM () . (.) . (.) . (.)
PINS () . (.) . (.) . (.)
COMBINED () . (.) . (.) . (.)
SMWLM () . (.) . (.) . (.)
PINS () . (.) . (.) . (.)
COMBINED () . (.) . (.) . (.)
. (.)

RMDQ, Rolland–Morris Disability Questionnaire; SBI, Sciatica Bothersome Index; VASL, Visual Analog Scale for leg pain; VASB, Visual Analog
Scale for back pain.

Table : Three-month to six-month follow-up change scores for clinical outcomes.

Variable Group (n) Three-month Six-month Within-group Between-group change p-value
follow-up follow-up change score score mean (% CI)
Mean (SD) Mean (SD)
Mean (SD)

RMDQ SMWLM () . (.) . (.) . (.) . (.–.) .
VASL PINS () . (.) . (.) . (.) . (.–.) .
VASB COMBINED () . (.) . (.) . (.) . (.–.) .
SBI SMWLM () . (.) . (.) . (.) . (.–.) .
PINS () . (.) . (.) . (.)
COMBINED () . (.) . (.) . (.)
SMWLM () . (.) . (.) . (.)
PINS () . (.) . (.) . (.)
COMBINED () . (.) . (.) . (.)
SMWLM () . (.) . (.) . (.)
PNS () . (.) . (.) . (.)
COMBINED () . (.) . (.) . (.)

RMDQ, Rolland–Morris Disability Questionnaire; SBI, Sciatica Bothersome Index; VASL, Visual Analog Scale for leg pain; VASB, Visual Analog
Scale for back pain.

Danazumi et al.: Randomized trial of manual therapy for lumbar radiculopathy 397

Table : Six-month to nine-month follow-up change scores for clinical outcomes.

Variable Group (n) Six-month Nine-month Within-group Between-group change p-value
follow-up follow-up change score score mean (% CI)
Mean (SD) Mean (SD)
Mean (SD)

RMDQ SMWLM () . (.) . (.) . (.) . (.–.) .
VASL PINS () . (.) . (.) . (.) . (.–.) .
VASB COMBINED () . (.) . (.) . (.) . (.–.) .
SBI SMWLM () . (.) . (.) . (.) . (.–.) .
PINS () . (.) . (.) . (.)
COMBINED () . (.) . (.) . (.)
SMWLM () . (.) . (.) . (.)
PINS () . (.) . (.) . (.)
COMBINED () . (.) . (.) . (.)
SMWLM () . (.) . (.) . (.)
PNS () . (.) . (.) . (.)
COMBINED () . (.) . (.) . (.)

RMDQ, Rolland–Morris Disability Questionnaire; SBI, Sciatica Bothersome Index; VASL, Visual Analog Scale for leg pain; VASB, Visual Analog
Scale for back pain.

the two treatment techniques (SMWLM and PINS) and data relief and improved motion in individuals with low back
were reported for long-term follow-ups. pain [37–39] while also improving disc hydration [40–42].
Given that SMWLM is a form of P–A gliding mobilization, it
Our findings also showed that there were significant is conceivable that this technique could correct the posi-
between-group differences in all outcomes with the com- tional fault by relieving pressure from the structures that
bined treatment group (SMWLM + PINS) having greater compress the nerves, improving disc hydration, and also
improvement in all outcomes (pain, disability, and sciatica reducing the extent of pain via centralization [19].
bothersomeness) at all timelines than the individual
treatment groups (SMWLM or PINS). Additionally, the On the other hand, PINS has been indicated to reduce
SMWLM group showed better improvement than the PINS lower limb radiating pain through increased hyperemia
group. These findings are similar to those of Das et al. [22], and washing away of the metabolites in the neuromuscular
Thakur et al. [31], and Yadav et al. [35], which reported structures following myofascial trigger point deactivation
better improvement in the SMWLM group than the com- [20]. Additionally, the ischemic compression applied on
parison groups. In contrast to our current findings, a pre- the sensitive neuromuscular structures using PINS may
vious study by Ahmed et al. [34] reported better also elicit counterirritant effect in which large, fast-
improvement in the comparison group than the SMWLM conducting afferent fibers in the dorsal horn of the spinal
group. Another previous study [18] reported no between- cord and collateral fibers in the substantia gelatinosa or
group differences in all outcomes which further contrasted adjacent interneurons inhibit the transmission of pain via
this current finding. the spinothalamic tract [20]. In this manner, pressure acts
as a stimulant to neighboring tissues, reducing the sensi-
The significant improvement achieved in our study tivity of the original tender point and thus resolving pain
after administering a combined treatment approach may [43]. Similar findings were also reported by a previous
not be unrelated to the ability of the two treatment tech- study [44] which indicated that sciatic nerve manipulation
niques in ameliorating nerve root compression and may promote healing of the soft tissues by stimulating the
sciatica. First, SMWLM has been indicated to relieve nerve functions of the nervous system to improve nervous system
compression through increased intervertebral disc space adaptability and decrease sensitivity, helping to alleviate
gapping, nucleus deformation, and simultaneous approx- the symptoms. All these effects of SMWLM and PINS may
imation in the alternate layers of the annulus, thereby be responsible for the better improvement achieved in the
producing favorable therapeutic effects on the interverte- combined treatment group.
bral disc [19]. While decreased diffusion of water and loss
of proteoglycans are hallmarks of disc degeneration, it has Moreover, in analyzing our data, it was observed that
been suggested that fluid exchange is integral to main- the participants’ body mass indices (BMIs) ranged from
taining disc nutrition [36]. Mobilizations such as posterior– the normal weight to overweight, with the combined
anterior (P–A) glides have been reported to result in pain

398 Danazumi et al.: Randomized trial of manual therapy for lumbar radiculopathy

treatment group having the highest mean BMI. Despite points (TrPs) along a neuromuscular structure and then
having a higher mean BMI, the combined treatment group applying ischemic compression to progressively deactivate
showed better improvement (in all outcomes) than the those TrPs. PINS technique is very similar to other indirect
individual manual therapy groups, indicating that this OMTs such as strain-counter strain (SCS) or positional
improvement may be attributed to the combined therapies release therapy (PRT), only TrPs are progressively deacti-
(SMWLM + PINS) administered in this group. This finding vated along a neuromuscular continuum in the PINS tech-
corroborates with those of Vismara et al. [45], who re- nique. It is therefore recommended that osteopaths and
ported greater improvement in biomechanical parameters other highly informed healthcare professionals may
of the thoracic spine in obese patients with chronic low consider the use of PINS either as a sole or an complementary
back pain managed with osteopathic manipulative treat- treatment in the management of patients with lumbar radi-
ment (OMT) than those treated with specific exercises. culopathy secondary to disc herniation.
Similar findings were also reported by Licciardone et al.
[46], who indicated that the OMT effects for LBP intensity Conclusions
and back-specific functioning were independent of
baseline patient demographic characteristics, comorbid Our results showed that combined SMWLM and PINS
medical conditions, and medication use. therapy along with a structured home exercise program
was better than the individual techniques alone in the
Although therapeutic exercises were incorporated along management of individuals with lumbar radiculopathy
with the manual therapy techniques in our study, we found secondary to disc herniation.
that these exercises cannot be used as sole treatments for in-
dividuals suffering from lumbar radiculopathy secondary to Research funding: None declared.
disc herniation [47–49]. However, given the significance of a Author contributions: Drs Bello, Yakasai, and Kaka and Mr
well-trained core in the for back stability, it is conceivable that Danazumi provided substantial contributions to
these exercises may be used as adjunct therapies [50, 51]. conception and design, acquisition of data, or analysis
Additionally, it is also worthy of note that the long-term and interpretation of data; Mr Danazumi drafted the article
improvement achieved in our study may be attributed to the and revised it critically for important intellectual content;
combined effects of both manual and exercise therapies. all authors gave final approval of the version of the article
to be published; and all authors agree to be accountable for
Limitations all aspects of the work in ensuring that questions related to
the accuracy or integrity of any part of the work are
This study was limited to individuals with unilateral lumbar appropriately investigated and resolved.
radiculopathy and data was collected in only one clinical Competing interests: Authors state no conflict of interest.
setting. Additionally, this study did not also assess psy- Informed consent: Written informed consent was obtained
chosocial variables that may be associated with chronic and from the participants and they were assured of their rights
persistent conditions like lumbar radiculopathy secondary to withdraw from the study at any point in time and their
to disc herniation. Further, it might have also been useful to full anonymity was maintained.
use a 4th group as a control, which could have allowed us to Ethical approval: Ethical approval to conduct this study
monitor the natural evolution of the disease and pain. was obtained from the Health Research Ethics Committee
Future studies should be conducted to address these issues. of the Federal Medical Centre (FMC), Nguru, Yobe State,
Nigeria (Registration number: FMC/N/CL.SERV/355/
Clinical relevance to osteopathic medicine VOL.IV/139). The trial was registered with the Pan
African Clinical Trial Registry (PACTR201907782710996)
Osteopathic manipulative treatment (OMT) is a set of hands- on July 11, 2019.
on techniques used by osteopathic physicians (DOs) to di-
agnose, treat, and prevent illness or injury. OMT is performed References
by a DO; while manipulative therapy in our study was not
performed by a DO, similarities may be appreciated. Our 1. Linqiu Z, Schneck CD, Shao Z. The anatomy of dorsal ramus nerves
study demonstrated the therapeutic efficacy of using PINS as and its implications in lower back pain. Neurosci Med 2012;3:
an indirect OMT in the management of LDHR. The applica- 192–201.
tion of PINS involves the palpation of myofascial trigger

Danazumi et al.: Randomized trial of manual therapy for lumbar radiculopathy 399

2. Casey E. Natural history of radiculopathy. Phys Med Rehabil Clin 20. Dowling DJ. Progressive inhibition of neuromuscular structures
2011;22:1–5. (PINS) technique. J Am Osteopath Assoc 2000;100:285–6.

3. Valat JP, Genevay S, Marty M, Rozenberg S, Koes B. Sciatica. Best 21. Faul F, Erdfelder E, Buchner A, Lang AG. Statistical power
Pract Res Clin Rheumatol 2010;24:241–52. analyses using G*Power 3.1: tests for correlation and regression
analyses. Behav Res Methods 2009;41:1149–60.
4. Schoenfeld AJ, Weiner BK. Treatment of lumbar disk herniation:
evidence based practice. Int J Med 2010;3:209–14. 22. Das MSS, Dowle P, Iyengar R. Effect of spinal mobilization with
leg movement as an adjunct to neural mobilization and
5. Ostelo RW. Physiotherapy management of sciatica. J Physiother conventional therapy in patients with lumbar radiculopathy:
2020;66:83–8. randomized controlled trial. J Med Sci Clin Res 2018;6:
356–64.
6. Skelly AC, Chou R, Dettori JR, Turner JA, Friedly JL, Rundell SD, et al.
Noninvasive nonpharmacological treatment for chronic pain: a 23. Suresh KP, Chandrashekara S. Sample size estimation and power
systematic review update. Rockville (MD): Agency for Healthcare analysis for clinical research studies. J Hum Reprod Sci 2012;5:
Research and Quality (US); 2020. Report No.: 20-EHC009. PMID: 7–13.
32338846.
24. Rabin A, Gerszten PC, Karausky P, Bunker CH, Potter DM, Welch
7. Maxwell CM, Lauchlan DT, Dall PM. The effects of spinal WC. The sensitivity of the seated straight-leg raise test compared
manipulative therapy on lower limb neurodynamic test with the supine straight-leg raise test in patients presenting with
outcomes in adults: a systematic review. J Man Manip Ther magnetic resonance imaging evidence of lumbar nerve root
2020;28:4–14. compression. Arch Phys Med Rehabil 2007;88:840–3.

8. Delitto A, George SZ, Van Dillen L, Whitman JM, Sowa G, 25. Majlesi J, Togay H, Unalan H, Toprak S. The sensitivity and
Shekelle P, et al. Low back pain. J Orthop Sports Phys Ther 2012; specificity of the Slump and the Straight Leg Raising tests in
42:A1–57. patients with lumbar disc herniation. J Clin Rheumatol 2008;14:
87–91.
9. Danish Health Authority. National Clinical Guideline: interventions
for recent onset lumbar radiculopathy [in Danish]; 2016. https:// 26. Walsh J, Hall T. Agreement and correlation between the straight
www.sst.dk/da/udgivelser/2016/lumbal-nerverodspaavirkning- leg raise and slump tests in subjects with leg pain. J Manip
ikke-kirurgisk-behandling [Accessed 30 Oct 2020]. Physiol Ther 2009;32:184–92.

10. Stochkendahl MJ, Kjaer P, Hartvigsen J, Kongsted A, Aaboe J, 27. Kamath SU, Kamath SS. Lase`gue’s sign. J Clin Diagn Res 2017;11:
Andersen M, et al. National Clinical Guidelines for non-surgical RG01–2.
treatment of patients with recent onset low back pain or lumbar
radiculopathy. Eur Spine J 2018;27:60–75. 28. Shi J, Jia L, Yuan W, Shi G, Ma B, Wang B, et al. Clinical
classification of cauda equina syndrome for proper treatment.
11. Hidalgo B, Detrembleur C, Hall T, Mahaudens P, Nielens H. The Acta Orthop 2010;81:391–5.
efficacy of manual therapy and exercise for different stages of
non-specific low back pain: an update of systematic reviews. J 29. Yao M, Xu BP, Li ZJ, Zhu S, Tian ZR, Li DH, et al. A comparison
Man Manip Ther 2013;22:59–74. between the low back pain scales for patients with lumbar disc
herniation: validity, reliability, and responsiveness. Health Qual
12. Slater SL, Ford JJ, Richards MC, Taylor NF, Surkitt LD, Hahne AJ. Life Outcome 2020;18:175.
The effectiveness of sub-group specific manual therapy for low
back pain: a systematic review. Man Ther 2012;17:201–12. 30. Haugen AJ, Grøvle L, Brox JI, Natvig B, Keller A, Soldal D, et al.
Estimates of success in patients with sciatica due to lumbar disc
13. Clar C, Tsertsvadze A, Court R, Hundt GL, Clarke A, Sutcliffe P. herniation depend upon outcome measure. Eur Spine J 2011;20:
Clinical effectiveness of manual therapy for the management of 1669–75.
musculoskeletal and non-musculoskeletal conditions:
systematic review and update of UK evidence report. Chiropr Man 31. Thakur A, Mahapatra RK, Mahapatra R. Effect of Mulligan spinal
Ther 2014;22:12. mobilization with leg movement and shacklock neural tissue
mobilization in lumbar radiculopathy: a randomised controlled
14. Swanson BT, Creighton D. The degenerative lumbar disc: not a trial. J Med Thesis 2015;3:27–30.
disease, but still an important consideration for OMPT practice: a
review of the history and science of discogenic instability. J Man 32. Mulligan BR. Manual therapy: “NAGS,” “SNAGS,”, “MWMS” etc,
Manip Ther 2020;28:191–200. 5th ed. Wellington, New Zealand: Plane View Services; 2004.

15. Berry JA, Elia C, Saini HS, Miulli DE. A review of lumbar 33. Streiner D, Geddes J. Intention to treat analysis in clinical trials
radiculopathy, diagnosis, and treatment. Cureus 2019;11:e5934. when there are missing data. Evid Base Ment Health 2001;4:70–1.

16. Dan-Azumi MS, Bello B, Rufai SA, Abdurrahman MA. Surgery 34. Ahmed N, Khan Z, Chawla C. Comparison of Mulligan’s spinal
versus conservative management for lumbar disc herniation with mobilisation with limb movement (SMWLM) and neural tissue
radiculopathy: a systematic review and meta-analysis. J Health mobilization for the treatment of lumbar disc herniation: a
Sci 2018;8:42–53. randomized clinical trial. J Nov Physiother 2016;6:304.

17. Chen B, Guo J, Zhang H, Zhang YJ, Zhu Y, Zhang J, et al. Surgical 35. Yadav S, Arora Nijhawan M, Panda P. Effectiveness of spinal
versus non-operative treatment for lumbar disc herniation: a mobilization with leg movement (SMWLM) in patients with
systematic review and meta-analysis. Clin Rehabil 2018;32: lumbar radiculopathy (L5/S1 nerve root) in lumbar disc
146–60. herniation. Int J Physiother Res 2014;2:712–8.

18. Bello B, Danazumi MS, Kaka B. Comparative effectiveness of 2 36. Johannessen W, Vresilovic EJ, Wright AC, Elliott DM. Intervertebral
manual therapy techniques in the management of lumbar disc mechanics are restored following cyclic loading and
radiculopathy: a randomized clinical trial. J Chiropr Med 2019;18: unloaded recovery. Ann Biomed Eng 2004;32:70–6.
253–60.
37. Bezci SE, Nandy A, O’Connell GD. Effect of hydration on healthy
19. Mulligan BR. Update on spinal mobilisations with leg movement. intervertebral disk mechanical stiffness. J Biomech Eng 2015;
J Man Manip Ther 1997;5:184–7. 137:10.

400 Danazumi et al.: Randomized trial of manual therapy for lumbar radiculopathy

38. Goodsell M. Short-term effects of lumbar posteroanterior 45. Vismara L, Cimolin V, Menegoni F, Zaina F, Galli M, Negrini S,
mobilization in individuals with low-back pain. J Manip Physiol et al. Osteopathic manipulative treatment in obese patients
Ther 2000;23:332–42. with chronic low back pain: a pilot study. Man Ther 2012;17:
451–5.
39. Powers CM, Beneck GJ, Kulig K, Landel RF, Fredericson M.
Effects of a single session of posterior-to-anterior spinal 46. Licciardone JC, Gatchel RJ, Aryal S. Targeting patient subgroups
mobilization and press-up exercise on pain response and with chronic low back pain for osteopathic manipulative
lumbar spine extension in people with nonspecific low back treatment: responder analyses from a randomized controlled
pain. Phys Ther 2008;88:485–93. trial. J Am Osteopath Assoc 2016;116:156–68.

40. Beattie PF, Arnot CF, Donley JW, Noda H, Bailey L. The immediate 47. Kreiner DS, Hwang SW, Easa JE, Resnick DK, Baisden JL, Bess S,
reduction in low back pain intensity following lumbar joint et al. North American Spine Society. An evidence-based clinical
mobilization and prone press-ups is associated with increased guideline for the diagnosis and treatment of lumbar disc
diffusion of water in the L5-S1 intervertebral disc. J Orthop Sports herniation with radiculopathy. Spine J 2014;14:180–91.
Phys Ther 2010;40:256–64.
48. Danazumi MS. Physiotherapy management of lumbar disc
41. Beattie PF, Donley JW, Arnot CF, Miller R. The change in the herniation with radiculopathy: a narrative review. Niger J Exp Clin
diffusion of water in normal and degenerative lumbar Biosci 2019;7:93–100.
intervertebral discs following joint mobilization compared to
prone lying. J Orthop Sports Phys Ther 2009;39:4–11. 49. Alzubeidi SA, Alfawaz AA, Algharawi MY, Alrashidi AS,
Alamrani MA, Alsayed AB. The effectiveness of stabilization
42. Beattie PF, Morgan PS, Peters D. Diffusion-weighted magnetic exercises in treating patients with chronic low back pain: a
resonance imaging of normal and degenerative lumbar systematic review. Asian J Orthop Res 2020;3:30–46.
intervertebral discs: a new method to potentially quantify the
physiologic effect of physical therapy intervention. J Orthop 50. Kriese M, Clijsen R, Taeymans J, Cabri J. Segmental stabilization
Sports Phys Ther 2008;38:42–9. in low back pain: a systematic review. Sportverletz Sportschaden
2010;24:17–25.
43. Wong CK. Strain counterstrain: current concepts and clinical
evidence. Man Ther 2012;17:2–8. 51. Ford JJ, Bower SE, Ford I, de Mello MM, Carneiro SR,
Balasundaram AP, et al. Effects of specific muscle activation
44. Jeong UC, Kim CY, Park YH, Hwang-Bo G, Nam CW. The effects of for low back pain on activity limitation, pain, work
self-mobilization techniques for the sciatic nerves on physical participation, or recurrence: a systematic review. Musculoskel
functions and health of low back pain patients with lower limb Sci Pract 2020. https://doi.org/10.1016/j.msksp.2020.
radiating pain. J Phys Ther Sci 2016;28:46–50. 102276.

J Osteopath Med 2021; 121(4): 401–415

Neuromusculoskeletal Medicine (OMT) Brief Report

Adrienne M. Kania, DO, Kailee N. Weiler, OMS III, Angeline P. Kurian, OMS III,
Marielle L. Opena, DO, Jennifer N. Orellana, OMS III and Harald M. Stauss*, MD, PhD

Activation of the cholinergic antiinflammatory
reflex by occipitoatlantal decompression and
transcutaneous auricular vagus nerve stimulation

https://doi.org/10.1515/jom-2020-0071 for parasympathetic modulation of cardiac function. In all
Received March 26, 2020; accepted November 9, 2020; three groups, the experimental protocol was associated with
published online February 24, 2021 a significant increase in salivary cytokine concentrations.
However, the increase in IL-1β was significantly less in the
Abstract taVNS group (+66 ± 13 pg/mL; p<0.05) than in the time
control group (+142 ± 24 pg/mL). A similar trend was
Context: The parasympathetic-mediated inflammatory re- observed in the taVNS group for TNF-α (+1.7 ± 0.3 pg/mL vs.
flex inhibits excessive proinflammatory cytokine produc- 4.1 ± 1.3 pg/mL; p<0.10). In the OA-D group baseline IL-6,
tion. Noninvasive techniques, including occipitoatlantal IL-8, and TNF-α levels on the third study day were signifi-
decompression (OA-D) and transcutaneous auricular vagus cantly lower than on the first study day (IL-6: 2.3 ± 0.4 vs.
nerve stimulation (taVNS), have been demonstrated to in- 3.2 ± 0.6 pg/mL, p<0.05; IL-8: 190 ± 61 vs. 483 ± 125 pg/mL, p
crease parasympathetic tone. <0.05; TNF-α: 1.2 ± 0.3 vs. 2.3 ± 0.4 pg/mL, p<0.05). OA-D
Objectives: To test the hypothesis that OA-D and taVNS decreased mean blood pressure from the first (100 ± 8 mmHg)
increase parasympathetic nervous system activity and to the second (92 ± 6 mmHg; p<0.05) and third (93 ± 8 mmHg;
inhibit proinflammatory cytokine mobilization and/or p<0.05) study days and reduced low frequency spectral po-
production. wer of systolic blood pressure variability (19 ± 3 mmHg2 after
Methods: Healthy adult participants were randomized to OA-D vs. 28 ± 5 mmHg2 before OA-D; p<0.05), a marker of
receive OA-D (5 min of OA-D followed by 10 min of rest; sympathetic modulation of vascular tone. OA-D also
n=8), taVNS (15 min; n=9), or no intervention (15 min, time increased baroreceptor-heart rate reflex sensitivity from the
control; n=10) on three consecutive days. Before and after first (13.7 ± 3.0 ms/mmHg) to the second (18.4 ± 4.3 ms/
these interventions, saliva samples were collected for mmHg; p<0.05) and third (16.9 ± 4.2 ms/mmHg; p<0.05)
determination of the cytokines interleukin-1β (IL-1β), study days.
interleukin-6 (IL-6), interleukin-8 (IL-8), and tumor ne- Conclusions: Both OA-D and taVNS elicited antiin-
crosis factor α (TNF-α). Arterial blood pressure and the flammatory responses that were associated with increases
electrocardiogram were recorded for a 30-min baseline, in heart rate variability-derived markers for para-
throughout the intervention, and during a 30-min recovery sympathetic function. These findings suggest that OA-D
period to derive heart rate and blood pressure variability and taVNS activate the parasympathetic antiinflammatory
markers as indices of vagal and sympathetic control. reflex. Furthermore, an antihypertensive effect was
Results: OA-D and taVNS increased root mean square of observed with OA-D that may be mediated by reduced
successive RR interval differences (RMSSD) and high fre- sympathetic modulation of vascular tone and/or increased
quency heart rate variability, which are established markers baroreceptor reflex sensitivity.

*Corresponding author: Harald M. Stauss, MD, PhD, Department of Keywords: blood pressure variability; cytokines; heart rate
Biomedical Sciences, Burrell College of Osteopathic Medicine, 3501 variability; hypertension; saliva.
Arrowhead Drive, Las Cruces, NM 88001-6056, USA,
E-mail: [email protected] Biological antiinflammatory drugs, such as tumor necrosis
Adrienne M. Kania, DO, Kailee N. Weiler, OMS III, Angeline P. Kurian, factor-α (TNF-α) antagonists, have revolutionized treat-
OMS III, Marielle L. Opena, DO and Jennifer N. Orellana, OMS III, ment of chronic inflammatory diseases and largely
Department of Clinical Medicine, Burrell College of Osteopathic improved prognosis and quality of life for affected patients
Medicine, Las Cruces, NM, USA [1, 2]. However, the high cost of these agents [3] and

Open Access. © 2020 Adrienne M. Kania et al., published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0
International License.

402 Kania et al.: Antiinflammatory reflex activation by OA-D and taVNS

potential adverse effects [4] and associated contraindica- craniosacral OMT compared to control subjects. Likewise, a
tions are prohibitive for a large number of patients. Thus, study by Henley et al. [14] found less tachycardia and higher
there is a need for affordable and effective alternative or high frequency HRV during an orthostatic challenge in
supplemental treatment strategies. Activation of the subjects receiving OMT as compared to subjects receiving
cholinergic antiinflammatory pathway [5] within the in- sham or no treatment. In that study, cervical myofascial
flammatory reflex [6] through osteopathic manipulative release techniques were used [14]. Giles et al. [15] demon-
treatment (OMT) or through direct vagus nerve stimulation strated that cervical OMT significantly increased overall
may offer an adjunctive therapeutic approach that could HRV (standard deviation of normal-to-normal intervals) and
allow for lower dosing of biological antiinflammatory high frequency HRV and, consequently, concluded that
drugs, therefore reducing the cost and potential adverse those OMT techniques modulate cardiac parasympathetic
effects of such treatments. The inflammatory reflex is an nervous system function. The OMT technique used in that
endogenous mechanism that prevents excessive responses study involved cervical soft tissue kneading and stretching
to acute and chronic proinflammatory stimuli such as local followed by suboccipital decompression [15]. Finally, Curi
tissue damage, invading pathogens, and rheumatoid or et al. [16] demonstrated that fourth ventricular compression
autoimmune disorders [5, 6]. This reflex is activated by reduced blood pressure in hypertensive patients and that
local inflammatory mediators such as cytokines or this antihypertensive effect was associated with an increase
pathogen-derived products that are sensed by afferent in high frequency HRV and a shift in autonomic balance to
vagal nerve fibers that project to the central nervous sys- parasympathetic dominance as indicated by a reduction in
tem. This afferent pathway activates the efferent reflex arc the low frequency to high frequency ratio of HRV. Taken
that utilizes efferent vagal nerve fibers to inhibit cytokine together, several OMT techniques have been demonstrated
production in reticular organs, such as the spleen via to increase parasympathetic activity. However, it is not
pathways that may depend on the sympathetic splenic known whether OMT-induced elevation of parasympathetic
nerve [7] and nicotinic α7 subtype acetylcholine receptors nervous system activity also results in activation of the
[8]. However, the exact neuronal pathways mediating the antiinflammatory efferent arc of the inflammatory reflex. If
efferent reflex response are still a matter of debate [9]. that were the case, OMT techniques could potentially be
Central nervous system processing of the afferent proin- developed to treat inflammatory conditions.
flammatory signals involves activation of postsynaptic
M1-muscarinic receptors because it has been demonstrated An alternative noninvasive approach to potentially
that intracerebroventricular administration of the M1- activate the cholinergic antiinflammatory pathway is
-muscarinic receptor agonist McN-A-343 decreases serum transcutaneous auricular vagus nerve stimulation (taVNS)
tumor necrosis factor (TNF) levels during endotoxemia [17, 18]. Evidence for an activation of the parasympathetic
[10]. This M1-muscarinic receptor-mediated pathway can nervous system by taVNS comes from studies that utilized
be augmented by inhibition of presynaptic M2-muscarinic HRV analysis, microneurography, and baroreceptor reflex
receptors that normally inhibit neuronal acetylcholine analysis [19–22]. Thus, as with OMT, there is a possibility
release [10]. Interestingly, intracerebroventricular admin- that taVNS may potentially be useful in the treatment of
istration of the M2-muscarinic antagonist methoctramine inflammatory conditions. Indeed, taVNS has been sug-
not only inhibited the TNF response to endotoxemia but gested to elicit antiinflammatory responses in depression
also increased high frequency heart rate variability (HRV) [23], postoperative ileus and endotoxemia [24], and in Par-
[10], a measure of parasympathetic modulation of cardiac kinson’s disease [25]. In a recent metaanalysis of OMT
function [11]. Thus, HRV analysis can be utilized as an in- techniques in inflammatory diseases [26], only two of 10
direct measure of the activation of the efferent anti- included studies utilized techniques (balanced ligamentous
inflammatory arc of the inflammatory reflex. tension in the occipitoatlantoid and cervicothoracic junc-
tions, and suboccipital decompression) that have been
Some OMT techniques have been demonstrated to in- demonstrated to increase parasympathetic tone, while the
crease parasympathetic nervous system activity, as other eight studies relied on rib raising, visceral, sacral, and
assessed by HRV analysis. Ruffini et al. [12] found that OMT lymphatic pump techniques. Overall, the results of the
significantly increased high frequency HRV. In that study, metaanalysis were inconsistent [26], which may be related
the OMT techniques were left at the discretion of the oper- to the fact that the majority of included studies did not uti-
ator but were limited to balanced ligamentous/membranous lize OMT techniques that have the potential to activate the
tension techniques and craniosacral techniques [12]. Fornari antiinflammatory efferent arc of the inflammatory reflex.
et al. [13] found that high frequency HRV was elevated
during an arithmetic stress test in subjects who received Thus, the hypothesis of our study was that the OMT
technique of decompression of the occipitoatlantal (OA-D)

Kania et al.: Antiinflammatory reflex activation by OA-D and taVNS 403

junction and noninvasive taVNS, both of which have been of $20.00 per study day, for a total amount of $60.00 for all three study
demonstrated to increase parasympathetic activity [12–16, days.
19–22] would also elicit an antiinflammatory response as
indicated by a reduction in proinflammatory cytokines. Our The study design and experimental protocol is shown in Figure 1.
rationale for selecting OA-D technique and taVNS is that The study consisted of three experimental groups: (1) OA-D; (2) taVNS;
these techniques are known to increase parasympathetic and (3) time control (no intervention). After written consent was ob-
tone and therefore have the potential to activate the tained, subjects were randomly assigned to a study group. For
cholinergic antiinflammatory reflex. Importantly, it has randomization into the three study groups, a six-sided die was rolled
been demonstrated that chronic inflammatory diseases like and numbers of one or two corresponded to the control group,
lupus are associated with reduced heart rate variability numbers of three or four corresponded to the OA-D group, and
markers of parasympathetic tone, even in the absence of overt numbers of five or six corresponded to the taVNS group. All subjects
cardiac involvement and symptoms [27]. Thus, it has been participated in three consecutive study days. At the beginning of the
suggested that interventions that increase parasympathetic first study day, body weight and height were determined and on all
tone also reduce inflammation in chronic inflammatory dis- three study days, upper-arm blood pressure (Omron 10 Series; Omron
orders [28]. Several OMT techniques have been demonstrated Healthcare, Inc.) was measured. Saliva samples were collected (Saliva
to increase parasympathetic tone. Curi et al. [16] demon- Collection Aid; Salimetrics, LLC) at the beginning and end of the ex-
strated that the technique of fourth ventricle compression is periments on each study day for determination of salivary cytokines.
associated with an increase in high-frequency HRV and a shift On each of the three study days, subjects were instrumented with
in autonomic balance to parasympathetic dominance. Like- electrocardiogram (ECG) electrodes and an inflatable finger cuff for
wise, Giles et al. [15] demonstrated a marked increase in high noninvasive continuous blood pressure monitoring (Finapres Fin-
frequency HRV when performing suboccipital decompres- ometer Pro; Finapres Medical Systems). During the protocol, subjects
sion, also suggesting increased parasympathetic tone. To our rested in the supine position on OMT examination tables and were
knowledge, there has been no prior study comparing the instructed to avoid unnecessary movements to ensure high quality
effectiveness of different OMT techniques in activating the ECG and blood pressure signals. Initially, a baseline recording of the
parasympathetic nervous system. We elected to utilize the ECG and arterial blood pressure was obtained for 30 min. Then OA-D
OA-D technique vs. the fourth ventricle compression tech- (Group 1: 5 min of OA-D followed by 10 min of rest), taVNS (Group 2:
nique because the fourth ventricle compression technique is 15 min of taVNS), or no intervention (Group 3: 15 min of rest) were
considered a more advanced technique compared to OA-D, applied, after which a final recovery recording of the ECG and arterial
which could be utilized by any osteopathic physician as it is blood pressure was obtained for another 30 min. Following the
included in the curriculum at all colleges of osteopathic baseline recording and following the 15-min intervention, the finger
medicine. As an experimental approach, we applied OA-D or cuff was deflated for 2–3 min and study participants were allowed to sit
taVNS to healthy study participants on three consecutive up or move their extremities.
days. The effect of these interventions on parasympathetic
modulation of cardiac function assessed by HRV and on Subjects
inflammation assessed by salivary cytokine content was
compared to the responses in participants who underwent a The study was conducted in healthy, adult (minimum age, 18 years)
time control without OA-D or taVNS application. subjects of both genders. Exclusion criteria included: pregnancy; any
medication that interferes with the autonomic nervous system or the
immune system (e.g., beta-blockers, steroids, TNF-α inhibitors); any
medical condition that affects the autonomic nervous system or the
immune system (e.g., autonomic neuropathy, pure autonomic failure,
rheumatic or autoimmune diseases, acquired autoimmune deficiency
syndrome); diabetes; and current drug or alcohol abuse.

Methods OMT intervention

Study design The overall premise of the study was that OMT techniques that in-
crease parasympathetic nervous system activity would also activate
The experiments of this study were conducted as part of the Burrell the cholinergic antiinflammatory reflex. OA-D is an osteopathic
College medical student summer research program in 2019 (June 2019 to technique that focuses on treating an articular compression between
September 2019). While this particular study was not publicly regis- the occiput and the atlas, which may improve conditions relating to
tered, an ongoing follow-up study is registered with ClinicalTrials.gov the path of the vagus nerve as it exits the skull. This technique has
(NCT04177264). The study was approved by the Institutional Review been demonstrated to markedly increase high frequency HRV, sug-
Board at Burrell College of Osteopathic Medicine (IRB# 0046_2019) and gesting increased cardiac parasympathetic modulation [15]. For OA-D,
all study participants provided written informed consent. Subjects were the subjects were lying in the supine position on an OMT examination
compensated for their time and effort with VISA gift cards in the amount table, while the investigator cradled the subjects’ heads with their
hands and finger pads along the inferior aspect of inion, reaching
toward the occipitoatlantal joint. The investigator then applied gentle

404 Kania et al.: Antiinflammatory reflex activation by OA-D and taVNS

Figure 1: Experimental protocol. The study consisted of three study days. At the beginning of the first study day, written consent was
obtained, potential exclusion criteria were assessed through a questionnaire, and height, weight, and arm cuff blood pressure were
measured. On each study, day saliva samples were collected at the beginning and end of the study day, for a total of six saliva samples from
each participant. Hemodynamic recordings of arterial blood pressure (finger plethysmography) and heart rate (ECG) were obtained on each
study day. The experimental protocol during the hemodynamic recordings on each study day consisted of a 30 min baseline recording, a 15 min
intervention, and a 30 min recovery recording. In three experimental groups, the interventions consisted of either 5 min of occipitoatlantal
decompression (OA-D) followed by 10 min of rest (OA-D group), or 15 min of transcutaneous auricular vagus nerve stimulation (taVNS group), or
15 min of rest (control group). The shown example of the hemodynamic recording is from a participant in the OA-D group.

anterior and cephalad traction to the occiput, while bringing the el- 500 µs resulted in the greatest bradycardic effect. Since the longest
bows together. This motion resulted in supination of the hands and pulse width of the EMS 7500 device is 300 µs, we used a combination
separation of the fingers, creating an anterolateral force vector to of 10 Hz stimulation frequency and 300 µs pulse width for the current
either side of the foramen magnum [29]. The gentle anterior and study. The stimulation current was determined individually for each
cephalad traction was then maintained for 5 min. All investigators subject by slowly increasing the stimulation current until the subjects
performing OA-D (K.N.W., A.P.K., M.L.O., and J.N.O.) had practiced felt a mild tingling sensation at the site of the electrode. Then the
modulating their palpatory efforts, specifically in light palpation, current was gradually reduced until the tingling sensation dis-
while applying the technique of OA-D such that there was no differ- appeared or was just barely felt. This current was then used for 15 min
ence among their application of OA-D when evaluated by a of senior of taVNS.
investigators with many years of experience in OMT (A.M.K.).

Transcutaneous auricular vagus nerve stimulation Baroreceptor-heart rate reflex analysis
(taVNS)
The baroreceptor–heart rate reflex sensitivity was calculated using the
Current literature [19–22] strongly suggest that taVNS increases car- sequence technique as first described by Bertinieri et al. [32]. Briefly,
diac parasympathetic tone and therefore may also activate the the freely available HemoLab software [33] was used to identify
cholinergic antiinflammatory reflex [10]. A bipolar clip electrode spontaneously occurring sequences of four or more consecutive heart
connected to a transcutaneous electrical nerve stimulator (EMS 7500; beats, where systolic blood pressure and pulse interval change in the
Current Solutions, LLC) was applied to the cymba conchae that is same direction. The average of the slopes of all identified sequences
innervated by the auricular branch of the vagus nerve (Arnold’s nerve) (changes in pulse interval on y-axis; changes in systolic blood pres-
[30]. The clip electrode was applied to the ear lobe such that the sure on x-axis) is a measure of the baroreceptor–heart rate reflex
cathode was placed at the cavum of the concha and the anode was sensitivity. This technique has been found to be reliable [34] when
placed at the opposing site of the back of the auricle that is also compared to the gold standard of the Oxford method [35]. The premise
innervated by the vagus nerve. Badran et al. [31] studied different of this study is that activation of the parasympathetic nervous system
stimulation frequencies (1, 10, and 25 Hz) and stimulation pulse widths by OA-D or taVNS elicits an antiinflammatory response; thus, we were
(100, 200, and 500 µs) and found that the combination of 10 Hz and primarily interested in parasympathetic baroreflex control. Since the
parasympathetic nervous system – other than the sympathetic

Kania et al.: Antiinflammatory reflex activation by OA-D and taVNS 405

nervous system – can affect heart rate rapidly within one heartbeat, we (Salimetrics, LLC; Carlsbad, CA). Saliva flow rate was determined as
did not use a time delay between the blood pressure and heart rate the ratio of the saliva volume produced to the time needed to pro-
time series when identifying baroreflex sequences. duce that volume. All saliva samples were stored at −80 °C. Upon
completion of the study, the saliva samples were shipped to Sali-
Parasympathetic modulation of cardiac function metrics for analysis (Salivary Cytokine Panel; Salimetrics, LLC). The
cytokine panel included interleukin-1β (IL-1β, detection range:
Parasympathetic modulation of cardiac function was assessed by HRV 0.05–2,256 pg/mL), interleukin-6 (IL-6, detection range: 0.06–
analysis. Specifically, the time-domain HRV parameter of the root 3,068 pg/mL), interleukin-8 (IL-8, detection range: 0.07–2,336 pg/mL),
mean square of successive RR interval differences (RMSSD) and the and TNF-α (detection range: 0.04–1,360 pg/mL) and were deter-
frequency-domain HRV parameter of high frequency spectral power mined by enzyme-linked immunosorbent assay. All measured
were determined. These two HRV parameters are well-established cytokine levels were within the detection range of the assays.
measures of parasympathetic modulation of cardiac function [11].
First, all RR intervals were derived from the ECG recordings (sample Statistical analyses
rate 1,000 Hz) for the 30 min of baseline recording, the recordings
during the interventions (15 min for control and taVNS and 5 min for All data are shown as means ± standard error of the mean (SEM). For
OA-D), and for the 30 min of recovery recording that followed the statistical comparison between groups, a one-way analysis of variance
intervention. For RMSSD, these beat-by-beat RR intervals were (ANOVA) for independent measures was used. Post-hoc Fisher tests were
segmented into 5 min segments with 50% overlap. From these seg- used if the ANOVA revealed statistical significance to identify differences
ments, RMSSD was calculated using the HemoLab software [33]. The between individual groups. For statistical comparison between data ob-
RMSSD values from all overlapping segments within the baseline, tained on the three study days, repeated-measures one-way ANOVA was
intervention, and recovery periods were averaged for statistical used. In case of statistical significance in the ANOVA, post-hoc Fisher tests
analysis. For frequency domain HRV analysis, beat-by-beat heart rate were used to identify significant differences between individual study
values were derived for the three sections of the experimental protocol days. For statistical comparison between different phases of the experi-
(baseline, intervention, recovery). These beat-by-beat heart rate time mental protocol within a study day (i.e., baseline recording, intervention,
series were spline interpolated and resampled at an equidistant or recovery recording) repeated-measures ANOVA was used. In case of
sampling rate of 25 Hz. From these 25 Hz time series, power spectra statistical significance in the ANOVA, post-hoc Fisher tests were used to
were calculated for overlapping segments (50% overlap, 4,096 data identify significant differences between the phases of the experimental
points, 164 s). The power spectra of the overlapping segments for each protocolInstead of a comprehensive power analysis for all parameters and
subject and each experimental condition (baseline, intervention, re- for all statistical comparisons made, we present the power analysis for the
covery) were averaged for statistical analysis. High frequency spectral comparison for the IL-1β responses among the three experimental groups
power was calculated as the area under the curve of the power spectra as a representative example. We chose this cytokine because our conclu-
in the frequency band of 0.15–0.4 Hz. sion of an antiinflammatory effect of taVNS is primarily based on the IL-1β
responses. The power analysis was performed using the power.anova.test
Sympathetic modulation of vascular tone function in the R statistical analysis software [38] according to Cohen [39].
The power analysis showed that the statistical power for comparing the
Analysis of sympathetic modulation of vascular tone was also IL-1β response between the three experimental groups was 59.4% at the
performed using the HemoLab software [33]. Using the blood p=0.05 level and 71.5% at the p=0.10 level. As with all statistical tests, there
pressure waveforms obtained from the Finapres device (1,000 Hz is a tradeoff between the p value (alpha error) and the statistical power
sampling rate), beat-by-beat systolic blood pressure values were (1-beta error). To obtain acceptable powers, we chose to present statistical
derived for the three sections of the experimental protocol (base- results at both the p=0.05 and p=0.10 levels as appropriate. We chose to
line, intervention, recovery). These beat-by-beat time series were use the terms “trend” or “tended” for results that only provided an
spline interpolated and resampled at an equidistant sampling rate acceptable statistical power (>70%) at the p=0.10 level.
of 25 Hz. From these 25 Hz time series, power spectra were calcu-
lated for overlapping segments (50% overlap, 4,096 data points, Results
164 s). The power spectra of the overlapping segments of each
subject and each experimental condition (baseline, intervention, Participant characteristics
recovery) were averaged for statistical analysis. Low frequency
spectral power of systolic blood pressure variability (LFSYS), a A total of 27 study participants were enrolled in the study
measure of sympathetic modulation of vascular tone [36, 37], was (n=10 for control; n=8 for OA-D; n=9 for taVNS). Two partic-
calculated as the area under the curve of the power spectra in the ipants (1 from the OA-D group and one from the taVNS group)
frequency range of 0.04–0.15 Hz. were excluded from hemodynamic analyses because the ECG
revealed sick sinus syndrome or atrial fibrillation, respec-
Determination of salivary cytokines tively. Two additional subjects (both from the control group)
did not complete all three study days.
Saliva samples were collected before and after the experiments on
all three study days using the SalivaBio Saliva Collection Aid

406 Kania et al.: Antiinflammatory reflex activation by OA-D and taVNS

Table : Participant characteristics. group, but not in the taVNS group (Figure 2). While there
was no change in MAP throughout the experimental pro-
Parameter Control (n=) OA-D (n=) taVNS (n=) tocol on individual study days in the control group, MAP
increased toward the end of the experimental protocol in
Sex  women/ men  women/ men  women/ men the OA-D and taVNS groups (pooled data, Figure 2). The
antihypertensive effect of OA-D observed during the
Age in years*  ±  (–)  ±  (–)  ±  (–) baseline recordings on the second and third study days
compared with the baseline recording on the first study day
BMI in kg/m* . ± . . ± . . ± . was confirmed by the blood pressure readings obtained at
the beginning of each study day using an arm cuff-based
(.–.) (.–.) (.–.)† blood pressure monitor. These blood pressure readings
showed statistically lower diastolic blood pressure values
*Values are presented as mean ± standard error of the mean (range); on the second (mean, 80 ± 4 mmHg; p<0.05) and third
†: p<. vs. control group. BMI, body mass index; OA-D, (mean, 79 ± 5 mmHg; p<0.05) study days compared with
occipitoatlantal decompression; taVNS, transcutaneous auricular the first study day (mean, 84 ± 5 mmHg) in the OA-D group,
vagus nerve stimulation. but not in the time control or taVNS groups. Heart rate
generally decreased during the time course of the experi-
The age, gender distribution, and BMI of the subjects in mental protocol in all three groups and on all three study
days with no significant differences between groups.
the three experimental groups are provided in Table 1. There
Effects of OA-D and taVNS on baroreceptor-
were more women (20) than men (7). The age of the partici- heart rate reflex sensitivity

pants (mean, 52.7 ± 3.8 years) was not statistically different In the time control and taVNS groups, baroreceptor-heart
between the three groups. However, body mass index (BMI) rate reflex sensitivity (BRS) did not change significantly
was higher in the taVNS group (mean, 31.0 ± 2.4 kg/m2; throughout the three study days or during the experimental
p<0.05) compared with the control group (mean, 24.5 ± 1.2 kg/ protocol on each study day (Figure 3). In contrast, BRS
m2). BMI of the subjects in the OA-D group (mean, baseline values increased significantly from the first to the
27.0 ± 1.4 kg/m2) was not different from the BMI of the sub-
jects in the two other groups.

Hemodynamic effects of OA-D and taVNS

In the time control group, baseline mean arterial blood
pressure (MAP) was similar on all three study days. In
contrast, baseline MAP was lower on the second and third
study day compared to the first study day in the OA-D

Figure 2: Effect of the control intervention
(CTR, top), occipitoatlantal decompression
(OA-D, middle), and transcutaneous auric-
ular vagus nerve stimulation (taVNS, bot-
tom) on mean arterial blood pressure (MAP)
on each of the three study days and for the
data from all three study days pooled.
Changes (Δ) from the baseline values on
the first study day are shown. The numbers
next to the baseline values of each study
day are the absolute baseline MAP
values ± SEM (in mmHg) and number of
participants (or number of experiments for
the pooled data), respectively. Data are
shown as means ± standard error of the
mean (SEM). †: p<0.05 vs. baseline MAP
value on the first study day; ‡: p<0.05 vs.
baseline MAP value on the same study day.

Kania et al.: Antiinflammatory reflex activation by OA-D and taVNS 407

Figure 3: Effect of the control intervention
(CTR, top), occipitoatlantal decompression
(OA-D, middle), and transcutaneous auric-
ular vagus nerve stimulation (taVNS, bot-
tom) on baroreceptor-heart rate reflex
sensitivity (BRS) on each of the three study
days and for the data from all three study
days pooled. Changes (Δ) from the baseline
values on the first study day are shown. The
numbers next to the baseline values of
each study day are the absolute baseline
BRS values ± SEM (in ms/mmHg) and
number of participants (or number of ex-
periments for the pooled data). Data are
shown as means ± standard error of the
mean (SEM). †: p<0.05 vs. baseline BRS
value on the first study day; ‡: p<0.05 vs.
baseline BRS value on the same study day.

second and third study days in the OA-D group. Further- increase in RMSSD (mean ± SEM, +5.0 ± 1.6 ms; p<0.05) at
more, pooling the data from all three study days revealed the end compared to the beginning of the experimental
that BRS increased significantly during the acute OA-D protocol in the taVNS group. Consistent with these RMSSD
intervention, but not during taVNS or during the control responses to taVNS, pooling the data from all three study
intervention. days revealed that taVNS increased high frequency spectral
power of HRV during the recovery period at the end of the
Cardiac autonomic responses to OA-D and experimental protocol compared to the baseline recording
taVNS at the beginning of the experimental protocol (mean ±
SEM,+0.65 ± 0.34 bpm2; p<0.05; Figure 5). These findings
In the time control group, there were no significant differ- suggest that taVNS increases cardiac parasympathetic tone.
ences for RMSSD (Figure 4) or high frequency spectral power However, this cardiac autonomic effect of taVNS did not last
of HRV (Figure 5) between the baseline values on the three into the next day because no significant changes in RMSSD
experimental days or throughout the experimental protocol or high frequency spectral power of HRV were observed
(baseline, intervention, recovery) on each individual study during the baseline recordings of the second and third study
day. In contrast, OA-D caused a significant increase in days compared to the first study day. While OA-D appears to
baseline RMSSD (mean ± standard error of the mean have had a more chronic effect on parasympathetic tone that
SEM,+19 ± 10 ms; p<0.05) on the second compared to the lasted into the next study day, taVNS appeared to elicit a
first study day (Figure 4) and trends for higher high fre- more acute effect because it was most apparent during the
quency spectral power of HRV on the second (mean ± recovery period and did not persist into the following study
SEM,+2.2 ± 1.6 bpm2; p=0.08) and third (mean ± day.
SEM,+2.0 ± 1.1 bpm2; p=0.11) study days (Figure 5). These
findings suggest that OA-D increases parasympathetic Effects of OA-D and taVNS on sympathetic
modulation of cardiac function and that this effect persists modulation of vascular tone
into the following day. In the taVNS group, RMSSD
increased significantly during the recovery period compared Low frequency systolic blood pressure variability (LFSYS)
to the baseline recording on the second study day (Figure 4). has been demonstrated to be modulated by fluctuations
Pooling the data from all three study days confirmed this in sympathetic-medicated vasomotor tone [40, 41]. In the

408 Kania et al.: Antiinflammatory reflex activation by OA-D and taVNS

Figure 4: Effect of the control intervention
(CTR, top), occipitoatlantal decompression
(OA-D, middle), and transcutaneous auric-
ular vagus nerve stimulation (taVNS, bot-
tom) on root mean square of successive
differences (RMSSD) on each of the three
study days and for the data from all three
study days pooled. Changes (Δ) from the
baseline values on the first study day are
shown. The numbers next to the baseline
values of each study day are the absolute
baseline RMSSD values ± SEM (in ms) and
number of participants (or number of ex-
periments for the pooled data). Data are
shown as means ± standard error of the
mean (SEM). †: p<0.05 vs. baseline values
on the first study day; ‡: p<0.05 vs. baseline
values on the same study day.

time control group and taVNS groups, no statistically Glandular autonomic responses to OA-D and
significant changes in LFSYS were observed throughout taVNS
the experimental protocol or between individual study days
(Figure 6). However, in the OA-D group, a marked decrease Parasympathetic innervation of the salivary glands increases
in LFSYS was apparent during the application of OA-D and salivary flow rate primarily through M3-muscarinic receptor
during the recovery period at the end of the experimental stimulation [42]. In all three study groups, saliva flow rate did
protocol. This reduction in sympathetic modulation of not differ before and after the experimental protocol on any
vascular tone in the OA-D group did not persist into the next study day. Only in the time control group saliva flow rate
study day because the baseline values on the second and increased from the first to the second (mean ± SEM,
third study days were not different from the baseline values +0.23 ± 0.09 mL/min; p<0.05) study day. No significant
on the first study day. This finding suggests that OA-D changes in saliva flow rate from the first to subsequent study
acutely reduces sympathetic vascular tone. days were observed in the OA-D or taVNS groups.

Figure 5: Effect of the control intervention
(CTR, top), occipitoatlantal decompression
(OA-D, middle), and transcutaneous auric-
ular vagus nerve stimulation (taVNS, bot-
tom) on high frequency (HF) heart rate (HR)
variability on each of the three study days
and for the data from all three study days
pooled. Changes (Δ) from the baseline
values on the first study day are shown. The
numbers next to the baseline values are
absolute baseline values for HF spectral
power ± SEM (units as on y-axes) and
number of participants (or number of ex-
periments for the pooled data). Data are
shown as means ± standard error of the
mean (SEM). (†): p<0.10 vs. baseline values
on the first study day; ‡: p<0.05 vs. baseline
values; (‡): p<0.1 vs. baseline values.

Kania et al.: Antiinflammatory reflex activation by OA-D and taVNS 409

Figure 6: Effect of the control intervention
(CTR, top), occipitoatlantal decompression
(OA-D, middle), and transcutaneous auric-
ular vagus nerve stimulation (taVNS, bot-
tom) on low frequency systolic blood
pressure variability (LFSYS) on each of the
three study days and for the data from all
three study days pooled. Changes (Δ) from
the baseline values on the first study day
are shown. The numbers below the base-
line values are absolute baseline values for
LFSYS ± SEM (in mmHg2) and number of
participants (or number of experiments for
the pooled data). Data are shown as
means ± standard error of the mean (SEM).
‡: p<0.05 vs. baseline values on the same
study day.

Effects of OA-D and taVNS on salivary cytokine values that were higher than 2 standard deviations
cytokines above the mean of the respective baseline cytokine levels on the
first study day (before any interventions were done). Outliers
In some study participants, salivary cytokine plasma levels were defined by this criterion were excluded from data analysis.
excessively high. Markedly elevated salivary cytokine levels
have been reported in patients with periodontal diseases [43, 44]. The baseline concentrations of all four salivary cytokines
Since we did not assess oral health status, we defined outliers as (IL-1β, IL-6, IL-8, and TNF-α) on each of the three study days
and for the three study days pooled are listed in Table 2.

Table : Baseline salivary cytokine concentrationsa.

Study day Control (n=–) OA-D (n=–) taVNS (n=–)

Interleukin-β (IL-β) in pg/mL (detection range: .–, pg/mL)

 . ± . . ± . (*) . ± . (*)
. ± . . ± . (*)
 . ± . . ± . . ± .
. ± . . ± .*
 . ± .
. ± . . ± .
– pooled . ± . . ± . (†) . ± .
. ± .
Interleukin- (IL-) in pg/mL (detection range: .–, pg/mL) . ± . † . ± .
. ± .
 . ± .  ± 
 ±   ± 
 . ± .  ±   ±  (†)
 ±  †  ± 
 . ± .  ± 
. ± .
– pooled . ± . . ± . . ± .
. ± . . ± .
Interleukin- (IL-) in pg/mL (detection range: .–, pg/mL) . ± . † . ± . (*)
. ± .
  ± 

  ±  (†)

  ± 

– pooled  ± 

Tumor necrosis factor-α (TNF-α) in pg/mL (detection range: .–, pg/mL)

 . ± .

 . ± .

 . ± .

– pooled . ± .

aValues are presented as mean ± standard error of the mean (range); *: p<. vs. control group; (*): p<. vs. control group. †: P<. vs. day
; (†): P<. vs. day . OA-D, occipitoatlantal decompression; taVNS, transcutaneous auricular vagus nerve stimulation.

410 Kania et al.: Antiinflammatory reflex activation by OA-D and taVNS

Figure 7: Effect of the control intervention
(left), occipitoatlantal decompression
(OA-D, middle), and transcutaneous auric-
ular vagus nerve stimulation (taVNS, right)
on inflammatory salivary cytokines inter-
leukin-1β (IL-1β), interleukin-6 (IL-6),
interleukin-8 (IL-8), and tumor necrosis
factor-α (TNF-α) on each of the three study
days and for the data from all three study
days pooled (bars). Differences (Δ) between
the cytokine levels following the experi-
mental protocol minus the cytokine levels
before the experimental protocol are
shown. Data are shown as means ± stan-
dard error of the mean (SEM). *: p<0.05 vs.
control group; (*): p<0.10 vs. control group.

Baseline cytokine levels on individual study days did not experimental groups. These findings indicate that the
differ significantly (p<0.05) between groups (Table 2). experimental protocol was associated with a proin-
However, when the data from all three study days were flammatory response that was blunted to some extent by
pooled baseline IL-1β (p<0.05) and TNF-α (p<0.10) levels taVNS. There were no significant differences in the increases
were lower in the taVNS group than in the time control in cytokine levels between the three study days in any group,
group. Importantly, in the OA-D group, baseline salivary suggesting that the proinflammatory effect of the experi-
IL-6, IL-8, and TNF-α concentrations were significantly mental protocol was consistent on all three study days.
(p<0.05) less on the third study day compared to the first
study day (Table 2), suggesting that successive OA-D ap- Discussion
plications elicited an antiinflammatory effect during the
3-day protocol. The major finding of this study is that both noninvasive taVNS
and OA-D, an OMT technique known to increase para-
Figure 7 shows the changes (Δ values) in salivary cyto- sympathetic tone, elicited changes in salivary cytokines
kine levels from the beginning to the end of each study day as consistent with antiinflammatory actions. Specifically, taVNS
well as for the data from all three study days pooled for the blunted the increase in salivary interleukin-1β (IL-1β) that was
three experimental groups. Generally, salivary cytokines associated with the experimental protocol (Figure 7) and
levels (IL-1β, IL-6, IL-8, TNF-α) increased from the beginning successive applications of OA-D on three consecutive days
to the end of the experimental protocol on each study day reduced salivary IL-6, IL8, and TNF-α baseline levels on the
(i.e., all Δ values in Figure 7 are positive). Specifically, when third compared to the first study day (Table 2). A secondary
the data from all three study days were pooled, all four cy- finding was that OA-D lowered diastolic and mean (Figure 2)
tokines increased significantly (p<0.05) from the beginning to blood pressure at the beginning of the second and third study
the end of the protocol in all three experimental groups. This day compared to the first study day.
finding suggests that the experimental protocol elicited a
proinflammatory response. However, pooling the data from Interestingly, in all three study groups, salivary concen-
all three study days also revealed that the increase in salivary trations of the inflammatory cytokines IL-1β, IL-6, IL-8, and
cytokines elicited by the experimental protocol differed TNF-α had significantly increased at the end of the experi-
among groups. In the taVNS group, the increase in salivary mental protocol compared to before the experimental proto-
IL-1β was significantly (p<0.05) less than in the control group. col (positive Δ values in Figure 7). This increase in
A similar trend (p<0.1) was observed for TNF-α in the taVNS inflammatory cytokines in response to the experimental
group. IL-6 and IL-8 increased similarly in all three

Kania et al.: Antiinflammatory reflex activation by OA-D and taVNS 411

protocol was highly robust and consistently observed in all delayed but prolonged effect on parasympathetic tone, while
study participants. It has been demonstrated that acute stress taVNS exerts a more immediate effect of shorter duration.
[45], including mental stress [46], increases salivary IL-1β,
IL-6, and TNF-α concentrations. Thus, it is not surprising that The HRV analysis of our data confirms previous studies by
these salivary cytokines increased in response to our experi- others [12–16, 19–22] suggesting that OA-D and taVNS increases
mental protocol, which presumably was perceived as a parasympathetic modulation of cardiac function, although the
stressful situation by most study participants. The increase in time course of this effect seems to differ for OA-D and taVNS.
proinflammatory cytokines in response to the experimental Based on this finding, we addressed the question of whether
protocol may be related to mobilization of preexisting cyto- this increased parasympathetic tone is associated with activa-
kine pools rather than de novo synthesis because the time tion of the cholinergic antiinflammatory pathway [5, 6]. To
between saliva sampling was less than 2 h and because address this question, we determined salivary cytokine con-
baseline levels on the second and third study days were not centrations before and after the experimental protocol on all
elevated compared with the first study day in any group. three study days. Interestingly, the differential time courses of
However, the lower baseline IL-6, IL-8, and TNF-α cytokine parasympathetic activation with OA-D and taVNS are in line
levels at the third study day compared with the first study day with the time courses of the antiinflammatory effects of OA-D
in the OA-D group may potentially be mediated by inhibition and taVNS. OA-D elicited delayed and prolonged effects on
of cytokine de novo synthesis. both the parasympathetic nervous system and salivary IL6,
IL-8, and TNF-α levels that were only significant when the
The hypothesis of this study was that noninvasive and baseline cytokine levels on the third study day were compared
nonpharmacologic interventions that have been suggested to with those from the first study day. In contrast, the effects of
increase parasympathetic tone [12–16, 19–22] also activate the taVNS on parasympathetic function and on salivary cytokine
cholinergic antiinflammatory pathway [5, 6]. To test this hy- levels were more acute. The effects of taVNS on IL-1β and its
pothesis, we assessed autonomic function by HRV analysis trend on TNF-α were only apparent as a blunted increase in
and the immune response to OA-D and taVNS by determining cytokine levels in response to the acute experimental protocol
salivary cytokine concentrations. RMSSD is a time-domain (Figure 7). In addition, this effect of taVNS did not last into the
HRV parameter that has been considered a specific marker of next study day, because the baseline levels on the three study
parasympathetic modulation of sinus node function [11]. Only days did not differ in the taVNS group (Table 2). The consistent
in the taVNS group, RMSSD increased significantly toward time courses of the parasympathetic and antiinflammatory
the end of the experimental protocol (pooled data in Figure 4), responses, with delayed responses in the OA-D group and more
suggesting that taVNS increases cardiac parasympathetic acute responses in the taVNS group, further suggest that the
tone. This finding is even more remarkable considering that antiinflammatory effects of OA-D and taVNS are linked to the
BMI was higher in the taVNS group than in the time control activation of the parasympathetic nervous system and medi-
group (Table 1) and that BMI is inversely correlated with ated through the cholinergic antiinflammatory pathway [5, 6].
RMSSD [47, 48]. Interestingly, OA-D but not taVNS applica-
tion on the first study day resulted in elevated RMSSD during While taVNS significantly (p<0.05) blunted the IL-1β
the initial baseline recording on the second study day response to the experimental protocol, there was only a
(Figure 4). These findings suggest that OA-D has a delayed but trend (p<0.10) for reduced TNF-α responses with taVNS. It
prolonged effect on cardiac parasympathetic tone, while is generally assumed that an increase in IL-1β precedes the
taVNS has a more immediate effect of shorter duration. This subsequent TNF-α response. It has been suggested that
suggestion is supported by frequency-domain HRV analysis IL-1β induces TNF-α gene expression [49] and an increase
that demonstrated that high frequency spectral power of in IL-1β plasma levels preceded TNF-α plasma levels in rats
HRV, another specific measure of parasympathetic modula- following a surgical intervention [50]. These different time
tion of cardiac function [11], was elevated during the recovery courses for the IL-1β and TNF-α responses to proin-
recording at the end of the experimental protocol compared to flammatory stimuli may explain why the effect of taVNS on
the baseline recording only in the taVNS group (pooled data salivary IL-1β was statistically significant, while there was
in Figure 5). In line with the finding for RMSSD, high fre- only a trend for the effect of taVNS on TNF-α.
quency spectral power of HRV tended (p<0.10) to be elevated
on the second and third study day compared to the first study Plasma levels of proinflammatory cytokines have been
day only in the OA-D group (Figure 5). Thus, the results of the reported to be markedly elevated in patients with chronic
HRV analysis is aligned with the notion that OA-D has a inflammatory diseases, including rheumatoid arthritis
[51, 52] and lupus [53]. Furthermore, it has been suggested
that cytokine levels correlate with disease severity [52].
There is also data suggesting that cytokine levels in plasma

412 Kania et al.: Antiinflammatory reflex activation by OA-D and taVNS

correlate with those in saliva [54, 55]. Thus, it is reasonable each study day. This would have required two blood draws on
to assume that salivary cytokine levels also correlate with each of the three study days for a total of six blood draws per
clinical disease severity. Our finding that both OA-D and study participants. Such a protocol might have limited the
taVNS reduced salivary cytokines offers the possibility that willingness of volunteers to participate in the study. While we
these techniques may also reduce the markedly elevated acknowledge that plasma cytokine levels would have been
plasma cytokine levels in patients with chronic inflam- ideal, there are data demonstrating correlations between
matory diseases, which may translate into less disease serum and salivary cytokine concentrations not only in con-
severity. While the results of our study are insufficient for ditions affecting the oral cavity [62] but also in response to
this conclusion, they provide a rationale for a clinical systemic conditions, such as pediatric obesity [55] and soci-
follow-up study to test this hypothesis. ocognitive stress [54]. Another limitation related is that
salivary cytokine levels can increase substantially in pa-
A secondary finding of this study was that OA-D per- tients with oral health issues [43, 44]. Some excessively
formed on the first and second study days reduced mean high salivary cytokine levels were observed and subse-
(Figure 2) and diastolic blood pressure on the following study quently identified as outliers and excluded from data
days, an effect that was not observed in the time control or analysis. While we did not assess oral health, it is possible
taVNS groups. This antihypertensive effect of OA-D is that these outliers originated from study participants with
consistent with findings by Curi et al. [16], who reported that oral health issues.
the OMT technique of fourth ventricle compression decreased
systolic and diastolic blood pressure in hypertensive study Conclusions
participants. In our study, the OMT technique of OA-D acutely
reduced low frequency spectral power of systolic blood The results of this study suggest that the osteopathic technique
pressure, a measure of sympathetic modulation of vascular of OA-D as well as non-invasive taVNS elicit antiinflammatory
tone. Although this sympatholytic effect did not persist into effects as indicated by a significantly blunted salivary IL-1β
the following study day, the acute reduction in vascular response to the experimental protocol in the taVNS group
sympathetic tone may still have triggered an antihypertensive compared to the time-control group and by significantly
response through activation of local vasodilator mechanisms, reduced baseline levels of salivary IL-6, IL-8, and TNF-α on the
such as the endothelial nitric oxide system or other mecha- third compared to the first study day in the OA-D group. It is
nisms that persisted into the following day. Furthermore, OA-D reasonable to assume that these antiinflammatory effects
increased baroreceptor-heart rate reflex sensitivity, an effect of OA-D and taVNS are mediated through activation of
that persisted into the following study day. It is well estab- the cholinergic antiinflammatory pathway, because both
lished that an increase in baroreceptor reflex sensitivity can techniques were associated with increased time- and
elicit antihypertensive effects [56–59]. In fact, stimulating frequency-domain HRV parameters that reflect para-
baroreflex afferents has recently been suggested as a thera- sympathetic modulation of cardiac function. Furthermore,
peutic approach for treatment-resistant hypertension [60, 61]. OA-D elicited an anti-hypertensive response that was reflected
Thus, the antihypertensive effect of OA-D may be the result of a in a significant decrease in baseline mean and diastolic blood
combination of acutely reduced vascular sympathetic tone and pressure on the second and third study days compared to the
chronically augmented baroreflex sensitivity. first study day. This antihypertensive effect of OA-D may be
mediated by an acute inhibition of vascular sympathetic tone
Limitations and/or chronic augmentation of baroreceptor reflex function.
Taken together, the results of this study provide a rationale for
This study enrolled relatively healthy subjects who did not conducting clinical studies that incorporate OMT techniques
have known autoimmune illnesses or other diseases associ- that elevate parasympathetic tone into the treatment plan for
ated with a high degree of inflammation such as diabetes chronic inflammatory conditions, such as rheumatoid dis-
mellitus. Thus, this serves as a baseline study with which eases or autoimmune disorders. Furthermore, the anti-
future studies can compare results to determine the efficacy hypertensive effect of OA-D observed in this study may be
of OA-D and/or taVNS in various disease states. Salivary utilized as supplemental treatment in patients with hy-
instead of plasma cytokines were used to investigate the pertension. It appears important to conduct follow-up
immune system response to OA-D or taVNS. The rationale for studies investigating the effects of OA-D and/or taVNS in
using salivary instead of plasma cytokine levels was that we patients with inflammatory or auto-immune diseases or
wanted to assess the temporal pattern of cytokine levels over in hypertensive patients, to test if the results obtained in
the three study days and the effect of our interventions on

Kania et al.: Antiinflammatory reflex activation by OA-D and taVNS 413

this preliminary study in healthy participants translates 8. Wang H, Liao H, Ochani M, Justiniani M, Lin X, Yang L, et al.
into improved clinical outcomes in patients. Cholinergic agonists inhibit HMGB1 release and improve survival
in experimental sepsis. Nat Med 2004;10:1216–21.
Research funding: This study was supported by funding
through the Research Office of Burrell College of 9. Martelli D, McKinley MJ, McAllen RM. The cholinergic anti-
Osteopathic Medicine, Las Cruces, NM. A follow-up study inflammatory pathway: a critical review. Auton Neurosci 2014;
is being supported by a grant from the American 182:65–9.
Osteopathic Association (Grant No.: 19137759).
Author contributions: All authors provided substantial 10. Pavlov VA, Ochani M, Gallowitsch-Puerta M, Ochani K, Huston JM,
contributions to conception and design, acquisition of Czura CJ, et al. Central muscarinic cholinergic regulation of the
data, or analysis and interpretation of data; all authors systemic inflammatory response during endotoxemia. Proc Natl
drafted the article or revised it critically for important Acad Sci U S A 2006;103:5219–23.
intellectual content; all authors gave final approval of the
version of the article to be published; and all authors agree 11. Task Force of the European Society of Cardiology and the North
to be accountable for all aspects of the work in ensuring that American Society of Pacing and Electrophysiology. Heart rate
questions related to the accuracy or integrity of any part of variability: standards of measurement, physiological
the work are appropriately investigated and resolved. interpretation and clinical use. Task Force of the European
Competing interests: Dr. Harald Stauss developed the Society of Cardiology and the North American Society of Pacing
HemoLab software that was used in this study and he and Electrophysiology. Circulation 1996;93:1043–65.
makes this software freely available through his website: http://
www.haraldstauss.com/HaraldStaussScientific/hemolab. 12. Ruffini N, D’Alessandro G, Mariani N, Pollastrelli A, Cardinali L,
Informed consent: All study participants provided written Cerritelli F. Variations of high frequency parameter of heart rate
informed consent. variability following osteopathic manipulative treatment in
Ethical approval: This study was approved by the healthy subjects compared to control group and sham therapy:
Institutional Review Board at Burrell College of Osteopathic randomized controlled trial. Front Neurosci 2015;9:272.
Medicine (IRB# 0046_2019).
13. Fornari M, Carnevali L, Sgoifo A. Single osteopathic manipulative
References therapy session dampens acute autonomic and neuroendocrine
responses to mental stress in healthy male participants. J Am
1. Strand V, Kimberly R, Isaacs JD. Biologic therapies in Osteopath Assoc 2017;117:559–67.
rheumatology: lessons learned, future directions. Nat Rev Drug
Discov 2007;6:75–92. 14. Henley CE, Ivins D, Mills M, Wen FK, Benjamin BA. Osteopathic
manipulative treatment and its relationship to autonomic
2. Strand V. The emerging role of biologics in rheumatic disease. J nervous system activity as demonstrated by heart rate variability:
Rheumatol Suppl 1992;33:40–5. a repeated measures study. Osteopath Med Prim Care 2008;2:7.

3. Joensuu JT, Huoponen S, Aaltonen KJ, Konttinen YT, Nordstrom D, 15. Giles PD, Hensel KL, Pacchia CF, Smith ML. Suboccipital
Blom M. The cost-effectiveness of biologics for the treatment of decompression enhances heart rate variability indices of cardiac
rheumatoid arthritis: a systematic review. PLoS One 2015;10: control in healthy subjects. J Altern Complement Med 2013;19:
e0119683. 92–6.

4. Singh JA, Wells GA, Christensen R, Tanjong Ghogomu E, Maxwell L, 16. Curi ACC, Maior Alves AS, Silva JG. Cardiac autonomic response
MacDonald JK, et al. Adverse effects of biologics: a network after cranial technique of the fourth ventricle (CV4) compression
meta-analysis and Cochrane overview. Cochrane Database Syst in systemic hypertensive subjects. J Bodyw Mov Ther 2018;22:
Rev 2011:Cd008794. https://doi.org/10.1002/14651858. 666–72.
CD008794.pub2.
17. Ellrich J. Transcutaneous auricular vagus nerve stimulation. J Clin
5. Tracey KJ. Physiology and immunology of the cholinergic Neurophysiol 2019;36:437–42.
antiinflammatory pathway. J Clin Invest 2007;117:289–96.
18. Butt MF, Albusoda A, Farmer AD, Aziz Q. The anatomical basis for
6. Tracey KJ. The inflammatory reflex. Nature 2002;420:853–9. transcutaneous auricular vagus nerve stimulation. J Anat 2019.
7. Rosas-Ballina M, Ochani M, Parrish WR, Ochani K, Harris YT, https://doi.org/10.1111/joa.13122.

Huston JM, et al. Splenic nerve is required for cholinergic 19. Bretherton B, Atkinson L, Murray A, Clancy J, Deuchars S,
antiinflammatory pathway control of TNF in endotoxemia. Proc Natl Deuchars J. Effects of transcutaneous vagus nerve stimulation in
Acad Sci U S A 2008;105:11008–13. individuals aged 55 years or above: potential benefits of daily
stimulation. Aging 2019;11:4836–57.

20. Kampusch S, Kaniusas E, Szeles JC. Modulation of muscle tone
and sympathovagal balance in cervical dystonia using
percutaneous stimulation of the auricular vagus nerve. Artif
Organs 2015;39:E202–212.

21. Clancy JA, Mary DA, Witte KK, Greenwood JP, Deuchars SA,
Deuchars J. Non-invasive vagus nerve stimulation in healthy
humans reduces sympathetic nerve activity. Brain Stimul 2014;7:
871–7.

22. Antonino D, Teixeira AL, Maia-Lopes PM, Souza MC, Sabino-
Carvalho JL, Murray AR, et al. Non-invasive vagus nerve
stimulation acutely improves spontaneous cardiac baroreflex

414 Kania et al.: Antiinflammatory reflex activation by OA-D and taVNS

sensitivity in healthy young men: a randomized placebo- 43. Kaczyński T, Wroński J, Głuszko P, Kryczka T, Miskiewicz A,
controlled trial. Brain Stimul 2017;10:875–81. Górski B, et al. Salivary interleukin 6, interleukin 8, interleukin
23. Liu CH, Yang MH, Zhang GZ, Wang XX, Li B, Li M, et al. Neural 17A, and tumour necrosis factor α levels in patients with
networks and the anti-inflammatory effect of transcutaneous periodontitis and rheumatoid arthritis. Cent Eur J Immunol 2019;
auricular vagus nerve stimulation in depression. J 44:269–76.
Neuroinflammation 2020;17:54.
24. Hong GS, Zillekens A, Schneiker B, Pantelis D, de Jonge WJ, 44. Riis JL, Ahmadi H, Hamilton KR, Hand T, Granger DA. Best practice
Schaefer N, et al. Non-invasive transcutaneous auricular vagus recommendations for the measurement and interpretation of
nerve stimulation prevents postoperative ileus and endotoxemia salivary proinflammatory cytokines in biobehavioral research.
in mice. Neuro Gastroenterol Motil 2019;31:e13501. Brain Behav Immun 2020. https://doi.org/10.1016/j.bbi.2020.
25. Jiang Y, Cao Z, Ma H, Wang G, Wang X, Wang Z, et al. Auricular 09.009.
vagus nerve stimulation exerts antiinflammatory effects and
immune regulatory function in a 6-OHDA model of Parkinson’s 45. Slavish DC, Graham-Engeland JE, Smyth JM, Engeland CG.
disease. Neurochem Res 2018;43:2155–64. Salivary markers of inflammation in response to acute stress.
26. Cicchitti L, Martelli M, Cerritelli F. Chronic inflammatory disease Brain Behav Immun 2015;44:253–69.
and osteopathy: a systematic review. PLoS One 2015;10:
e0121327. 46. Maydych V, Claus M, Watzl C, Kleinsorge T. Attention to emotional
27. Yorgun H, Canpolat U, Aytemir K, Ates¸ AH, Kaya EB, Akdog˘an A, information is associated with cytokine responses to
et al. Evaluation of cardiac autonomic functions in patients with psychological stress. Front Neurosci 2018;12:687.
systemic lupus erythematosus. Lupus 2012;21:373–9.
28. Koopman FA, van Maanen MA, Vervoordeldonk MJ, Tak PP. 47. Koenig J, Jarczok MN, Warth M, Ellis RJ, Bach C, Hillecke TK, et al.
Balancing the autonomic nervous system to reduce inflammation Body mass index is related to autonomic nervous system activity
in rheumatoid arthritis. J Intern Med 2017;282:64–75. as measured by heart rate variability – a replication using short
29. DeStefano LA. Greenman’s principles of manual medicine. term measurements. J Nutr Health Aging 2014;18:300–2.
Baltimore, MD: Lippincott Williams & Wilkins; 2011.
30. He W, Wang X, Shi H, Shang H, Li L, Jing X, et al. Auricular 48. Mehta RK. Impacts of obesity and stress on neuromuscular
acupuncture and vagal regulation. Evid Based Complement fatigue development and associated heart rate variability. Int J
Alternat Med 2012;2012:786839. Obes2015;39:208–13.
31. Badran BW, Mithoefer OJ, Summer CE, LaBate NT, Glusman CE,
Badran AW, et al. Short trains of transcutaneous auricular vagus 49. Bethea JR, Gillespie GY, Benveniste EN. Interleukin-1 beta
nerve stimulation (taVNS) have parameter-specific effects on induction of TNF-alpha gene expression: involvement of protein
heart rate. Brain Stimul 2018;11:699–708. kinase C. J Cell Physiol 1992;152:264–73.
32. Bertinieri G, Di Rienzo M, Cavallazzi A, Ferrari AU, Pedotti A,
Mancia G. A new approach to analysis of the arterial baroreflex. J 50. Loram LC, Themistocleous AC, Fick LG, Kamerman PR. The time
Hypertens 1985;3(Suppl 3):S79–81. course of inflammatory cytokine secretion in a rat model of
33. Stauss HM. HemoLab software. http://www.haraldstauss.com/ postoperative pain does not coincide with the onset of
HaraldStaussScientific/hemolab. mechanical hyperalgesia. Can J Physiol Pharmacol 2007;85:
34. Stauss HM, Moffitt JA, Chapleau MW, Abboud FM, Johnson AK. 613–20.
Baroreceptor reflex sensitivity estimated by the sequence
technique is reliable in rats. Am J Physiol Heart Circ Physiol 2006; 51. Azizieh FY, Al Jarallah K, Shehab D, Gupta R, Dingle K, Raghupathy
291:H482–3. R. Patterns of circulatory and peripheral blood mononuclear
35. Smyth HS, Sleight P, Pickering GW. Reflex regulation of arterial cytokines in rheumatoid arthritis. Rheumatol Int 2017;37:
pressure during sleep in man. A quantitative method of assessing 1727–34.
barorelex sensitivity. Circ Res 1969;24:109–21.
36. Julien C, Malpas SC, Stauss HM. Sympathetic modulation of 52. Koopman FA, Chavan SS, Miljko S, Grazio S, Sokolovic S,
blood pressure variability. J Hypertens 2001;19:1707–12. Schuurman PR, et al. Vagus nerve stimulation inhibits
37. Stauss HM. Identification of blood pressure control mechanisms cytokine production and attenuates disease severity in
by power spectral analysis. Clin Exp Pharmacol Physiol 2007;34: rheumatoid arthritis. Proc Natl Acad Sci U S A 2016;113:
362–8. 8284–9.
38. Chambers JM. Software for data analysis: programming with R
(statistics and computing). New York: Springer; 2008. 53. Guimarães PM, Scavuzzi BM, Stadtlober NP, Franchi Santos L,
39. Cohen J. A power primer. Psychol Bull 1992;112:155–9. Lozovoy MAB, Iriyoda TMV, et al. Cytokines in systemic lupus
40. Stauss HM, Kregel KC. Frequency response characteristic of erythematosus: far beyond Th1/Th2 dualism lupus: cytokine
sympathetic-mediated vasomotor waves in conscious rats. Am J profiles. Immunol Cell Biol 2017;95:824–31.
Physiol 1996;271:H1416–22.
41. Stauss HM, Anderson EA, Haynes WG, Kregel KC. Frequency 54. La Fratta I, Tatangelo R, Campagna G, Rizzuto A, Franceschelli
response characteristics of sympathetically-mediated vasomotor S, Ferrone A, et al. The plasmatic and salivary levels of IL-1β,
waves in humans. Am J Physiol 1998;274:H1277–83. IL-18 and IL-6 are associated to emotional difference during
42. Proctor GB, Carpenter GH. Regulation of salivary gland function stress in young male. Sci Rep 2018;8:3031.
by autonomic nerves. Auton Neurosci 2007;133:3–18.
55. Pîrsean C, Negut C, Stefan-van Staden RI, Dinu-Pirvu CE, Armean
P, Udeanu DI. The salivary levels of leptin and interleukin-6 as
potential inflammatory markers in children obesity. PLoS One
2019;14:e0210288.

56. Barrett CJ, Malpas SC. Problems, possibilities, and pitfalls in
studying the arterial baroreflexes’ influence over long-term
control of blood pressure. Am J Physiol Regul Integr Comp Physiol
2005;288:R837–45.

57. Floras JS, Hassan MO, Jones JV, Osikowska BA, Sever PS, Sleight P.
Consequences of impaired arterial baroreflexes in essential
hypertension: effects on pressor responses, plasma noradrenaline
and blood pressure variability. J Hypertens 1988;6:525–35.

Kania et al.: Antiinflammatory reflex activation by OA-D and taVNS 415

58. Head GA. Cardiac baroreflexes and hypertension. Clin Exp 61. Lohmeier TE, Iliescu R. Chronic lowering of blood pressure by
Pharmacol Physiol 1994;21:791–802. carotid baroreflex activation: mechanisms and potential for
hypertension therapy. Hypertension 2011;57:880–6.
59. Head GA. Baroreflexes and cardiovascular regulation in
hypertension. J Cardiovasc Pharmacol 1995;26(Suppl 2):S7–16. 62. Bagan L, Sáez GT, Tormos MC, Labaig-Rueda C, Murillo-Cortes J,
Bagan JV. Salivary and serum interleukin-6 levels in
60. Jordan J, Heusser K, Brinkmann J, Tank J. Electrical carotid sinus proliferative verrucous leukoplakia. Clin Oral Invest 2016;20:
stimulation in treatment resistant arterial hypertension. Auton 737–43.
Neurosci 2012;172:31–6.



J Osteopath Med 2021; 121(4): 417–428

Neuromusculoskeletal Medicine (OMT) Review Article

Yasir Rehman*, MD, MSc, PhD (C), Hannah Ferguson, M.OMSc, BSc, Adelina Bozek, MD,
Joshua Blair, M.OMSc, BSc, Ashley Allison, BA and Robert Johnston, M.OMSc

Dropout associated with osteopathic manual
treatment for chronic noncancerous pain in
randomized controlled trials

https://doi.org/10.1515/jom-2020-0240 eligible for inclusion. In this sub-study of a previous,
Received January 30, 2020; accepted November 18, 2020; larger systematic review, 11 studies (n=1,015) reported
published online March 16, 2021 data that allowed the authors to perform meta-analyses
on ACD and dropouts due to AE. The risk of bias (ROB)
Abstract was assessed with the Cochrane ROB tool and the quality
of evidence was determined with the Grading of Recom-
Context: Reviews exploring harm outcomes such as mendations Assessment, Development, and Evaluation
adverse effects (AE), all cause dropouts (ACD), dropouts (GRADE) approach.
due to inefficacy, and dropouts due to AE associated with Results: The pooled analysis showed that ACD was not
osteopathic manipulative treatment (OMT) or osteopathic significantly different for visceral OMTh (vOMTh) vs.
manual therapy (OMTh) are scant. OMTh control (odds ratio [OR]=2.66 [95% confidence
Objectives: To explore the overall AE, ACD, dropouts due interval [[CI]], 0.28, 24.93]) or for OMTh vs. standard care
to inefficacy, and AE in chronic noncancerous pain (CNCP) (OR=1.26 [95% CI, 0.84, 1.89]; I2=0%). Single study
patients receiving OMTh through a systematic review of analysis showed that OMTh results were nonsignificant
previous literature. in comparison with chemonucleolysis, gabapentin, and
Methods: For this systematic review and meta-analysis, the exercise. OMTh in combination with gabapentin (vs.
authors searched MEDLINE, Embase, Cochrane Central gabapentin alone) and OMTh in combination with
Register of Controlled Trials (CENTRAL), Physiotherapy exercise (vs. exercise alone) showed nonsignificant ACD.
Evidence Database (PEDro), EMCare, and Allied and Com- Dropouts due to AE were not significantly different, but
plementary Medicine Database (AMED), and Ostmed.Dr, the results could not be pooled due to an insufficient
as well as the bibliographical references of previous number of studies.
systematic reviews evaluating OMTh for pain severity, Conclusions: Most articles did not explicitly report AEs,
disability, quality of life, and return to work outcomes. ACD rates, or dropouts due to AEs and inefficacy. The
Randomized controlled trials with CNCP patients 18 years or limited data available on dropouts showed that OMTh
older with OMTh as an active or combination intervention was well tolerated compared with control interventions,
and the presence of a control or combination group were and that the ACD and dropouts due to AEs were not
significantly different than comparators. Future trials
*Corresponding author: Yasir Rehman, MD, MSc, PhD (C), Department should focus on explicit reporting of dropouts along with
of Health Research Methodology, The Michael G. DeGroote Institute beneficial outcomes to provide a better understanding of
for Pain Research and Care, Hamilton, ON, Canada; McMaster OMTh efficacy.
University, Hamilton, ON, Canada; and Department of Medical
Sciences at Canadian Academy of Osteopathy, 66 Ottawa Street Keywords: adverse events; chronic noncancerous pain
North, L8H 3Z1, Hamilton, ON, Canada, (CNCP); dropouts; osteopathic manipulative therapy
E-mail: [email protected]. (OMTh); osteopathic manipulative treatment (OMT);
https://orcid.org/0000-0002-2817-3638 tolerability.
Hannah Ferguson, M.OMSc, BSc, Adelina Bozek, MD, Joshua Blair,
M.OMSc, BSc, Ashley Allison, BA and Robert Johnston, M.OMSc, and Chronic noncancerous pain (CNCP) is among the most
Department of Medical Sciences at Canadian Academy of Osteopathy, common reasons for patients to consult general practi-
Hamilton, ON, Canada tioners and specialist pain clinics, with back pain being the
most common reason for osteopathic consultations [1–4].

Open Access. © 2020 Yasir Rehman et al., published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0
International License.

418 Rehman et al.: Dropout after OMT for chronic pain in randomized trials

Management of CNCP is multidisciplinary and, depending AE. The goal of this review was to explore the tolerability,
on the underlying cause, treatments for CNCP range from all cause dropout (ACD), and dropouts due to AE from OMTh.
pharmacological interventions and psychotherapy to
physical treatments such as physiotherapy, chiropractic Methods
treatments, massage therapy, and osteopathic manipula-
tive treatment (OMT) or osteopathic manual therapy To describe harm outcomes such as ACD rates and dropouts due to AE,
(OMTh) [5, 6]. Although the same principles are followed in we reviewed the data from individual studies that were previously
either manual treatment approach, OMT is performed reported in a separate systematic review that explored the effective-
by physicians, whereas OMTh is performed by nonphysi- ness of OMTh in CNCP [22]. The original study was registered with
cians [7]. Since OMT and OMTh are defined differently Prospero (#CRD42019125659).
based on the specific licensure of the practitioner, we will
be referring to both as OMTh throughout the rest of the We developed keywords using the MeSH word analyzer (http://
article. OMTh is reportedly a safe and effective treatment MeSH.med.yale.edu/) and developed a broad search strategy that was
approach and is employed as a primary or adjunctive implemented from database inception to July 2019. We utilized Ovid to
treatment to manage CNCP [6–9]. OMTh requires a search MEDLINE, Embase, EMCare, Allied and Complementary Medi-
comprehensive understanding of anatomy and physiology cine (AMED), Physiotherapy Evidence Database (PEDRO), and
based on osteopathic principles, and practitioners use that Cochrane Central Register of Controlled Trials (CENTRAL) databases, as
knowledge to mobilize and influence the patient’s body. well as Ostmed.Dr, for eligible studies. The bibliographical references of
The risk of complications and adverse effects with OMTh is previous systematic reviews were also searched for eligible trials. As our
lesser than with other manipulation techniques [10]; the goal was to report the AE and dropouts associated with OMTh in rela-
scope and acceptance of OMTh continues to broaden as tion to studies included in our previous review [22], we did not update
considerable advances continue to be made by the osteo- our search strategy. The search strategy is provided in the Supplemental
pathic profession in both research and the politics of Material, which was also published with our previous review [22].
healthcare [11–14]. Considering both the benefits and
potential risks of any treatment is vital for both patients Our eligibility criteria included RCT enrolling patients 18 years or
and healthcare providers to establish realistic expectations older with CNCP that employed OMTh as an active or combination
and to make informed decisions [15, 16]. intervention and involved comparison with any other intervention or
control. Eligible trials explored the effectiveness of OMTh on pain
Most systematic reviews focus on beneficial outcomes severity, disability, QOL, or RTW in CNCP. The methodology, inclusion
for intervention; however, less than 10% of systematic criteria, and exclusion criteria are reported in the previously published
reviews and meta-analyses explore harm outcomes such as study [22] and we conducted our meta-analysis according to Preferred
tolerability and adverse effects (AE) [17, 18]. One potential Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)
reason that authors of review articles do not focus on harm guidelines [27]. For this review, we included the studies among those
outcomes is inconsistent reporting of harm outcomes and reviewed that reported tolerability outcomes such as AE, as well as
lack of standardized reporting methods [19]. A meta- ACD due to AE and inefficacy.
analysis as part of a systematic review aims to provide a
thorough, comprehensive, and unbiased statistical sum- We excluded RCT that enrolled fewer than 10 participants per arm
mary of data from the literature [20, 21]. We recently pub- at baseline or if the author(s) reported composite score. Studies with a
lished another systematic review and meta-analysis focus on cancerous pain, pain developed during pregnancy, head-
exploring the effectiveness of OMTh in CNCP patients [22]. aches, pain due to gynecological abnormalities, irritable bowel syn-
During that review, it became apparent that there is a need drome (IBS), or other visceral pain such as prostatitis were also
to report transparent and critically appraised results from excluded. We excluded crossover trials, as there is a possibility of
existing literature about the harm effects of OMTh in the carry-over effect and methodological challenges limits their applica-
management of CNCP. bility [20, 28]. Given that OMTh is governed by a set of principles and
uses a combination of methods to treat the body [29], we excluded
The existing studies supporting OMTh safety and effi- studies that specifically investigated a single technique without
cacy suffer from limitations that may contribute to OMTh employing additional use of general OMTh.
being misunderstood and underutilized [23–25]. Previous
studies [26] have explored the tolerability of OMTh by CNCP Data collection and analysis plan
in cross-sectional surveys or prospective observational
studies; however, since OMTh is a therapeutic hands-on Title, abstract, full text screening, data abstraction, and risk of bias
intervention, randomized controlled trials (RCT) are consid- (ROB) were compiled in duplicates and independently by a team of
ered a more suitable study design to assess dropout rates and reviewers (H.F., J.B., A.B., A.A.). Included studies and results were
compared between independent reviewers and teams, and any dis-
crepancies were reviewed with adjudication with third reviewer (Y.R.).
We extracted information about the number of events associated with
ACD and dropouts due to AE according to the definition previously
reported [30–32]. In previous reviews, all comparators in varying
osteopathic studies would be pooled together resulting in high het-
erogeneity and, therefore, less reliable results. We employed the same

Rehman et al.: Dropout after OMT for chronic pain in randomized trials 419

Figure 1: PRISMA flowchart of the included studies.

strategy from the primary study [22] to pool according to the homo- Indirectness, inconsistency: The clinical characteristics of the par-
geneity between OMTh and the comparator type(s). A previous sys- ticipants and width of the 95% CI, respectively, were assessed for
tematic review with meta-analysis [33] reported no significant indirectness and inconsistency.
difference between exercise and physiotherapy; therefore, in our
meta-analysis, we merged physiotherapy and group exercise com- Analysis: When possible, we performed meta-analysis in random ef-
parators. Albers et al. [34] and Licciardone et al. [35] were three-arm fect model (REM) and reported results with odds ratio (OR) with 95%
studies and, based on the similarities between the OMTh applied and CI. The statistical analysis was performed with Review Manager 5.3
comparators, respectively, we merged those two arms. (Review Manager RevMan; computer program, version 5.3. Copenha-
gen: The Nordic Cochrane Centre, the Cochrane Collaboration, 2014).
Quality of evidence

The quality of evidence was assessed using the Grading of Recom- Results
mendations Assessment, Development, and Evaluation (GRADE) tool
[36–39]. GRADE consists of five components: ROB, inconsistency, Our search from all databases collectively yielded 2,956
imprecision, publication bias, and indirectness. studies by title and abstract screening (Figure 1) [22]. Of the
2,956 titles and abstracts, we included 16 studies in our
Risk of bias: ROB was analyzed using modified Cochrane ROB tools on initial review [22], but only 11 studies [34, 35, 41–49]
the following components: random sequence generation, allocation reported AE and dropout rates in patients (n=1,015).
concealment, blinding of participants, health care providers, outcome Descriptions of the study characteristics and reported
assessors, and dropout rates [40]. dropouts are given in Table 1 and Table 2.

Heterogeneity (Inconsistency): The heterogeneity of the pooled There was variety in the type of pain between the
studies was determined by visual inspection of forest plots and using included studies: five included patients with low back pain
the I2 statistic [20]. For heterogeneity, we used a cutoff of 60% or [35, 41, 43, 44, 46], while one study was specific to those with
inconsistency in the effect estimates of individual studies on visual sciatica related to lumbar disc herniation (LDH) [41]. Two
inspection of the forest plot [35]. studies included patients with fibromyalgia [34, 45], one
included patients with osteoarthritis [42], one with unspeci-
Imprecision: Imprecision was determined with 95% confidence in- fied musculoskeletal pain [47], one study included patients
terval (CI). with chronic shoulder pain [44], and one study with chronic
neck pain [48]. The follow up duration ranged from 42 to
Publication bias: We did not have 10 or more studies in the pooled 365 days. Only three studies explicitly reported [35, 42, 44]
analysis and therefore could not assess publication bias. that cointerventions were allowed during the trial period.

Table : Characteristics of the included studies. 420

Study Mean age, Female Pain region Were patients with Was disability/litigation Were other Duration of trial Number of Number of
comorbid psychological considered part of medications
years (SD) patients, % and medical conditions exclusion criteria? allowed? or follow up participants in participants in
included?
period, days intervention control or Rehman et al.: Dropout after OMT for chronic pain in randomized trials

group comparison groups

Albers  [] . (.) . Fibromyalgia No NR NR   

Altinbilek  [] . (.) . Knee OA Yes NR No; except for   

paracetamol

Burton  [] . (.) . LBP/LDH Yes Yes NR   

Chown  [] . . Chronic LBP Yes NR Steroids were   

Knebl  [] -* . Chronic Yes NR not allowed  +
NR

shoulder pain

Licciardone  [] . (.) . Chronic Yes Yes Yes   

nonspecific

LBP

Licciardone  [] . (.) . Chronic Yes Yes Yes   

nonspecific

LBP

Marske  [] . . Fibromyalgia No NR NR   

Marti Salvador  [] -* NR Chronic No NR No   

nonspecific

LBP

Papa  []  (.) . Chronic Yes NR NR   

nonspecific

body pain

Schwerla  [] .–.* . Chronic No Yes NR   

nonspecific

neck pain

OA, Osteoarthritis; LBP, low back pain; LDH, lumbar disc herniation; NR, not reported. *Studies did not report a mean age, only provided a range with no SD. +Knebl [] only reported the total
number of patients involved but did not distinguish between intervention or control groups.

Table : Summary of the OMTh and the comparator interventions.

Study Treatment groups Osteopathic manipulation Comparator Per protocol or Cointerventions use allowed
Albers  [] OMTh clinician-directed

Description

Direct and indirect techniques (muscle SC Clinician directed Pharmaceutical use NR but patients were
excluded if they underwent manual
energy, MFR, HVLA, functional techniques, therapy treatment or alternative treatment
procedures during the study period
and balanced ligamentous tension) and No – Prevented from taking non-steroidal
antiinflammatory drugs  week before and
cranial/sacral techniques during the study. Paracetamol up to  g/
daily was allowed and drugs for systemic
Altinbilek  [] OMTh + Exercise Standardized mobilization and compres- Exercise consisted of quadriceps Per protocol conditions continued

sion for bilateral patellofemoral and muscle strengthening exercise, No – patients with previous manipulations
on the same area were excluded
tibiofemoral joints followed by a lower leg lifting, and muscle stretching
No – patients with manipulations such as
extremity pumping technique. These such as iliotibial band, hamstring physiotherapy, acupuncture in the previous
 months were excluded as well as users of
techniques are taught to the patient to stretching, strengthening steroids and anticoagulants

apply at home as well abductor and adductor muscles NR

of the hip Yes – Usual or other low back care allowed
in both arms except OMTh or chiropractic
Burton  [] OMTh Soft tissue stretching in combination with Chemonucleolysis under Clinician directed manipulation

low amplitude passive articulatory ma- general anesthesia, a single Rehman et al.: Dropout after OMT for chronic pain in randomized trials

neuvers and high velocity thrust to the dose of chymopapain

lumbar spine and buttock musculature

Chown  [] OMTh Soft tissue massage, inhibition, muscle – Group exercise: Problem Clinician directed

stretching, muscle energy, high velocity identification, anatomy ed-

thrust (varying), articulation, mobilization, ucation, home stretching

exercise advice, discussion of psychoso- exercise programme, basic

cial issues, education, nutritional/dietary postural setting use of

advice transverses, multifidus

– Manipulative physiotherapy:

Education/advice, joint

mobilization, soft tissue

mobilization, global exercise

for mobility, electrotherapy,

postural correction

Knebl  [] OMTh Treatment was defined as administration The positions of the Spencer Per protocol

of the seven-step Spencer technique technique without administration

which is an articular treatment involving of the actual corrective forces

compression, traction, and muscle energy (isometric muscle contraction)

in various planes and axes

Licciardone  [] OMTh Combination of MFR, strain–counter ROM activities, light touch, and Clinician directed

strain, muscle energy, soft tissue, high- simulated OMTh techniques

velocity–low-amplitude thrusts, and cra-

nial–sacral

421

Table : (continued) 422

Study Treatment groups Osteopathic manipulation Comparator Per protocol or Cointerventions use allowed
clinician-directed
Rehman et al.: Dropout after OMT for chronic pain in randomized trials
Description

Licciardone  [] OMTh The lumbosacral, iliac, and pubic regions Sham OMTh involved hand con- Clinician directed Yes – patients could self-initiate LB
cotreatments including prescription and
were targeted using high-velocity, tact, active and passive range of non-prescription drugs, exercise programs,
lumbar supports, complementary, alterna-
low-amplitude thrusts; moderate velocity, motion, and techniques that tive medicine, and physical therapies

moderate-amplitude thrusts; soft tissue simulated OMTh but used such Concurrent medications were continued

stretching, kneading, and pressure; MFR; maneuvers as light touch, Yes – Cointerventions were not analyzed, a
noted limitation of the study
positional treatment of myofascial tender improper patient positioning,
Yes – patients were subjected to the usual
points; and isometric muscle activation purposefully misdirected move- therapy established as a result of medical
specialist visits but were not reported
ments, and diminished physi-
Yes – Excluded patients with concomitant
cian force physical therapy, corticosteroid use,
anticoagulants. Patients kept a diary of
Marske  [] OMTh & Treatment modalities included MFR, Gabapentin with variable dose Clinician directed analgesic and muscle relaxant use
OMTh + Gabapentin
muscle energy, counter strain, facilitated ( – mg/day)

positional release, articular ligamentous,

high velocity/low amplitude, and cranial

sacral OMTh. Indirect moving to direct as

tolerated by the patient

Marti Salvador  [] Diaphragm OMTh Lumbar MFR, normalization of the ilio- Manual contact was applied but Per protocol

lumbar ligament, pumping, traction to the with no therapeutic intention

lumbar and sacral regions, techniques

directed to the diaphragm included

pumping, inhibition, muscle stretching,

and a global abdominal hemodynamic

maneuver

Papa  [] OMTh Different techniques were used based on Postural examination and Clinician directed

the results of the exam, objective tech- palpation of non-specific

niques were performed on any body part different parts of the body in

that the osteopath found to be correlated different positions supine

with the disorder and the patient’s func-

tional limitation with considerations to

myofascial, visceral, articulating and head

structures

Schwerla  [] OMTh Osteopathic techniques included direct Inert therapeutic ultrasound Clinician directed

techniques (high velocity, muscle energy,

MFR), indirect techniques (functional

techniques, balanced ligamentous ten-

sion), visceral and/or cranial techniques

MT, manual therapy; MFR, myofascial release; HVLA, high velocity low amplitude; OMTh, osteopathic manual therapy; CCT, conventional conservative therapy; ROM, range of motion; SC, standard
care; vOMTh, visceral osteopathic manual treatment; SE, specific exercise; NR, not reported.

Rehman et al.: Dropout after OMT for chronic pain in randomized trials 423

Table : Risk of bias in the included studies.

Study Random sequence Allocation Blinding of the Blinding of the Blinding of the Drop

generation concealment participants health care provider outcome assessors out >%`

Albers  [] Low risk Low risk High risk High risk High risk Low risk
Altinbilek  [] Low risk Low risk High risk High risk Low risk Low risk
Burton  [] Low risk Low risk High risk High risk High risk High risk
Chown  [] Low risk Low risk High risk High risk Low risk High risk
Knebl  [] Low risk Low risk Low risk High risk Low risk High risk
Licciardone  [] Low risk Low risk High risk High risk Low risk High risk
Licciardone  [] Low risk Low risk High risk High risk Low risk High risk
Marske  [] Low risk Low risk High risk High risk Low risk Low risk
Marti Salvador  [] Low risk Low risk High risk High risk Low risk High risk
Papa  [] Low risk Low risk High risk High risk Low risk Low risk
Schwerla  [] Low risk Low risk High risk High risk High risk Low risk

Descriptions of the OMTh approach for each study are intervention [45]. The number of reported AE decreased
given in Table 2. Overall, methods of osteopathic treatment with time in all groups but there was a significant differ-
used were directed to superficial, intermediate, and deep ence in OMTh only (mean difference, −5.7; p<0.01) and
structures with a combination of direct or indirect methods combination groups (mean difference, −3.7; p=0.03) [43].
using a type of activating force (compression or traction) with
the patient active or passive (respiration, isometric contraction). Two patients (5.5%) from the control group in Schwerla
Comparators varied and included exercise [42, 43], pharmaco- et al. [48] dropped out due to aggravation in pain.
logical [45], standard care [35, 44], and general OMTh [46].
All cause dropout rate

Risk of bias Two hundred and 87 out of total 1,015 participants (28.28%)
had ACD. The overall ACD rate in OMTh vs. controls was
A summary of the ROB is given in Table 3. None of the 22.67 vs. 30.07%, respectively (n=107 and n=156, respec-
included studies met all criteria of the ROB. All studies were tively). Some reported reasons for dropouts included
high risk for blinding of participants, health care providers, patients who were unable to meet the study demands [34],
and high dropout rates. Except for three studies [34, 41, 48], patients who chose to cease treatment [42], patients had flare
all studies performed independent outcome assessments. up in previously diagnosed conditions [38], or patients who
became pregnant [46]. Unfortunately, in most studies, the
Adverse events and drop out due to adverse events reasons for ACD were not specified [35, 41, 43, 48, 49]. In
Papa et al. [47] and Knebl et al. [49], the patients’ reasons for
A summary of adverse events outcomes is reported in dropping out were not reported nor did the authors report
Table 4. Only one study [44] explicitly reported an AE asso- ACD rates for each arm; therefore, we did not include Papa
ciated with OMTh that led to dropout. In Licciardone et al. et al. [47] and Knebl et al. [49] in the pooled analysis. One
[44], one patient reported recurring increase in back spas- study [38] reported high ACD in both arms, OMTh vs. exercise
ticity after treatments, which was attributed to OMTh. Overall at 50.64 and 63.12% respectively. The authors noted major
in Licciardone et al. [44], 27 (6%) patients experienced AE; 16 problems with recruitment and retention of patients, finding
(6.95%) from OMTh and 11 (4.88%) from the control group. that patients were more likely to show up to one-on-one
Six patients from OMTh and three patients from the control sessions. They cited dual pressures of scheduling with both
group were considered to have serious AE; however, no the patients and the practitioners and a lack of administra-
events were explicitly related to the study intervention and tive time, which impacted retrieval of follow up statistics [38].
there were no significant differences between the main ef-
fects’ group in the frequency of AEs. ACDs: pooled analysis

In Marske et al. [45], one patient from each arm – OMTh OMTh vs. standard care
and OMTh combined with gabapentin – reported an AE;
however, the authors did not specify what the AE were. The Three studies [35, 44, 48] (n=587) reported ACD in compar-
AE were mild-to-moderate in severity and none required ison with 20.78% ACDs (122 of 587 participants; 22.59 vs.

424 Rehman et al.: Dropout after OMT for chronic pain in randomized trials

Table : Outcomes of the unpooled studies.

Study Interventions ACD, n(%) Dropout due Dropout Comments
to AE, n(%) due to
inefficacy

Albers  [] vOMTh (n=)  (.)  NR n= dropped out due to not able to meet with
*Altinbilek  [] study demand
*Burton  [] Control (n=)  NR NA
*Chown  [] OMTh + Exercise  (%) NR NR Patients in both arms dropped out on their
(n=) own will
Exercise (n=)  (%) NR NR ES=. [., .]
OMTh (n=)  (.%) NR NR No explanation about the dropout rate was
Chemonucleolysis  (.%) NR NR given
(n=) ES=. [., .]
OMTh (n=)  (.%) NR NR Difficulties with scheduling and administrative
Exercise (n=)  (.%) NR NR time to conduct follow-ups was reported. One
on one OMT sessions were more likely to be
Knebl  [] OMTh group + control  (.%) NR NR attended compared to group sessions
Licciardone  [] group (n=) ES=. [., .]
Licciardone  [] OMTh (n=)  (.%) NR NR One patient died and one patient refused. Not
*Marske  [] Control (n=)  (.%) NR NR clearly stated which arm patients were in
OMTh (n=)  (.)  (.%) NR ACD was .%. Further detail for dropout
Control (n=) NR rates was not given
OMTh (n=)  (.%)  NR n= developed back spasticity due to OMT
 (.)  (.%) No further detail given
Total two patients dropped from OMTh group,
Gabapentin (n=)  (.%)  (.%) NR n= developed AE due to OMTh, trans-
portation problems for nd loss
OMTh + Gabapentin  NR *ES (OMT vs. Gabapentin) =. [., .]
(n=) Notes treatment reaction, transportation and
 (.%)  NR pregnancy but did not note how many in each
Martí-Salvador  [] vOMTh (n=) category
 NR NA
Papa  [] Control (n=)  NR NR *ES (OMTh + Gabapentin vs. Gabapentin)
OMTh (n=) NR =. [., .]
Control (n=) NR Two patients in OMTh group got pregnant due
to which patients had to discontinue OMTh
Schwerla  [] OMTh (n=)  (.%)  NR NA
Control (n=)  (.%)  (.%) NR n= dropped out from the study. Author did
not specified number of dropouts for each
arm; n= dropped out due to underlying
medical conditions, n= dropped out due to
transportation problems
 dropped out for reason not given;  had
incomplete data
 dropped due to aggravation in pain in con-
trol group

ACD, all cause dropouts; AE, adverse events; ES, effect size; OMTh, osteopathic treatment; SC, standard care; vOMTh, visceral osteopathic
treatments; gOMTh, general osteopathic treatment. *Studies were not poolable/outcome was reported by single study. ES (odd ratio) with %

confidence interval was calculated to determine the significance level between OMT and control group.

18.88% for OMTh vs. control groups, respectively; Figure 2). Visceral OMTh vs. OMTh control

There is moderate quality evidence that ACD rates between Two studies [34, 46] (n=116) reported ACD in this compar-
ison with 2.5% ACDs (three of 116 participants; 4.34 vs. 0%
OMTh and standard care were not significant (OR=1.26 [CI,
0.84, 1.89]; I2=0%].

Rehman et al.: Dropout after OMT for chronic pain in randomized trials 425

Figure 2: All cause dropouts (comparison: osteopathic manipulative therapy vs. control).

Figure 3: All cause dropouts (comparison: general osteopathic treatments vs viscerall osteopathic treatment).

Table : GRADE quality of evidence.

Outcome Number of studies Risk of bias Inconsistency Imprecision Indirectness Publication bias Quality of evidence

All cause dropouts

OMTh vs. SC  (n=) High Low Low Not detected Not detected Moderate
High Low
vOMhT vs. gOMTh  (n=) Low Not detected Not detected Moderate

GRADE, Grading of Recommendations, Assessment, Development, and Evaluations; OMTh, osteopathic treatment; SC, standard care; vOMTh,
visceral osteopathic treatments; gOMTh, general osteopathic treatment.

for OMTh and control group, respectively; Figure 3). Mod- attempted to specifically explore the risks and reasons for
erate quality evidence showed no statistically significant dropouts from OMTh in comparison to other interventions
difference in ACD between visceral OMTh (vOMTh) and in CNCP RCT studies. Our results indicated that although
control OMTh (OR=2.66 [CI, 0.28, 24.93]; I2=0%]. ACD were higher in OMTh groups compared with control
groups, there was no significant difference noted. Aggra-
The GRADE assessment for the ACD and dropouts vation of pain was specified within the OMTh group in two
due to AE is reported in Table 5. The quality of evidence studies [44, 48] and one study [48] within the control
for pooled analyses was moderate due to a high ROB group. One study [45] noted a treatment reaction to the
result. control medication. Differences between OMTh and control
group dropouts due to AE were not significant. Unfortu-
Dropout rates due to inefficacy nately, only five studies [34, 44–46, 48] out of 11 reported
specific data on AE. From our main review, we found that
No study reported dropouts due to inefficacy. with important patient outcomes such as pain, disability,
QOL, and RTW, OMTh was well tolerated and was not
Discussion statistically different from the comparators. From our
initial review [22] including 16 studies, there were only 11
In this review, we critically appraised tolerability outcomes studies that included any data on ACD or AEs which could
reported as AE, ACD, dropouts due to AE, and inefficacy be included in this review. Few of those studies included
within RCTs, exploring the effectiveness of OMTh in CNCP details on why patients dropped out or provided insights
after we noticed a degree of oversight in the currently- on how to improve the problem in the future, and fewer
available literature and were unable to address this fully in reported complete data on AE (Table 4). Overall, the cur-
our previous systematic review and meta-analysis [22]. To rent literature reporting AE of OMTh is lacking and our
the best of our knowledge, this is the first review that has hope is that this review will encourage improved reporting