A brief whole-body vibration intervention to avoid weight gain in college students: A randomized controlled pilot trial

Abstract

Objective:

Over one-third of college students are overweight or obese and rates are rising. Whole body vibration (WBV) training could prevent weight gain but has not been tested in college students.

Methods:

Randomized controlled trial comparing thrice weekly WBV for 6 months to controls (CON) in undergraduate students. Feasibility included retention, adherence and safety and outcomes included changes in weight, body mass index (BMI) and fat mass.

Results:

77 students enrolled in the trial (WBV: n=40, CON: n=37), 81% completed the study. Adherence to WBV averaged 59%. Average group differences were 1% body fat (p=0.049) and 1kg fat mass (p<0.01), favoring WBV. Among students completing >80% of prescribed WBV sessions significant group differences widened, while group differences in BMI (p=0.026) and weight (p=0.02) change became significant.

Conclusions:

WBV may be a feasible, safe and effective approach to weight management in college students, though strategies to optimize adherence should continue.

Introduction

According to the most recent American College Health Association health survey, the self-reported rate of overweight and obesity among college students in the U.S. averages 37%,¹ a figure that has risen sharply over the last decade.^2,3 According to data from the 2015–2016 National Health and Nutrition Examination Survey, the rates of obesity escalate from 20.6% in adolescents (12–19 years) to 35.7% among adults 20–39 years of age,^4,5 suggesting that the transition from living at home to independent residency comes with unfavorable changes in diet and physical activity patterns. If the weight trends are not addressed during the transition years, they are likely to be carried through into middle and later adulthood and become precursors to obesity-related chronic illnesses, such as stroke, heart disease, diabetes, and cancer.^6–8 Even more troubling is the appearance of obesity-related chronic illnesses in young adults. Over the past few decades the rates of colon and rectal cancer, which are linked to obesity,⁹ rose 1%−2% and 3% per year among 20–39 year olds and 20–29 year olds, respectively.¹⁰ Not all college students are younger adults, but the transition from regular work/home schedules to a college environment and associated stressors with schooling can create barriers to managing healthy weight practices. Moreover, these students may already be overweight or obese when returning to school and thus the implications of continued or accelerated weight gain while pursuing undergraduate education could be higher than for younger adults.

A variety of strategies have previously been evaluated to prevent weight gain in college students, ranging from daily monitoring of body weight,¹¹ the use of modest dietary and physical activity strategies,^11,12 to educational/behavioral interventions.^13,14 Although healthy weight loss practices have been associated with lower rates of weight gain, the rate of weight change in the treatment groups did not differ significantly from controls.^11,13 Developing weight gain prevention programs that require minimal time, effort and cost and are easily integrated into the environment of college life might be more feasible and sustainable than commercial diet programs and heavy exercise or programs introduced to reverse weight gain. College may offer an ideal setting for introducing weight gain prevention programs because students may have increased access to resources, are in a homogeneous social environment and can be targeted in larger numbers, however, relatively few randomized trials have specifically focused on weight loss interventions in the college setting.¹⁵

Alternative approaches that can stimulate increases in energy expenditure in a brief time period could be attractive to college students. Whole body vibration (WBV), that requires a person to passively resist small but frequent vertical accelerations, is a novel training modality that has emerged as an alternative to traditional exercise.¹⁶ An increasing body of literature suggests that short, regular bouts of WBV alone, or in conjunction with low-level exercise are safe and may lower % body fat just as well conventional exercise training.^17–20 However, to our knowledge WBV has yet to be tested for feasibility and efficacy as a weight gain prevention strategy in college students. We conducted a pilot study to test WBV as an innovative weight management approach to improve body composition in college students with or at risk for overweight and obesity. Our study aims were to evaluate the feasibility (enrollment, retention, compliance and safety) and preliminary efficacy of a WBV training program on prevention of weight gain and unhealthy shifts in body composition (increased total and central fat mass) in undergraduate college students.

Methods

Study design

We conducted a 6-month randomized controlled trial comparing 2 parallel groups assigned to either whole-body vibration training (WBV) or a control group. The Oregon Health and Science University (OHSU) institutional review board approved the study and every participant consented to participate in the study prior to data collection. The trial is registered in ClinicalTrials.gov (NCT02134275).

Population and recruitment

The target population for this study was undergraduate college students who were overweight/obese or were at risk for becoming overweight (BMI>22 kg/m²). To power this study, we used a conservative effect size estimate of d=.40 when determining sample size. Using assumptions of traditional repeated measures ANOVA and assuming power=.80 and alpha=.05, we needed to randomize 37 persons to each of the two study groups to detect a moderate effect of WBV (d=.40) on our efficacy outcomes of weight and fat mass between groups over time^17–20. Accounting for an anticipated 20% attrition, we aimed to recruit 89 students into the study.

Inclusion criteria for the study were as follows: 1) Age 18 years or older, 2) self-reported BMI between 22–49.0 kg/m², 3) no self-reported weight change >5 lbs. in the previous month, 4) not currently on or planning to begin a prescribed weight loss program, 5) not sufficiently physically active (<150 minutes of self-reported moderate intensity physical activity per week), 6) plans to reside in the study area for 6 months, 7) currently enrolled in a 4-year undergraduate degree program, 8) not pregnant and no plans to become pregnant during the study period, 9) no self-reported health conditions that would prevent participation in moderate intensity physical activity per American College of Sports Medicine pre-participation guidelines, 10) not currently smoking or using tobacco products, 11) not self-reporting consumption of ≥3 alcoholic drinks per day, on average. Eligibility based on BMI was later confirmed in the laboratory during the baseline visit. Participants were recruited from the OHSU School of Nursing baccalaureate degree nursing program and from 4-year undergraduate degree programs at nearby universities. Recruitment strategies included classroom announcements, tabling at university events, posting flyers within the community, postings on websites (e.g., Craig’s list) and social media campaigns (e.g, Reddit, Twitter, Facebook). Recruitment occurred from July 2015 to November 2016 and final post-intervention testing was completed by July 2017.

Procedures and setting

Outcomes were measured at baseline, three, and six months. Testing took place at the OHSU School of Nursing from 2015–2017. Participants were provided a small monetary remuneration for each testing visit ($40) and participants randomized to WBV were incentivized to adhere to training, potentially earning up to $200 that was prorated based on attendance. Training occurred at both OHSU and at a local private university to improve accessibility for participants.

Participants were initially screened by phone and interested and eligible students were scheduled for a baseline testing appointment which consisted of anthropometric measurements and surveys completed via computer. For female participants, pregnancy was ruled out by administering a urine-based pregnancy test prior to body composition testing by dual energy x-ray absorptiometry (DXA). After baseline testing participants were randomized to WBV or control groups.

Randomization and masking

Using a computer-generated randomization sequence and a block size of 6, eligible participants were assigned in a 1:1 ratio to either WBV or control groups. The randomization scheme was created prior to enrolling the first participant and placed in concealed envelopes then assigned in order according to the scheduled baseline visit. After baseline testing participants opened the sealed envelope to reveal their assignment, thus participants were not blinded to their assigned group nor to the purpose of the study. The primary outcome, body composition, was obtained via DXA and scans were analyzed by staff who were blinded to study group.

Study arms

Whole Body Vibration Training (WBV).

The training intervention included 6-months of WBV training, three times per week (Power Plate pro5 AIRdative™, Power Plate International Ltd, Irvine, CA). The WBV protocol was based on other published reports where WBV improved body composition and physical function,^17,21 in which frequency progressed from 30Hz to 50Hz and amplitude from 1mm to 4mm across the course of the intervention (Table 1). Each WBV session consisted of 20 minutes of stand time on the vibration platform, with the legs in slight flexion at the knees, while the platform oscillated. The sessions were monitored by a research assistant and perceived pain/discomfort on a 1–10 Likert scale (none to worst pain) was recorded before, during, and immediately after each training session. If participants reported a high level of pain/discomfort (8–10) for 3 or more WBV sessions, the duration of training was reduced by 25% for the subsequent visit(s) until pain/discomfort subsided below 8 for >3 sessions in a row.

Table 1.

Training protocol for WBV sessions over 6-month training program

Week	Amplitude*	Frequency	# of bouts	Duration/bout	Rest interval
Week 1	2mm	30 Hz	2–3	5 min	1 min
Week 2	2mm	30 Hz	3–4	5 min	1 min
Week 3	4mm	30 Hz	4	5 min	1 min
Week 4	4mm	35 Hz	4	5 min	1 min
Week 5	4mm	40 Hz	4	5 min	1 min
Week 6	4mm	45 Hz	4	5 min	1 min
Week 7–24	4mm	50 Hz	4	5 min	1 min

Control.

Participants randomized to the control group were asked to not change their physical activity and dietary habits across the intervention period.

Measures

Demographics.

To describe the sample we obtained age, race, ethnicity, presence of chronic conditions, current medications, and health habits (e.g., smoking status) at baseline by questionnaire.

Pre-specified feasibility outcome.

Feasibility was based on the ability to accrue the target sample, retention of participants throughout the trial, adherence to prescribed WBV sessions, and tolerability of any pain/discomfort to WBV training. We hypothesized that we could fully accrue the target sample, retain 80% of participants, achieve 80% adherence to WBV sessions and report low (<2) average pain scores post-WBV sessions.

Pre-specified efficacy outcome.

Our study was powered on primary outcomes of weight and fat mass determined by DXA (Hologic Discovery Wi; APEX software v4.0). From the DXA scan we could also explore additional measures of bone-free lean mass; visceral, android, and gynoid fat mass (kg); and percent body fat. Coefficients of variation in our laboratory are <1.5% for soft tissue.²² Two clinical measures of adiposity were also measured by waist circumference (WC) and body mass index (BMI; kg/m²).

Potential confounders: Energy balance.

We hypothesized that WBV would improve weight and body composition independent of energy intake and additional physical activity energy expenditure, thus we assessed these variables before and after the intervention period in both study arms. An online version of the 2014 Block Food Frequency Questionnaire (FFQ) was used to assess habitual dietary intake and physical activity. The Block FFQ is a validated dietary assessment measure and also includes questions about physical activity habits in the past month.²³ The food and beverage list includes 127 items, plus additional questions to adjust for fat, protein, carbohydrate, sugar, and whole grain content. The activity screener includes 11 items that represent the most important sources of energy expenditure in the US. The variables of interest were total daily energy intake and energy expended in moderate-vigorous intensity physical activity, both in kcals per day.

Statistical analysis

Analyses.

Standard t-test and chi-squared tests were used to compare demographic, body composition, and energy balance variables between study arms at baseline. The primary analyses were conducted using a linear mixed effects modeling approach implemented in the nlme package for the R statistical computing environment.^24,25 The base model included fixed effects for group (WBV vs. control), time (baseline, 3 months, 6 months), and the group x time interaction to test whether the change in outcomes across time differed between groups. All analyses were performed as intent-to-treat (ITT). We also assessed intervention effects for those with higher adherence to WBV training, defined > 50% of WBV sessions over 6 months, compared to controls. Alpha was set at p = 0.05 for all analyses.

Results

Students in the trial were, on average, slightly older than a traditional undergraduate population (x̄ = 27.1 ± 8.2 years) due to the inclusion of older students enrolled in undergraduate nursing programs. The sample was overweight based on average BMI (x̄ = 28.3 ± 5.7 kg/m²) and had an average waist circumference in the unhealthy range based on gender (x̄ = 97.8 + 13.1 cm and 101.0 + 16.5 cm, for women and men, respectively). Students reported expending less than 200 kcal in moderate-vigorous intensity physical activities per day. There were no significant group differences in self-reported energy intake or physical activity energy expenditure at baseline nor over time. No participants reported taking up smoking during the study period nor did self-report alcohol intake change significantly between groups over time. Results were unchanged when considering high adherers to WBV only.

In response to recruitment efforts, 435 persons inquired about the study but 43% were ineligible and 31% did not respond to follow-up calls by study staff (Fig 1). Of those students contacted and eligible, few refused to participate (n=32) and 80 (18% of all inquiries) were scheduled for baseline testing. At baseline testing, three persons failed further screening resulting in 77 persons randomized to WBV (n=37) or control (n=40). Over the course of the 6-month intervention period 11 students dropped out of the trial (WBV: n=6; Control: n=5), resulting in a retention rate of 81%. The majority of dropouts (91%) participated in training at the OHSU site. Reasons cited for dropout were lack of time (n=6), disinterest (n=3), or moving out of the study area (n=2).

Figure 1.

CONSORT diagram showing participant flow throughout the trial

Among the 29 WBV participants who completed baseline and 6-month visits, adherence to thrice weekly training sessions averaged 59% or 1.7 sessions per week. Fifteen participants (41% of the WBV group) completed more than half of the prescribed WBV sessions and were considered “high adherers”, averaging 82% adherence.

Three mild adverse events (AE) out of a total of 1,335 completed WBV sessions (0.2%) were reported on separate occasions and included leg muscle cramps (n=1), malaise (n=1), and knee/back pain (n=1). The muscle cramps and malaise preceded the WBV sessions, while the knee/back pain subsided within one week of the session. There were no serious adverse events and no deviations from the prescribed training protocol for any participant. On a 1–10 Likert pain scale, the average pain rating prior to WBV sessions was 1.3 ± 0.3 and the average pain rating immediately following a WBV session was 1.5 ± 0.7. There were no significant differences in pain ratings before or after WBV sessions nor between high and low adherers.

In the ITT analysis, the WBV group reduced % body fat and fat mass compared to controls. The average differences were 1% body fat and 1kg fat mass between groups and were significantly different (p=.049 and p<0.01, for body fat percent and fat mass, respectively). Differences in weight and android fat mass were nearly significantly different between groups (p=0.08 and p=0.06, for weight and android fat mass, respectively) where controls gained weight and android fat compared to no or little change in the WBV group. When only comparing high adherers in the WBV group to controls, group differences for changes in %body fat and fat mass widened and remained significant, while changes in BMI and weight became significantly different and android fat mass was nearly significantly different between groups (p=0.05) (Table 4). The difference in weight between groups averaged 2kg and in fat mass averaged 1.5kg.

Table 4.

Mixed effects models among WBV participants with high adherence (≥80% of prescribed sessions) controlling for age and baseline BMI.

Characteristic	Control (n = 40)		WBV (n = 15)		Difference in 6-mo. change (95% CI)	p-value for group differences
	Adjusted mean (95% CI)	6-mo, change (95% CI)	Adjusted mean (95% CI)	6-mo. change (95% CI)
BMI (kg/m2)*	28.28 (26.67, 29.90)	0.64 (0.32, 0.95)	27.02 (22.34, 31.71)	−0.01 (−1.63, 0.86)	−0.64 (−1.95, −0.09)	0.026
Body fat (%)	37.88 (35.95, 39.81)	0.46 (−0.06, 0.98)	36.63 (31.02, 42.25)	−0.47 (−1.89, 0.96)	−0.93 (−1.83, −0.02)	0.049
Waist circumference (cm)	98.34 (97.17, 99.52)	2.11 (0.76, 3.46)	97.86 (94.46, 101.26)	0.14 (−3.60, 3.87)	−1.97 (−4.35, 0.41)	0.108
Weight (kg)	78.32 (75.67, 80.98)	1.55 (0.69, 2.42)	78.44 (70.71, 86.17)	−0.29 (−2.68, 2.09)	−1.85 (−3.37, −0.33)	0.020
Fat mass (kg)	30.05 (28.73, 31.36)	0.93 (0.31, 1.56)	29.38 (25.56, 33.21)	−0.60 (−2.33, 1.13)	−1.54 (−2.64, −0.43)	0.008
Lean mass (kg)	48.29 (45.55, 51.02)	0.69 (0.21, 1.17)	49.05 (41.09, 57.01)	0.31 (−1.02, 1.65)	−0.38 (−1.23, 0.47)	0.388
Android fat mass (kg)	2.40 (2.27, 2.54)	0.10 (0.03, 0.17)	2.54 (2.15, 2.93)	−0.02 (−0.22, 0.17)	−0.13 (−0.25, 0.00)	0.051
Visceral fat mass (kg)	0.51 (0.46, 0.55)	0.01 (0.00, 0.03)	0.56 (0.43, 0.69)	−0.01 (−0.06, 0.05)	−0.02 (−0.05, 0.01)	0.234
Energy expenditure (kcal/d)	177.74 (116.93, 238.56)	1.04 (−63.46, 65.53)	289.72 (113.05, 466.39)	−36.22 (−219.95, 147.51)	−37.26 (−156.49, 81.98)	0.544
Energy intake (kcal/d)	2027.99 (1745.79, 2310.19)	96.84 (−229.93, 423.62)	1902.91 (1082.44, 2723.38)	−340.16 (−1264.89, 584.57)	−437.00 (−1034.96, 160.95)	0.160

^*Controlling only for baseline age in BMI model

Discussion

In our sample of overweight undergraduate college students, WBV was both feasible and effective at preventing gains in weight and fat over six months. To our knowledge this is the first study to evaluate WBV as a weight management strategy in college students who are vulnerable to beginning the upward trajectory toward obesity. Though our study sample included a broad mix of students from different types of undergraduate settings, students were willing to enroll in the study and to participate in WBV training but at a slightly lower frequency than prescribed. Across the entire sample, WBV was well tolerated and prevented gains in fat mass and percent body fat and students who met adherence goals for WBV training experienced greater weight and body composition changes than those who trained less often.

Overall, WBV was a feasible and safe weight management approach among students enrolled in undergraduate degree programs. Seventy-one percent of eligible students agreed to participate in the study, slightly short of the target sample of n=89. A change in the study institution and principal investigator contributed to a delay in study start up and shortened the recruitment period. Our study remains one of the largest samples to test WBV as a weight loss strategy in any population.¹⁹ We met our retention goal with 80% of students completing the six-month trial, which was comparable to the only other similar length WBV weight loss trials in obese adults (average age 43 ±10 years).²⁶ Dropouts were nearly all from the medical school site, where students were either enrolled in a rigorous baccalaureate nursing program or at a nearby university that required car or bus travel to the study site. On the other hand, dropout was minimal at the campus where the WBV platform was housed in a student facility and easily accessible, thus future studies or implementation of WBV should ensure platforms are situated close to students. Though adherence to WBV was below target goals of 80%, the attained volume of training, which averaged 1.7 sessions per week and took 41 minutes per week to complete, still prevented weight and fat gain. Most other WBV weight loss studies in adults older than our sample are only 8–12 weeks long and report better adherence to sessions, though it is unclear whether these adherence rates would be maintained long term.¹⁹ Forty percent of WBV students met adherence goals of 80% and in turn experienced a greater degree of benefit from training. Strategies to get students to train 3 times per week, which would only take a little over an hour weekly, should be incorporated into future trials. WBV was very well tolerated as evidenced by an extremely low number of mild adverse events and negligible pain ratings that were unchanged from pre to post WBV session. Recent reviews have raised the concern that some WBV protocols used in research exceed the ISO safety standards for vibration exposure.¹⁹ The vibration dose in our protocol was within the tolerable to safe range and no adverse events or worsening pain was reported by participants. The long-term safety of WBV needs to be studied further and future research can determine the minimum dose of WBV that is both safe and effective for specific outcomes.

WBV training prevented unhealthy changes in body composition at a lower dose of prescribed training, but when training goals were met weight and fat mass dropped. Our findings were similar to other trials that tested 6–12 week-long WBV programs in obese adults where seven out of eight studies reported 2%−6% differences in fat mass between WBV groups and controls.^27–33 Among students who reached the target adherence goal of 80%, group differences in fat mass rose to 5%, a 1.8kg difference in weight between groups became significant, and android fat patterning began to differ between groups. While android fat patterning began to improve over time with WBV, we did not observe a significant change in visceral adipose tissue. We may have been underpowered to detect changes in visceral adipose tissue as this represents a much smaller body composition compartment, but it is also possible that WBV alone may not be enough to elicit changes in this fat depot. The two studies that have reported reductions in visceral adipose tissue with WBV training also included caloric restriction across intervention arms.^26,30 Regardless, the avoidance of weight and fat gain across the broader group is meaningful, because adults typically gain pounds over time and this can eventually lead to or worsen overweight and obesity. The frequency of training achieved by the WBV arm averaged 1.7 sessions per week, translating to 34 minutes per week of training time. High adherers trained an additional 15 minutes per week, still under an hour weekly, and achieved greater benefits. An hour or more of moderate-intensity physical activity every day may be required to prevent weight gain in adulthood.³⁴ Unfortunately, less than 20% of college students report participating in 30 minutes or more of physical activity for 5–7 days a week,¹ and time is a commonly cited barrier to exercise and healthy weight management.^35,36 Since the time commitment for WBV training is relatively short, this approach could be more attractive to busy college students and help them avoid the weight gain that can happen when they are unwilling to engage in enough activity.

The mechanism by which WBV may have an effect on body weight and fat, including visceral adipose tissue, is currently unclear though hypotheses have emerged. Animal studies have provided evidence that WBV reduced adipogenesis and slowed fat acquisition in rats.¹⁷ In human studies, Goto et al reported that brief exposures to WBV stimulated an increase in secretion of epinephrine and norepinephrine, subsequently triggering lipolysis as evidenced by increased circulating free fatty acid concentration during the recovery period.³⁷ WBV may also favorably alter counterregulatory hormones by improving insulin action,³⁸ and promoting growth hormone release,³⁹ which would boost metabolism. Participation in WBV programs has also been reported to increase energy expenditure.^40,41 Future studies should incorporate indirect calorimetry techniques to quantify the degree of energy expenditure from WBV in relation to weight changes.

While our study of WBV in college students was novel and rigorously designed, it has limitations that may impact the interpretation of our findings and which should be considered in the design of future research. Our study included a mix of undergraduate programs, including students enrolled in a baccalaureate nursing program at an academic medical center, students enrolled in undergraduate programs in a large nearby university where students commute to campus and students enrolled at a traditional 4-year undergraduate university where students live on or near campus. Dropout differed between students who traveled to the training site and students whose training site was on the campus where they lived. Furthermore, undergraduate students in our study varied widely in age, from young to middle-aged adults. This allows our findings to generalize to both young and mature students, though the barriers to healthy weight management may differ between these two age groups and thus the suitability of WBV as a weight loss strategy for different aged students should be considered. Also notable is that our sample was not completely gender balanced and lacked much diversity in race/ethnicity, thus future studies should aim for a more representative sample. Our sample size further prevented us from being able to look at the feasibility and efficacy of WBV within specific subgroups of students. As the first study of a novel intervention approach, we offered remuneration for participants who completed testing and training sessions. A small remuneration for completion of testing visits in voluntary human subjects research to offset time and travel costs is common practice and was approved for this study by our institutional review board. We also offered a small incentive for completing training sessions, also approved by our IRB and not considered coercive, that amounted to just under $3/session. Though we cannot be certain that WBV would be as feasible among students who are not paid, knowing now that WBV takes little time and can fend off weight gain may be equally incentivizing to students as a small monetary reward. Our study lacked a dietary component, which could have enhanced weight and fat loss. While it was important to first establish the feasibility of WBV in the college setting before testing the addition of weight loss or a full factorial design, optimal long-term weight management strategies may require combining multiple strategies. For example, Sanudo et al reported that reductions in % body fat and improvements in blood lipids among obese adults were only significant among participants in a combined group of WBV + exercise training + a hypocaloric diet compared to other study arms of diet alone or diet plus exercise.²⁰

Chronic diseases associated with obesity are largely preventable, yet obesity rates in young adults are increasing.⁵ Achieving and maintaining weight loss among overweight and obese individuals is often ineffective or poorly sustained,⁴² thus evidence-based effective strategies for preventing or slowing the rate of weight or fat gain early in the transition years may offer the best overall strategy. According to 2017 census data, 66.7% of high school graduates (aged 16–24 years) enrolled in an undergraduate college program,⁴³ suggesting that a university-based weight management program could be a particularly effective setting for reaching large numbers of students. Despite the well-known role of physical activity in weight management, more than 50% of U.S. adults do not get enough physical activity to provide health benefits; and 2% are not active at all in their leisure time.⁴⁴ Though WBV training may not provide the full array of benefits of physical activity, it may be a feasible starting point for students who are yet unable or unwilling to engage in adequate amounts of aerobic and/or resistance exercise training as a means to maintain a healthy weight and reduce the risk of chronic diseases. Overall, WBV appears to be a feasible, safe and time effective approach to weight management in college students, though strategies to optimize uptake and adherence should continue to be considered as these may lead to the best benefits.

Table 2.

Baseline demographics, body composition and energy balance for the full sample and by study arm.

	Full Sample (N = 77)	Control (N = 40)	WBV (N = 37)
Characteristic	Mean (SD) or %	Mean (SD) or %	Mean (SD) or %	Range	p-value
Age	27.11 (8.21)	26.87 (8.48)	27.36 (8.03)	18.24 – 55.11	0.800
Gender					1.000
Female	75.00%	74.36%	75.68%
Male	25.00%	25.64%	24.32%
Race					0.164
White	73.02%	78.13%	67.74%
Asian	17.46%	18.75%	16.13%
Black	4.76%	0.00%	9.68%
Pacific Islander	3.17%	0.00%	6.45%
American Indian/Alaskan Native	1.59%	3.13%	0.00%
Non-Hispanic	85.33%	82.05%	88.89%		0.610
BMI (kg/m2)	28.25 (5.68)	28.24 (5.05)	28.25 (6.36)	21.37 – 50.27	0.996
Waist circumference (cm)	98.56 (13.87)	98.15 (11.92)	98.98 (15.87)	75.42 – 144.02	0.793
Female	97.80 (13.06)	98.26 (12.25)	97.32 (14.05)	75.90 – 135.91
Male	101.04 (16.52)	98.19 (12.15)	104.21 (20.65)	75.42 – 144.02
Weight (kg)	78.15 (17.43)	77.96 (14.94)	78.36 (19.99)	48.87 – 143.54	0.922
Fat mass (kg)	29.85 (10.58)	29.94 (10.17)	29.76 (11.14)	9.77 – 66.70	0.943
Lean mass (kg)	48.30 (10.14)	48.02 (9.09)	48.59 (11.29)	29.21 – 76.85	0.807
Body fat (%)	37.66 (7.32)	37.94 (7.77)	37.35 (6.88)	15.94 – 53.83	0.725
Android fat mass (kg)	2.46 (1.27)	2.38 (1.12)	2.55 (1.44)	0.52 – 7.06	0.572
Visceral fat mass (kg)	0.52 (0.29)	0.50 (0.26)	0.53 (0.33)	0.09 – 1.67	0.669
Energy expenditure (kcal/d)	185.33 (175.93)	176.85 (154.45)	194.49 (198.32)	0.00 – 797.74	0.666
Energy intake (kcal/day)	1933.60 (822.63)	2038.87 (839.62)	1819.80 (799.57)	668.22 – 4565.81	0.245

Table 3.

Intent-to-treat mixed effects models controlling for age and baseline BMI.

Characteristic	Control (n = 40)		WBV (n = 37)		Difference in 6-mo. change (95% CI)	p-value for group differences
	Adjusted mean (95% CI)	6-mo. change (95% CI)	Adjusted Mean (95% CI)	6-mo. change (95% CI)
BMI (kg/m2)*	28.27 (26.48, 30.07)	0.64 (0.31, 0.96)	28.24 (23.87, 32.60)	0.22 (−0.59, 1.03)	−0.42 (−0.90, 0.07)	0.094
Body fat (%)	37.88 (36.03, 39.73)	0.45 (−0.06, 0.97)	37.28 (32.60, 41.78)	−0.34 (−1.62, 0.94)	−0.79 (−1.56, −0.03)	0.046
Waist circumference (cm)	98.35 (97.01, 99.69)	2.11 (0.67, 3.56)	98.92 (95.64, 102.19)	0.62 (−2.96, 4.21)	−1.49 (−3.64, 0.65)	0.176
Weight (kg)	78.34 (75.61, 81.07)	1.55 (0.67, 2.48)	78.42 (71.78, 85.06)	0.35 (−1.86, 2.60)	−1.21 (−2.53, 0.11)	0.076
Fat mass (kg)	30.05 (28.74, 31.37)	0.91 (0.28, 1.54)	29.73 (26.53, 31.44)	−0.24 (−1.81, 1.34)	−1.15 (−2.09, −0.21)	0.019
Lean mass (kg)	48.29 (45.59, 50.99)	0.67 (0.19, 1.16)	48.69 (42.11, 51.10)	0.52 (−0.70, 1.73)	−0.16 (−0.89, 0.57)	0.673
Android fat mass (kg)	2.40 (2.26, 2.55)	0.10 (0.03, 0.17)	2.54 (2.18, 2.90)	0.00 (−0.18, 0.17)	−0.10 (−0.21, 0.00)	0.055
Visceral fat mass (kg)	0.51 (0.46, 0.56)	0.01 (−0.01, 0.03)	0.53 (0.41, 0.65)	0.00 (−0.05, 0.05)	−0.02 (−0.04, 0.01)	0.302
Energy expenditure (kcal/d)	177.61 (121.77, 233.44)	0.79 (−61.35, 62.93)	194.46 (58.57, 330.34)	10.95 (−146.23, 168.13)	10.16 (−84.88, 105.20)	0.835
Energy intake (kcal/d)	2028.22 (1767.31, 2289.12)	92.19 (−212.25, 396.62)	1815.26 (1180.30, 2450.22)	−191.73 (−967.81, 584.35)	−283.92 (−755.56, 187.73)	0.244

^*Controlling only for baseline age in BMI model.

Data Availability

Data sharing will include individual participant data that underlie the results reported in this article, after deidentification (text, tables, figures, and appendices) and the Study Protocol beginning 9 months and ending 36 months following article publication. Data will be shared with researchers who provide a methodologically sound proposal for individual participant data meta-analysis. Proposals should be directed to wintersk@ohsu.edu. To gain access, data requestors will need to sign a data access agreement. Data will be stored in a university data repository and provided by request.