Disability and Health Journal 8 (2015) 61e69 www.disabilityandhealthjnl.com
Research Paper
A comparison of two weight management programs for adults with mobility impairments Amanda Reichard, Ph.D.a,*, Muriel D. Saunders, Ph.D.a, Richard R. Saunders, Ph.D.a, Joseph E. Donnelly, Ph.D.b, Eric Lauer, M.P.H.c, Debra K. Sullivan, Ph.D., R.D.d, and Lauren Ptomey, Ph.D.b b
a Institute for Life Span Studies, University of Kansas, 1000 Sunnyside Avenue, Lawrence, KS 66045, USA Cardiovascular Research Institute, Division of Internal Medicine, The University of Kansas Medical Center, 3901 Rainbow Boulevard, Kansas City, KS 66160, USA c Institute on Disability, The University of New Hampshire, 10 West Edge Dr., Durham, NH 03824, USA d Department of Dietetics and Nutrition, The University of Kansas Medical Center, 3901 Rainbow Boulevard, Kansas City, KS 66160, USA
Abstract Background: Individuals with physical disabilities experience disparities in obesity; yet few interventions have incorporated accommodations necessary for weight loss in this population. Objective/Hypothesis: This project compared the effectiveness of two weight loss interventions among individuals with physical disabilities. Methods: Adults with physical disabilities who were overweight or obese were randomized across two diet approaches: a modified version of the MyPlate diet (usual care (UC)) and a modified Stoplight Diet (SLDm) supplemented with portion-controlled meals. Project staff met monthly with each participant to measure weight, 24-recalls of diet intake, self-tracking of foods and beverages, and physical activity during the preceding month. Results: Of 126 enrollees, 70% completed the initial 6-month diet phase and 60% of these completed a 6-month follow-up phase. Participants in the SLDm group reduced weight and BMI during the 6 month intervention, and maintained or lost more weight during the 6 month maintenance period. Alternately, the UC diet resulted in a reduction in weight and BMI only at 6 months. BMI from baseline was significantly more improved for SLDm than UC and, among those who lost weight, the SLDm group lost more weight at 6 and 12 months. Conclusions: These results demonstrate that interventions with proper design and accommodations can overcome the barriers to weight loss unique to individuals with mobility impairments with low income. Additionally, the results suggest that using portion control may be more effective than teaching portion sizes. Published by Elsevier Inc. Keywords: Physical disability; Mobility impairment; Weight loss; Intervention
According to the Centers for Disease Control and Prevention, 35.8 million non-institutionalized adults over the age of 18 living in the U.S. reported a physical functioning difficulty (PD) and 16.7% of adults reported they were unable or had difficulty walking.1 Further, the prevalence of physical difficulties, including mobility impairments, We presented parts of the content of this manuscript at a poster presentation for the American Public Health Association. We have no conflicts of interest to report. This work was funded under a Field Initiated Project from the National Institute on Disability and Rehabilitation Research, award number H133G090230. * Corresponding author. Institute on Disability, University of New Hampshire, 10 West Edge Dr., Suite 101, Durham, NH 03824, USA. E-mail address: [email protected] or Amanda.Reichard@ unh.edu (A. Reichard). 1936-6574/$ - see front matter Published by Elsevier Inc. http://dx.doi.org/10.1016/j.dhjo.2014.08.002
increases with age. More than 60% of adults over the age of 65 report difficulty in at least one basic action or complex activity limitation.1 In addition, research has identified obesity as a health disparity for individuals with PD compared to the general population.2 Specifically, analyses of the Medical Expenditures Panel Survey show an ageadjusted prevalence of overweight and obese Body Mass Indexes (BMIs) among 70.6% of people with PD compared to 59.7% of the general population.3 In the general population, people who are obese also have a higher prevalence of total mortality,4 heart disease,5 diabetes,6 hypertension,7,8 and some cancers.9,10 Studies show that these negative outcomes may be elevated for obese individuals with PD. For example, people with PD and obesity have a higher risk of chronic conditions such as coronary heart disease and diabetes than obese people without a
62
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69 11,12
disability. Studies also suggest that obesity leads to additional risk of skeletal stress and atherogenesis, leading to further physical disabilities.13,14 Additionally, obese individuals with PD, especially disabilities involving lower extremities, have a higher risk of secondary conditions including pressure sores, physical inactivity, depression, and fatigue.12,15 Moreover, as Liou et al (2005)12 state: ‘‘For people with physical disabilities, obesity is doubly disturbing. It is not only linked to an increase in potentially disabling chronic conditions, but when paired with existing functional limitations, may also limit a person’s ability to engage in physical activity and participate in social events and community activities (p. 321).’’ Moreover, people with PD are vulnerable to the same risk factors related to obesity as those without disabilities, but at greater frequency. For example, poverty is a widely accepted risk factor for overweight and obesity, and significantly more people with disabilities (21.2%) than those without a disability (7.5%) have an income below the poverty level.16 People with PD also face additional risks not faced by the general population. They encounter barriers to exercising such as limited availability of accessible exercise facilities, pain and muscle weakness,17 cost of programs,18 lack of accessible, affordable transportation18,19 poor understanding about the capacity and skills needed for exercise, limited social support, and inaccessible equipment at fitness facilities.20,21 People with disabilities also report barriers to a healthy diet including reliance on others to shop and/or cook.22 In addition, biologic characteristics of the disability or effects of prescription medication may predispose them to weight gain.12,23 Yet, few weight loss programs have been designed and implemented to address the unique needs and barriers individuals with physical disability face.24 This intervention addresses the multi-dimensional and complex issues surrounding weight loss for individuals with PD, a population that is at significantly greater risk for obesity and morbidity associated with obesity than those without PD. Specifically, the purpose of this study was to assess the weight loss effectiveness of a diet intervention for individuals with mobility impairments based on portion-controlled meals and education compared to a usual care diet recommended by the United States Department of Agriculture (USDA).
Methods Participants Low-income adults with mobility difficulties were enrolled in the weight management study and randomly assigned to one of the two diet groups, using a 1:1 computergenerated randomization table. Project staff (a registered dietitian and an exercise physiologist) recruited participants by hanging flyers and talking with health care and other
providers in hospitals, clinics, doctor’s offices, and agencies serving individuals with physical disabilities. Participants were eligible if they (a) had a self-reported mobility impairment, (b) were overweight or obese ((BMI) > 25), (c) qualified or were eligible for Medicaid, and (d) lived within 60 miles of Wichita, KS. Limiting the sample to those with low income had a two-fold purpose. First, as indicated above a large proportion of people with disabilities live in poverty and are overweight. Second, we wanted to examine the effect of participation on health utilization by analyzing Medicaid claims for participants pre- and post-intervention (to be reported in a future publication). Potential participants were excluded if they had a diagnosis of Type 1 diabetes, acute heart disease, cancer, or other medical conditions that would affect energy metabolism, or if they had participated in another weight loss program within the last year. Staff conducted an initial meeting with eligible participants to provide information about the diet program. Interested persons also were encouraged to invite someone to attend future meetings and act as a study partner, preferably someone who assisted with grocery shopping and food preparation. Study partners were optional, however, and several chose not to have another person at their meetings. The University of Kansas’ Human Subjects Committee approved all procedures before the project start. A personal consent to participate and a physician’s consent were required prior to enrollment. Those who wished to participate in an exercise program required an additional physician’s consent to exercise. Overview of diet and exercise program The diet programs consisted of 6 months of active dieting, followed by either 6 months of additional dieting or weight maintenance (chosen by the participant, 12 months total). Throughout the project, participants met once a month with project staff, in their homes or other place of their choosing. Staff obtained baseline data and project eligibility through interview and measurement, including height, weight, a standardized multiple-pass 24-h dietary recall of foods and beverages, current medications, type of mobility impairment, use of assistive devices, exercise status, and demographic information. The presence of a physical disability was determined by asking if the applicant had difficulty walking, climbing stairs, or standing for long periods; what if any, assistive device was used to aid in ambulation, and if the physical disability hindered the ability to work or schooling in or outside the home. Assessment of regular exercise routines was determined by asking the applicant exercised, and if so, how many days a week, what were the forms of exercise, and where the exercise was performed. Following the baseline meeting, participants were assigned one of two diets (discussed below) and a research staff member met with the participant and the study partner for 60e90 min to explain the diet and exercise program.
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69
63
Subsequent monthly meetings with the participant, study partner and research staff member lasted approximately 30e45 min for the remainder of the project. At each meeting, participants were weighed and interviewed to obtain a 24-h diet recall. Staff distributed incentives (discussed below) and addressed problems from a strengthsbased perspective. Staff members achieved inter-observer agreement for measuring height, weight, and 24-h food recall prior to beginning the study. All scales used in weight measurement were calibrated to agree within one tenth of a pound. Interobserver agreement was achieved for height when two staff persons measured 10 different people 2 times. Staff agreed on the height 90% of the time. When staff persons conducted recalls, they agreed on 90% of foods recalled across five 24 h-records.
Participants were taught portion sizes using food models, the use of measuring cups, and other visual aids. Participants using the UC diet were also provided with a chart depicting foods in each of the five food groups, and the portion sizes in each depicted on an iconic pyramid.
Intervention components
Data collection and incentives
Each participant was randomly assigned to one of the diets with approximately 1200e1500 calories (adjusted upward based on initial weight), both clinically proven to promote weight loss in average adults. All participants were encouraged to regularly drink water or other zero calorie beverages and to exercise. One diet, referred to as the modified Stop Light Diet (SLDm diet), followed the model used successfully by Saunders et al25 with adults with intellectual or developmental disabilities. This diet consisted of:
Diet intake tracking forms Participants were provided with forms for self-recording daily foods and beverages consumed and types as well as frequency and length of exercise completed. The SLDm tracking form is shown in Fig. 1. The UC diet tracking form was similar, but in place of shakes and entrees, it depicted a carton of milk and cheese, cuts of meat, and a loaf of bread. Participants were instructed to circle or mark the appropriate icon for each food or beverage they consumed; to circle or mark the tennis shoe for each day they exercised, enter what type of exercise, and indicate duration of the exercise. Staff members used the completed forms as a point of discussion at the monthly meetings.
(a) at least 5 daily servings of (fresh, canned or frozen) fruits and vegetables. (b) 2 meal replacement shakes (Provided by Health Management Resources (HMRÒ)). (c) 2 packaged entrees of 300 calories or less, typically found in grocery stores. The SLDm diet program used a visual aid similar to a Stoplight Guide described by Epstein and Squires (1988). This aid listed a typical serving size for approximately 150 food items. Items in the 40e60 calorie range per serving were categorized in the green group, or ‘‘Eat all you want’’; items in the 60e100 range in the yellow group, or ‘‘use caution’’; and items over 100 calories in the red group ‘‘eat rarely/never.’’ Participants were told they if they wanted additional foods, they should choose one from the green group, or occasionally, yellow group on the chart; red foods should only be those found in the packaged entrees. The second diet, referred to the Usual Care diet (UC diet) was adapted from the United Stated Department of Agriculture, MyPlate diet, formally the MyPyramid diet, consisting of: (a) (b) (c) (d)
At least 5 portions of fruits and vegetables 3 portions of dairy products 2 portions of protein 4 portions of grains.
Exercise A simple exercise program was recommended for each participant based on his/her physical capabilities. Those that were ambulatory were encouraged to walk. Those that were unable to bear weight but had movement in their legs or arms were encouraged use one of several ergometers available through the project or to purchase one. In addition, exercises employing therabands were taught and interested participants were provided with therabands, with project funds.
Twenty-four hour diet recalls At monthly meetings, the participant was interviewed to obtain standardized, multiple-pass 24-h recalls. Project staff used food models to show samples of portion sizes, to help participants provide a reasonably accurate report. Food brands were requested. These data were entered into the Nutrition Data System for Research (NDSR, 2008) software, nutritional composition and caloric intake was reviewed by a dietitian, and used for suggesting changes to improve success, or for positive feedback for success with the diet. Incentives At monthly meetings during the first 6 months, participants received 5 cents for each food icon marked on the tracking form. An additional 5 cents per item was deposited in a ‘savings account.’ At each meeting, if a participant’s BMI had decreased by one point, the cumulative amount in the savings account was paid immediately. After the first 6 months, a flat rate of $20 was paid for attending each monthly meeting (and anything remaining in the savings account). No money was paid for walking or other forms of exercise but all types of exercise were encouraged and praised.
64
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69
Fig. 1. Diet log used for the SLDm diet to track food and exercise during the 6 month intervention phase of the diet.
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69
Final monthly meeting When a participant had either completed all monthly sessions or indicated a desire to leave the study, he/she was asked to complete a Satisfaction Survey. The survey was comprised of eight short statements about components of the diet with a 3-item Likert-type rating scale. Statements included: I liked the diet, The diet was easy to follow, The monthly meetings were enjoyable, I liked eating fruits and vegetables, The tracking form helped me stay on the diet, Payments motivated me to stay on the diet, Demonstrations of food models were helpful, and I plan to stay on the diet. Additionally, two openended questions were asked: What was liked most about the diet?, and What was not so good about the diet? Surveys were filled out anonymously and mailed directly to the PI. Data analysis We conducted descriptive and inferential analyses on outcomes for the SLDm diet and the UC diet groups, making within- and between-subject comparisons of baseline, 6-month (post-intervention), and 12-month follow-up assessments. Individuals who did not complete at least 60% of the 6-month intervention were considered to have insufficient participation and data, and their assessments were completely removed from the analysis. Missing outcome data were imputed using multiple imputations under the multivariate normal model, on continuous missing data. The imputation model included baseline weight, age, race, sex and treatment group. Variances within- and acrossimputations were estimated for each of the five resulting imputations. Values for 28 (22%) and 66 (52%) participants were imputed at 6 and 12 months, respectively. Additionally, analysis for the total group and within each weight loss group demonstrated no significant differences between those who had insufficient participation in the first six months and those who completed at least six months of the study, with the exception of diabetes status in the usual care diet; diabetics in the usual care diet were more likely to continue to participate. Similarly, no significant differences existed in the total group or within each type of diet for those who completed at least 6 months compared to those who completed 12 months of the study, with the exception of wheelchair use in the usual care diet; wheelchair users in the usual care diet were more likely to continue to participate. A review of the parametric nature of the data and model fit suggested the use non-parametric statistics. Descriptive results included medians (and associated means) and proportions, for the baseline, post-intervention and 12-month follow-up groups. Baseline results were compared to results after the intervention, and results at 12-month follow-up using Fisher’s exact tests (categorical outcomes) and Wilcoxon rank-sum tests (continuous outcomes). The Wilcoxon signed ranks test was used to compare changes between
65
SLDm diet and UC diet groups at 6-month and 12-month follow-up and to effectively control for all covariates. In addition, a Wilcoxon rank-sum test was used to compare the total weight loss after the intervention and after 12-month follow-up, between the SLDm and UC diet groups. A stringent cut-off of 0.05 was applied to twotailed p-values for statistical significance.
Results Participants At baseline, the SLDm diet group had a median age of 51.7 (mean 51.4) years and a median BMI of 43.3 (mean 43.8); the UC diet group had a median age of 52.8 (mean 52.3) years and a median BMI of 45.6 (mean 45.6) (Table 1). The majority of participants in both groups were white, non-Hispanic, and female. Ninety-one percent reported difficulty in walking, 95% reported difficulty in climbing stairs and/or standing, 86% had difficulty in completing simple daily routines. Seventy-one percent reported using a wheelchair, cane, or other appliance to assist with walking. Fifty-six percent self-reported a diagnosis of Type II diabetes. The majority of participants in both diet groups bought and prepared their own food and nearly half elected to participate in exercise. Wilcoxon signed-rank tests of demographic variables demonstrated no significant differences between the two diet groups at any stage of the intervention (Table 1), and no significant differences from those who dropped out of the study. Intention to treat analyses comparing outcomes between treatments resulted in the same findings as analysis of completers, alone. Weight-loss measures The median BMI and weight for all study participants post-intervention (BMI (ZWSR 5 5.17, p 5 .000); weight (ZWSR 5 5.108, p 5 .000)), and at 12-month follow-up (BMI (ZWSR 5 2.93, p 5 .003); weight (ZWSR 5 2.89, p 5 .004)) differed significantly from weight and BMI at baseline (Fig. 2). Within groups, a Wilcoxon signed ranks test showed that, after the intervention and at 6-month follow-up, the SLDm diet group showed a statistically significant reduction in BMI (ZWSR 6 mos 5 3.79, p 5 .000; ZWSR 12 mos 5 3.44, p 5 .001) and weight loss (ZWSR 6 mos 5 3.68, p 5 .000; ZWSR 12 mos 5 3.44; p 5 .001) among individuals with mobility impairments (Table 2). Median BMI was reduced by more than 2 points and median weight by 13.0 pounds post-intervention, and BMI was reduced by more than 2.5 points and weight by 17.0 pounds after the 12-month follow-up. The Wilcoxon signed ranks test of the UC group showed a statistically significant reduction in BMI (ZWSR 5 3.63, p 5 .000) and weight (ZWSR 5 3.59, p 5 .000) only at 6 months; by 12 months the mean BMI and weight were nearly equivalent to the baseline BMI, with no significant difference. More
66
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69
Table 1 Demographics of participants at baseline, post-intervention, and follow-up Baselinea Post-interventiona
BMI [Median (mean)] Age [Median (mean)] Race White Black 2þ races Hispanic Yes Gender Male Diabetes Yes Wheelchair Yes Living arrangement Alone With 1 other With O1 other Employment Part time Full time Not employed Food buyer Self Family/friend Paid assistant Food prep Self Family/friend Paid assistant Exercise Yes
12-Month follow-upa
Intervention
Control
Intervention
Control
Intervention
Control
n 5 64
n 5 62
n 5 44
n 5 54
n 5 29
n 5 31
42.5 (43.6) 52.4 (52.0)
45.9 52.8
43.3 (43.8) 51.7 (51.4)
45.6 (45.6) 52.8 (52.3)
41.2 (43.0) 51.1 (51.8)
46.3 (45.7) 53.8 (54.3)
68.8 4.7 25.0
61.3 16.1 22.6
70.5 4.5 25.0
63.0 14.8 22.2
75.9 3.4 20.7
61.3 19.4 19.4
7.8
4.8
4.5
5.6
3.4
3.2
14.1
17.7
11.4
14.8
10.3
16.1
60.9
51.6
61.4
48.1
58.6
51.6
45.3
30.6
47.7
35.2
44.8
48.4
51.6 28.1 20.3
41.9 27.4 30.6
52.3 22.7 25.0
44.4 27.8 27.8
55.2 20.7 24.1
45.2 32.3 22.6
4.7 6.3 89.1
17.7 4.8 77.4
6.8 4.5 88.6
13.0 3.7 83.3
10.3 3.4 86.2
16.1 0.0 83.9
67.2 21.9 10.9
75.8 14.5 9.7
65.9 22.7 11.4
75.9 13.0 11.1
62.1 27.6 10.3
71.0 16.1 12.9
64.1 23.4 12.5
75.8 11.3 12.9
61.4 27.3 11.4
72.2 13.0 14.8
65.5 20.7 13.8
67.7 12.9 19.4
48.4
59.7
43.9
56.1
49.0
51.0
a
Fisher’s Exact analyses revealed no statistically significant differences between the intervention and control groups on any of the above characteristics, at any stage of the study.
than 75% (n 5 22) of the SLDm diet group lost weight by 6 months and 72% (n 5 21) were still below baseline weight at 12 months. In comparison, 67% of the UC diet group (n 5 21; mean change in BMI 5 0.74) lost weight by 6 months, and 54.8% (n 5 17) were still below baseline weight at 12 months. No statistically significant difference in the amount of change in BMI and weight between the total samples of the SLDM and UC diet groups was found at the end of the intervention period among completers, per Wilcoxon rank-sum testing. However, intention-to-treat analysis of all randomized participants demonstrated a significant difference in weight change for this time period; SLDm dieters lost more weight than UC dieters. At the 12-month followup, among those who lost weight, the SLDm diet group lost significantly more weight than the UC diet group, at 6 and 12 months (tWRS intervention 5 2.39, p 5 .022; tWRS follow-up 5 2.24, p 5 .032). These results for completers were confirmed by intention-to-treat analyses.
Impact of personal characteristics on weight loss In the SLDm diet group and the total sample, more participants who did not use wheelchairs were more likely to lose weight than wheelchair users in both the SLDm diet group (Fisher’s Exact 5 8.385, p 5 .005) and the total sample (Fisher’s Exact 5 4.172, p 5 .036); however among those who lost weight, there was no difference in the amount of weight lost regardless of wheelchair usage. Those in the SLD diet group who exercised were more likely to lose weight (Fisher’s Exact 5 5.511, p 5 .025) at 6 months, but not at 12 months. In addition, more participants with diabetes than those without diabetes in the UC group lost weight. Although not statistically significant at the .05 level, the difference approached significance (Fisher’s Exact 5 2.901, p 5 .081). For both diet groups, living alone or with others, employment status, food buying and food preparation (self or other) appeared to have no significant effect on weight loss.
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69
67
not good about the diet, although many respondents responded that there was ‘‘nothing bad about the diet.’’ Diet intake analysis Analysis of 24-h diet recall data using the NDSR showed no significant energy intake differences between the UC and the SLDm diet groups at baseline or 12-month follow up. At the end of the 6 month diet period, however, those on the UC diet had a significantly greater energy intake compared to those on the SLDm diet (1361.5 6 661.6 kcals vs. 1142.7 6 621.6 kcals; p 5 .005), and also had greater intake of fat, carbohydrates, and protein than those on the SLDm diet (all p ! .05).
Discussion
Fig. 2. Weight change from baseline by intervention type and time period.
Satisfaction survey Fifty-five (56%) participants returned satisfaction surveys, 29 from the SLDm diet and 26 from the UC diet. More than 86% of each group scored the diet favorably. Those who had unfavorable comments in the open-ended section most frequently cited the ‘‘cost of the foods’’ as what was
The results of this study suggest that the Meal Replacement diet can induce weight loss among individuals with mobility impairments, as it did for individuals with IDD.25 Overall, the SLDm diet group lost more weight than the UC diet group during the intervention phase (median of 13.0 vs. 7.0 pounds) and was able to maintain the loss, or lose more, during the 12-month follow up. In addition, among those who lost weight, the SLDm diet group lost significantly more than the UC diet group, at both measurement points. Weight reduction of at least 5% has been shown to effect significant clinical outcomes such as cardiovascular disease and prevention of diabetes (Perez et al, 2007). In this study, more than 56% of those on the SLDm diet who lost weight lost at least 7% of their baseline weight at the end of the intervention, and more than 51% had maintained or lost at least 7% at the 6 month follow up. These results suggest that, for this population, similar to the general population26,27 controlling portion size and reducing the consumption of high-calorie foods by using
Table 2 Weight measures after intervention and at 12-month follow-up, among completers Post-intervention
Median change in weighta (mean) Median change in BMIa (mean) Number who lost weight n (%) Mean weight loss, among those who lost weightb Lost at least 5%, n (%) At least 7% At least 10% At least 15%
12-Month follow-up
SLDm diet
UC diet
SLDm diet
UC diet
13.00 (10.87)c 2.27 (2.64)c 32 (72.2%) 18.97e 23 (52.3%) 18 (40.9%) 11 (25.0%) 3 (6.8%)
7.00 (5.96)d 1.31 (1.08)d 40 (74.1%) 11.86 14 (27.5%) 8 (12.9%) 3 (5.9%) e
17.00 2.58 21 22.9 24 15 9 5
4.00 (1.39) 0.65 (0.27) 17 (54.8%) 13.76 8 (25.8%) 4 (12.7%) 3 (9.7%) 1 (3.2%)
(14.6)c,f (2.64)c,f (72.4%) (40.0%) (51.7%) (31.0%) (17.2%)
Note: Intention-to-treat analysis demonstrated similar changes in weight post-intervention and at 12-month follow-up. Post intervention and 12-month followup values were 13.1 and 12.5, respectively, for the SLDm diet group and 6.95 and 4.89 for the UC diet group. a Wilcoxon rank-sum test comparing baseline to post-intervention, and baseline to 6-month follow-up. b Fisher’s Exact test comparing MR diet group values to UC Diet values at post-intervention and at 6-month follow-up. c p 5 .000. d p ! .01. e p ! .05. f SLDm dieters significantly different UC dieters ( p ! .05).
68
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69
portion-controlled, healthy meals may be more effective than limiting portions when preparing one’s own meals. Some participants on the SLDm diet plan indicated that they would likely continue to follow the meal replacement diet because they do not have assistance in meal preparation and are looking for an easy method of preparing meals themselves. Others indicated they were not likely to use prepackaged meals, but they learned portion sizes using prepackaged meals and learned to identify high-calorie food items for portion control through use of the Stop Light Chart. Notably, however, the UC diet group also lost weight, on average, and maintained the loss at the 6 month follow up. The mean amount of weight loss was modest and the dropout rate by 6-month follow up was high; however, the finding that more than 73% of all participants lost weight despite the large obstacles (e.g., very high BMI, poverty, chronic health conditions, etc.) indicates that this methodology of combining education about approaches to healthy eating and monthly support meetings, regardless of the diet, is promising for people with mobility impairments who have low income. Although most of the personal characteristics of the participants did not predict weight loss success in either group, there were two exceptions. First, in the SLDm diet group and the total group of participants, more non-wheelchair users than wheelchair users lost weight. Interestingly, however, a greater proportion (62.5 vs. 51.7%, not significant) of wheelchair users than non-users reported exercising. Surprisingly, exercise did not significantly influence weight loss success overall. Additional research is needed to tease out the cause of this, to determine the ways in which those who exercise differ from those who do not. For example, did those who exercise replace more kcals? Was exercise new or was it a continuation? Second, although diabetes management was not the focus of the project, more diabetics than non-diabetics in the UC diet group lost weight. Most of the participants were enrolled in Kansas Medicaid for health insurance, which does not cover the costs of diabetes management. Thus, we hypothesize that the relationship between weight loss and diabetes likely stems from the education participants received about the relationship between what they eat, how it impacts their blood sugar, and how this influences weight loss or gain. This finding points to a need to further investigate the effectiveness of diet and diabetes education on the control of blood sugar and weight loss. Moreover, a greater number of participants in the SLDm diet who prepared their own food than who had it prepared by others lost weight. A better understanding of what underlies this difference could benefit future interventions. Information to tease out includes whether those who have to have their food prepared for them have less control over the food selection? Are those who prepare the food for the participants not supportive of the diet and therefore unwilling to make the recommended healthy dietary changes? Do
those who are able and do prepare their own meals have a greater mental investment in weight loss in some way? Results of this intervention should be viewed as preliminary due to the small sample size and diversity of the sample population. Specifically, the type and severity of the mobility impairment varied widely. In addition, there was a wide range of co-existing disabilities. The age of the participants also was diverse, ranging from 23 to 78 years. The range in BMI scores at baseline also varied from 25.1 to 65.2. Overall, both groups had an average BMI score in the extreme obese range, which may have made weight loss and exercise more difficult to achieve. Thus, future studies of weight management should control for differences in age, BMI, type and severity of mobility impairment, and extent of co-existing disabilities or chronic health conditions. We experienced several barriers in conducting this study. Retention may have been a problem for those with a bias against portion-controlled meals. Eight participants (6%) discontinued participation after learning they had been assigned to the SLDm diet. Only one participant discontinued immediately after assignment to the UC diet. Retention and meeting times were also affected by issues related to poor health and poverty. Ten participants (8%) were discontinued simply because staff was unable to schedule a meeting, despite numerous attempts. Another barrier was in finding portable scales for weighing wheelchairs users, especially those with electric wheelchairs. Our portable scales only weighed reliably up to 450 pounds and many motorized wheelchairs are above that weight. Also, portable scales typically use runners instead of a platform for measuring weight. Wheelchair widths did not always conform to the width of chairs. Thus, for some of our participants, platform scales suited for wheelchair use were the only option. A potential problem for individuals to continue to maintain weight loss or continue to lose weight by following what they learned in this intervention is costs of healthy food.28,29 Most of our participants were on very limited incomes. Many rely on food stamps and community food banks to obtain adequate nourishment. Thus, maintaining a diet that is high in fruits and vegetables and low in high carbohydrates and fat presents a challenge. Monetary incentives were used in this study, in part because food cost was a known obstacle and therefore, payments have augmented how much participants could spend on foods.
Conclusion People with mobility impairments face disparities in obesity compared to the general population. Moreover, they frequently encounter many obstacles for losing weight or maintaining a healthy weight. This study demonstrated that a program of diet and exercise education and monthly counseling with trained professionals can help overweight/obese people with mobility impairments lose weight. In addition,
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69
this intervention highlights the need to continue this line of research, especially to investigate how to better address obesity among those with diabetes and wheelchair users. References 1. Centers for Disease Control and Prevention. Disability and Functioning (Adults) [2/24/2012]. Available from: http://www.cdc.gov/ nchs/fastats/disable.htm; 2011. 2. Centers for Disease Control and Prevention CDC. State-specific prevalence of obesity among adults with disabilitieseeight states and the District of Columbia, 1998-1999. MMWR Morb Mortal Wkly Rep. 2002;51(36):805e808. 3. Reichard A, Stolzle H, Fox MH. Health disparities among adults with physical disabilities or cognitive limitations compared to individuals with no disabilities in the United States. Disabil Health J. 2011;4(2): 59e67. 4. Fontaine KR, Redden DT, Wang C, Westfall AO, Allison DB. Years of life lost due to obesity. J Am Med Assoc. 2003;289(2): 187e193. 5. Grotto I, Huerta M, Kark JD, Shpilberg O, Meyerovitch J. Relation of parental history of coronary heart disease to obesity in young adults. Int J Obes Relat Metab Disord. 2003;27(3):362e368. 6. Woloshin S, Schwartz LM. Press releases: translating research into news. J Am Med Assoc. 2002;287(21):2856e2858. 7. Carnethon MR, Gidding SS, Nehgme R, Sidney S, Jacobs DR Jr, Liu K. Cardiorespiratory fitness in young adulthood and the development of cardiovascular disease risk factors. J Am Med Assoc. 2003;290(23):3092e3100. 8. Neter JE, Stam BE, Kok FJ, Grobbee DE, Geleijnse JM. Influence of weight reduction on blood pressure: a meta-analysis of randomized controlled trials. Hypertension. 2003;42(5):878e884. 9. Evenson KR, Stevens J, Cai J, Thomas R, Thomas O. The effect of cardiorespiratory fitness and obesity on cancer mortality in women and men. Med Sci Sports Exerc. 2003;35(2):270e277. 10. Parker ED, Folsom AR. Intentional weight loss and incidence of obesity-related cancers: the Iowa Women’s Health Study. Int J Obes Relat Metab Disord. 2003;27(12):1447e1452. 11. Doshi JA, Polsky D, Chang VW. Prevalence and trends in obesity among aged and disabled U.S. Medicare beneficiaries, 1997-2002. Health Aff (Millwood). 2007;26(4):1111e1117. 12. Liou T, Pi-Sunyer FX, Laferrere B. Physical disability and obesity. Nutr Rev. 2005;63(10):321e331. 13. Must A, Spadano J, Coakley EH, Field AE, Colditz G, Dietz WH. The disease burden associated with overweight and obesity. J Am Med Assoc. 1999;282:1523e1529.
69
14. Okoro CA, Hootman JM, Strine TW, Balluz LS, Mokdad AH. Disability, arthritis and body weight among adults 45 years and older. Obes Res. 2004;12:854e861. 15. Kinne S, Patrick DL, Doyle DL. Prevalence of secondary conditions among people with disabilities. Am J Public Health. 2004;94(3):443e445. 16. Harris BH, Hendershot G, Stapleton DC. Disability Statistics User Guide Series: A Guide to Disability Statistics from the National Health Interview Survey. Ithaca, NY: Rehabilitation Research and Training Center, Cornell University; October 2005. 17. Weil E, Wachterman M, McCarthy EP, et al. Obesity among adults with disabling conditions. J Am Med Assoc. 2002;288(10):1265e1268. 18. Rimmer JH, Wang E, Smith DS. Barriers associated with exercise and community access for individuals with stroke. J Rehabil Res Dev. 2008;45:315e322. 19. Iezzoni LI, Killeen MB, O’Day BL. Rural residents with disabilities confront substantial barriers to obtaining primary care. Health Serv Res. 2006;41(4):1258e1275. 20. Nary DE, Froehlich K, White GW. Accessibility of fitness facilities for persons with physical disabilities using wheelchairs. Top Spinal Cord Inj Rehabil. 2000;6:87e98. 21. Rimmer JH, Riley B, Wang E, Rauworth A. Accessibility of health clubs for people with mobility disabilities and visual impairments. Am J Public Health. 2005;95:2022e2028. 22. Center for Research on Women With Disabilities. In: Improving the Health and Wellness of Women With Disabilities. Symposium to Establish a Research Agenda. Baylor College of Medicine: Centers for Disease Control and Prevention; 2003. 23. Laino C. Is Your Medicine Cabinet Making You Fat?: MedicineNet. com [January 23, 2009]. Available from: http://www.medicinenet. com/script/main/art.asp?articlekey556339; 2005. 24. Rimmer J, Wang E, Pellegrini CA, Lullo C, Gerber BS. Telehealth weight management intervention for adults with physical disabilities: a randomized controlled trial. Am J Phys Med Rehabil. 2013;92(12): 1084e1094. 25. Saunders RR, Saunders MD, Donnely JE, et al. Evaluation of an approach to weight loss in adults with intellectual or developmental disabilities. Intellect Dev Disabil. 2011;49(2):103e112. 26. Anderson JW, Reynolds LR, Bush HM, Rinsky JL, Washnock C. Effect of a behavioral/nutritional intervention program on weight loss in obese adults. Postgrad Med. 2011;123(5). 27. Houston KA. Systematic review of literature on the effectiveness of behavioral weight loss programs to achieve weight reduction. J Obes Weight Loss Ther. 2012;3(156):2. 28. Drewnowski A, Darmon N. Food choices and diet costs: an economic analysis. J Nutr. 2005;135(4):900e904. 29. Drewnowski A, Specter SE. Poverty and obesity: the role of energy density and energy costs. Am J Clin Nutr. 2004;79(1):6e16.
Research Paper
A comparison of two weight management programs for adults with mobility impairments Amanda Reichard, Ph.D.a,*, Muriel D. Saunders, Ph.D.a, Richard R. Saunders, Ph.D.a, Joseph E. Donnelly, Ph.D.b, Eric Lauer, M.P.H.c, Debra K. Sullivan, Ph.D., R.D.d, and Lauren Ptomey, Ph.D.b b
a Institute for Life Span Studies, University of Kansas, 1000 Sunnyside Avenue, Lawrence, KS 66045, USA Cardiovascular Research Institute, Division of Internal Medicine, The University of Kansas Medical Center, 3901 Rainbow Boulevard, Kansas City, KS 66160, USA c Institute on Disability, The University of New Hampshire, 10 West Edge Dr., Durham, NH 03824, USA d Department of Dietetics and Nutrition, The University of Kansas Medical Center, 3901 Rainbow Boulevard, Kansas City, KS 66160, USA
Abstract Background: Individuals with physical disabilities experience disparities in obesity; yet few interventions have incorporated accommodations necessary for weight loss in this population. Objective/Hypothesis: This project compared the effectiveness of two weight loss interventions among individuals with physical disabilities. Methods: Adults with physical disabilities who were overweight or obese were randomized across two diet approaches: a modified version of the MyPlate diet (usual care (UC)) and a modified Stoplight Diet (SLDm) supplemented with portion-controlled meals. Project staff met monthly with each participant to measure weight, 24-recalls of diet intake, self-tracking of foods and beverages, and physical activity during the preceding month. Results: Of 126 enrollees, 70% completed the initial 6-month diet phase and 60% of these completed a 6-month follow-up phase. Participants in the SLDm group reduced weight and BMI during the 6 month intervention, and maintained or lost more weight during the 6 month maintenance period. Alternately, the UC diet resulted in a reduction in weight and BMI only at 6 months. BMI from baseline was significantly more improved for SLDm than UC and, among those who lost weight, the SLDm group lost more weight at 6 and 12 months. Conclusions: These results demonstrate that interventions with proper design and accommodations can overcome the barriers to weight loss unique to individuals with mobility impairments with low income. Additionally, the results suggest that using portion control may be more effective than teaching portion sizes. Published by Elsevier Inc. Keywords: Physical disability; Mobility impairment; Weight loss; Intervention
According to the Centers for Disease Control and Prevention, 35.8 million non-institutionalized adults over the age of 18 living in the U.S. reported a physical functioning difficulty (PD) and 16.7% of adults reported they were unable or had difficulty walking.1 Further, the prevalence of physical difficulties, including mobility impairments, We presented parts of the content of this manuscript at a poster presentation for the American Public Health Association. We have no conflicts of interest to report. This work was funded under a Field Initiated Project from the National Institute on Disability and Rehabilitation Research, award number H133G090230. * Corresponding author. Institute on Disability, University of New Hampshire, 10 West Edge Dr., Suite 101, Durham, NH 03824, USA. E-mail address: [email protected] or Amanda.Reichard@ unh.edu (A. Reichard). 1936-6574/$ - see front matter Published by Elsevier Inc. http://dx.doi.org/10.1016/j.dhjo.2014.08.002
increases with age. More than 60% of adults over the age of 65 report difficulty in at least one basic action or complex activity limitation.1 In addition, research has identified obesity as a health disparity for individuals with PD compared to the general population.2 Specifically, analyses of the Medical Expenditures Panel Survey show an ageadjusted prevalence of overweight and obese Body Mass Indexes (BMIs) among 70.6% of people with PD compared to 59.7% of the general population.3 In the general population, people who are obese also have a higher prevalence of total mortality,4 heart disease,5 diabetes,6 hypertension,7,8 and some cancers.9,10 Studies show that these negative outcomes may be elevated for obese individuals with PD. For example, people with PD and obesity have a higher risk of chronic conditions such as coronary heart disease and diabetes than obese people without a
62
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69 11,12
disability. Studies also suggest that obesity leads to additional risk of skeletal stress and atherogenesis, leading to further physical disabilities.13,14 Additionally, obese individuals with PD, especially disabilities involving lower extremities, have a higher risk of secondary conditions including pressure sores, physical inactivity, depression, and fatigue.12,15 Moreover, as Liou et al (2005)12 state: ‘‘For people with physical disabilities, obesity is doubly disturbing. It is not only linked to an increase in potentially disabling chronic conditions, but when paired with existing functional limitations, may also limit a person’s ability to engage in physical activity and participate in social events and community activities (p. 321).’’ Moreover, people with PD are vulnerable to the same risk factors related to obesity as those without disabilities, but at greater frequency. For example, poverty is a widely accepted risk factor for overweight and obesity, and significantly more people with disabilities (21.2%) than those without a disability (7.5%) have an income below the poverty level.16 People with PD also face additional risks not faced by the general population. They encounter barriers to exercising such as limited availability of accessible exercise facilities, pain and muscle weakness,17 cost of programs,18 lack of accessible, affordable transportation18,19 poor understanding about the capacity and skills needed for exercise, limited social support, and inaccessible equipment at fitness facilities.20,21 People with disabilities also report barriers to a healthy diet including reliance on others to shop and/or cook.22 In addition, biologic characteristics of the disability or effects of prescription medication may predispose them to weight gain.12,23 Yet, few weight loss programs have been designed and implemented to address the unique needs and barriers individuals with physical disability face.24 This intervention addresses the multi-dimensional and complex issues surrounding weight loss for individuals with PD, a population that is at significantly greater risk for obesity and morbidity associated with obesity than those without PD. Specifically, the purpose of this study was to assess the weight loss effectiveness of a diet intervention for individuals with mobility impairments based on portion-controlled meals and education compared to a usual care diet recommended by the United States Department of Agriculture (USDA).
Methods Participants Low-income adults with mobility difficulties were enrolled in the weight management study and randomly assigned to one of the two diet groups, using a 1:1 computergenerated randomization table. Project staff (a registered dietitian and an exercise physiologist) recruited participants by hanging flyers and talking with health care and other
providers in hospitals, clinics, doctor’s offices, and agencies serving individuals with physical disabilities. Participants were eligible if they (a) had a self-reported mobility impairment, (b) were overweight or obese ((BMI) > 25), (c) qualified or were eligible for Medicaid, and (d) lived within 60 miles of Wichita, KS. Limiting the sample to those with low income had a two-fold purpose. First, as indicated above a large proportion of people with disabilities live in poverty and are overweight. Second, we wanted to examine the effect of participation on health utilization by analyzing Medicaid claims for participants pre- and post-intervention (to be reported in a future publication). Potential participants were excluded if they had a diagnosis of Type 1 diabetes, acute heart disease, cancer, or other medical conditions that would affect energy metabolism, or if they had participated in another weight loss program within the last year. Staff conducted an initial meeting with eligible participants to provide information about the diet program. Interested persons also were encouraged to invite someone to attend future meetings and act as a study partner, preferably someone who assisted with grocery shopping and food preparation. Study partners were optional, however, and several chose not to have another person at their meetings. The University of Kansas’ Human Subjects Committee approved all procedures before the project start. A personal consent to participate and a physician’s consent were required prior to enrollment. Those who wished to participate in an exercise program required an additional physician’s consent to exercise. Overview of diet and exercise program The diet programs consisted of 6 months of active dieting, followed by either 6 months of additional dieting or weight maintenance (chosen by the participant, 12 months total). Throughout the project, participants met once a month with project staff, in their homes or other place of their choosing. Staff obtained baseline data and project eligibility through interview and measurement, including height, weight, a standardized multiple-pass 24-h dietary recall of foods and beverages, current medications, type of mobility impairment, use of assistive devices, exercise status, and demographic information. The presence of a physical disability was determined by asking if the applicant had difficulty walking, climbing stairs, or standing for long periods; what if any, assistive device was used to aid in ambulation, and if the physical disability hindered the ability to work or schooling in or outside the home. Assessment of regular exercise routines was determined by asking the applicant exercised, and if so, how many days a week, what were the forms of exercise, and where the exercise was performed. Following the baseline meeting, participants were assigned one of two diets (discussed below) and a research staff member met with the participant and the study partner for 60e90 min to explain the diet and exercise program.
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69
63
Subsequent monthly meetings with the participant, study partner and research staff member lasted approximately 30e45 min for the remainder of the project. At each meeting, participants were weighed and interviewed to obtain a 24-h diet recall. Staff distributed incentives (discussed below) and addressed problems from a strengthsbased perspective. Staff members achieved inter-observer agreement for measuring height, weight, and 24-h food recall prior to beginning the study. All scales used in weight measurement were calibrated to agree within one tenth of a pound. Interobserver agreement was achieved for height when two staff persons measured 10 different people 2 times. Staff agreed on the height 90% of the time. When staff persons conducted recalls, they agreed on 90% of foods recalled across five 24 h-records.
Participants were taught portion sizes using food models, the use of measuring cups, and other visual aids. Participants using the UC diet were also provided with a chart depicting foods in each of the five food groups, and the portion sizes in each depicted on an iconic pyramid.
Intervention components
Data collection and incentives
Each participant was randomly assigned to one of the diets with approximately 1200e1500 calories (adjusted upward based on initial weight), both clinically proven to promote weight loss in average adults. All participants were encouraged to regularly drink water or other zero calorie beverages and to exercise. One diet, referred to as the modified Stop Light Diet (SLDm diet), followed the model used successfully by Saunders et al25 with adults with intellectual or developmental disabilities. This diet consisted of:
Diet intake tracking forms Participants were provided with forms for self-recording daily foods and beverages consumed and types as well as frequency and length of exercise completed. The SLDm tracking form is shown in Fig. 1. The UC diet tracking form was similar, but in place of shakes and entrees, it depicted a carton of milk and cheese, cuts of meat, and a loaf of bread. Participants were instructed to circle or mark the appropriate icon for each food or beverage they consumed; to circle or mark the tennis shoe for each day they exercised, enter what type of exercise, and indicate duration of the exercise. Staff members used the completed forms as a point of discussion at the monthly meetings.
(a) at least 5 daily servings of (fresh, canned or frozen) fruits and vegetables. (b) 2 meal replacement shakes (Provided by Health Management Resources (HMRÒ)). (c) 2 packaged entrees of 300 calories or less, typically found in grocery stores. The SLDm diet program used a visual aid similar to a Stoplight Guide described by Epstein and Squires (1988). This aid listed a typical serving size for approximately 150 food items. Items in the 40e60 calorie range per serving were categorized in the green group, or ‘‘Eat all you want’’; items in the 60e100 range in the yellow group, or ‘‘use caution’’; and items over 100 calories in the red group ‘‘eat rarely/never.’’ Participants were told they if they wanted additional foods, they should choose one from the green group, or occasionally, yellow group on the chart; red foods should only be those found in the packaged entrees. The second diet, referred to the Usual Care diet (UC diet) was adapted from the United Stated Department of Agriculture, MyPlate diet, formally the MyPyramid diet, consisting of: (a) (b) (c) (d)
At least 5 portions of fruits and vegetables 3 portions of dairy products 2 portions of protein 4 portions of grains.
Exercise A simple exercise program was recommended for each participant based on his/her physical capabilities. Those that were ambulatory were encouraged to walk. Those that were unable to bear weight but had movement in their legs or arms were encouraged use one of several ergometers available through the project or to purchase one. In addition, exercises employing therabands were taught and interested participants were provided with therabands, with project funds.
Twenty-four hour diet recalls At monthly meetings, the participant was interviewed to obtain standardized, multiple-pass 24-h recalls. Project staff used food models to show samples of portion sizes, to help participants provide a reasonably accurate report. Food brands were requested. These data were entered into the Nutrition Data System for Research (NDSR, 2008) software, nutritional composition and caloric intake was reviewed by a dietitian, and used for suggesting changes to improve success, or for positive feedback for success with the diet. Incentives At monthly meetings during the first 6 months, participants received 5 cents for each food icon marked on the tracking form. An additional 5 cents per item was deposited in a ‘savings account.’ At each meeting, if a participant’s BMI had decreased by one point, the cumulative amount in the savings account was paid immediately. After the first 6 months, a flat rate of $20 was paid for attending each monthly meeting (and anything remaining in the savings account). No money was paid for walking or other forms of exercise but all types of exercise were encouraged and praised.
64
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69
Fig. 1. Diet log used for the SLDm diet to track food and exercise during the 6 month intervention phase of the diet.
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69
Final monthly meeting When a participant had either completed all monthly sessions or indicated a desire to leave the study, he/she was asked to complete a Satisfaction Survey. The survey was comprised of eight short statements about components of the diet with a 3-item Likert-type rating scale. Statements included: I liked the diet, The diet was easy to follow, The monthly meetings were enjoyable, I liked eating fruits and vegetables, The tracking form helped me stay on the diet, Payments motivated me to stay on the diet, Demonstrations of food models were helpful, and I plan to stay on the diet. Additionally, two openended questions were asked: What was liked most about the diet?, and What was not so good about the diet? Surveys were filled out anonymously and mailed directly to the PI. Data analysis We conducted descriptive and inferential analyses on outcomes for the SLDm diet and the UC diet groups, making within- and between-subject comparisons of baseline, 6-month (post-intervention), and 12-month follow-up assessments. Individuals who did not complete at least 60% of the 6-month intervention were considered to have insufficient participation and data, and their assessments were completely removed from the analysis. Missing outcome data were imputed using multiple imputations under the multivariate normal model, on continuous missing data. The imputation model included baseline weight, age, race, sex and treatment group. Variances within- and acrossimputations were estimated for each of the five resulting imputations. Values for 28 (22%) and 66 (52%) participants were imputed at 6 and 12 months, respectively. Additionally, analysis for the total group and within each weight loss group demonstrated no significant differences between those who had insufficient participation in the first six months and those who completed at least six months of the study, with the exception of diabetes status in the usual care diet; diabetics in the usual care diet were more likely to continue to participate. Similarly, no significant differences existed in the total group or within each type of diet for those who completed at least 6 months compared to those who completed 12 months of the study, with the exception of wheelchair use in the usual care diet; wheelchair users in the usual care diet were more likely to continue to participate. A review of the parametric nature of the data and model fit suggested the use non-parametric statistics. Descriptive results included medians (and associated means) and proportions, for the baseline, post-intervention and 12-month follow-up groups. Baseline results were compared to results after the intervention, and results at 12-month follow-up using Fisher’s exact tests (categorical outcomes) and Wilcoxon rank-sum tests (continuous outcomes). The Wilcoxon signed ranks test was used to compare changes between
65
SLDm diet and UC diet groups at 6-month and 12-month follow-up and to effectively control for all covariates. In addition, a Wilcoxon rank-sum test was used to compare the total weight loss after the intervention and after 12-month follow-up, between the SLDm and UC diet groups. A stringent cut-off of 0.05 was applied to twotailed p-values for statistical significance.
Results Participants At baseline, the SLDm diet group had a median age of 51.7 (mean 51.4) years and a median BMI of 43.3 (mean 43.8); the UC diet group had a median age of 52.8 (mean 52.3) years and a median BMI of 45.6 (mean 45.6) (Table 1). The majority of participants in both groups were white, non-Hispanic, and female. Ninety-one percent reported difficulty in walking, 95% reported difficulty in climbing stairs and/or standing, 86% had difficulty in completing simple daily routines. Seventy-one percent reported using a wheelchair, cane, or other appliance to assist with walking. Fifty-six percent self-reported a diagnosis of Type II diabetes. The majority of participants in both diet groups bought and prepared their own food and nearly half elected to participate in exercise. Wilcoxon signed-rank tests of demographic variables demonstrated no significant differences between the two diet groups at any stage of the intervention (Table 1), and no significant differences from those who dropped out of the study. Intention to treat analyses comparing outcomes between treatments resulted in the same findings as analysis of completers, alone. Weight-loss measures The median BMI and weight for all study participants post-intervention (BMI (ZWSR 5 5.17, p 5 .000); weight (ZWSR 5 5.108, p 5 .000)), and at 12-month follow-up (BMI (ZWSR 5 2.93, p 5 .003); weight (ZWSR 5 2.89, p 5 .004)) differed significantly from weight and BMI at baseline (Fig. 2). Within groups, a Wilcoxon signed ranks test showed that, after the intervention and at 6-month follow-up, the SLDm diet group showed a statistically significant reduction in BMI (ZWSR 6 mos 5 3.79, p 5 .000; ZWSR 12 mos 5 3.44, p 5 .001) and weight loss (ZWSR 6 mos 5 3.68, p 5 .000; ZWSR 12 mos 5 3.44; p 5 .001) among individuals with mobility impairments (Table 2). Median BMI was reduced by more than 2 points and median weight by 13.0 pounds post-intervention, and BMI was reduced by more than 2.5 points and weight by 17.0 pounds after the 12-month follow-up. The Wilcoxon signed ranks test of the UC group showed a statistically significant reduction in BMI (ZWSR 5 3.63, p 5 .000) and weight (ZWSR 5 3.59, p 5 .000) only at 6 months; by 12 months the mean BMI and weight were nearly equivalent to the baseline BMI, with no significant difference. More
66
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69
Table 1 Demographics of participants at baseline, post-intervention, and follow-up Baselinea Post-interventiona
BMI [Median (mean)] Age [Median (mean)] Race White Black 2þ races Hispanic Yes Gender Male Diabetes Yes Wheelchair Yes Living arrangement Alone With 1 other With O1 other Employment Part time Full time Not employed Food buyer Self Family/friend Paid assistant Food prep Self Family/friend Paid assistant Exercise Yes
12-Month follow-upa
Intervention
Control
Intervention
Control
Intervention
Control
n 5 64
n 5 62
n 5 44
n 5 54
n 5 29
n 5 31
42.5 (43.6) 52.4 (52.0)
45.9 52.8
43.3 (43.8) 51.7 (51.4)
45.6 (45.6) 52.8 (52.3)
41.2 (43.0) 51.1 (51.8)
46.3 (45.7) 53.8 (54.3)
68.8 4.7 25.0
61.3 16.1 22.6
70.5 4.5 25.0
63.0 14.8 22.2
75.9 3.4 20.7
61.3 19.4 19.4
7.8
4.8
4.5
5.6
3.4
3.2
14.1
17.7
11.4
14.8
10.3
16.1
60.9
51.6
61.4
48.1
58.6
51.6
45.3
30.6
47.7
35.2
44.8
48.4
51.6 28.1 20.3
41.9 27.4 30.6
52.3 22.7 25.0
44.4 27.8 27.8
55.2 20.7 24.1
45.2 32.3 22.6
4.7 6.3 89.1
17.7 4.8 77.4
6.8 4.5 88.6
13.0 3.7 83.3
10.3 3.4 86.2
16.1 0.0 83.9
67.2 21.9 10.9
75.8 14.5 9.7
65.9 22.7 11.4
75.9 13.0 11.1
62.1 27.6 10.3
71.0 16.1 12.9
64.1 23.4 12.5
75.8 11.3 12.9
61.4 27.3 11.4
72.2 13.0 14.8
65.5 20.7 13.8
67.7 12.9 19.4
48.4
59.7
43.9
56.1
49.0
51.0
a
Fisher’s Exact analyses revealed no statistically significant differences between the intervention and control groups on any of the above characteristics, at any stage of the study.
than 75% (n 5 22) of the SLDm diet group lost weight by 6 months and 72% (n 5 21) were still below baseline weight at 12 months. In comparison, 67% of the UC diet group (n 5 21; mean change in BMI 5 0.74) lost weight by 6 months, and 54.8% (n 5 17) were still below baseline weight at 12 months. No statistically significant difference in the amount of change in BMI and weight between the total samples of the SLDM and UC diet groups was found at the end of the intervention period among completers, per Wilcoxon rank-sum testing. However, intention-to-treat analysis of all randomized participants demonstrated a significant difference in weight change for this time period; SLDm dieters lost more weight than UC dieters. At the 12-month followup, among those who lost weight, the SLDm diet group lost significantly more weight than the UC diet group, at 6 and 12 months (tWRS intervention 5 2.39, p 5 .022; tWRS follow-up 5 2.24, p 5 .032). These results for completers were confirmed by intention-to-treat analyses.
Impact of personal characteristics on weight loss In the SLDm diet group and the total sample, more participants who did not use wheelchairs were more likely to lose weight than wheelchair users in both the SLDm diet group (Fisher’s Exact 5 8.385, p 5 .005) and the total sample (Fisher’s Exact 5 4.172, p 5 .036); however among those who lost weight, there was no difference in the amount of weight lost regardless of wheelchair usage. Those in the SLD diet group who exercised were more likely to lose weight (Fisher’s Exact 5 5.511, p 5 .025) at 6 months, but not at 12 months. In addition, more participants with diabetes than those without diabetes in the UC group lost weight. Although not statistically significant at the .05 level, the difference approached significance (Fisher’s Exact 5 2.901, p 5 .081). For both diet groups, living alone or with others, employment status, food buying and food preparation (self or other) appeared to have no significant effect on weight loss.
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69
67
not good about the diet, although many respondents responded that there was ‘‘nothing bad about the diet.’’ Diet intake analysis Analysis of 24-h diet recall data using the NDSR showed no significant energy intake differences between the UC and the SLDm diet groups at baseline or 12-month follow up. At the end of the 6 month diet period, however, those on the UC diet had a significantly greater energy intake compared to those on the SLDm diet (1361.5 6 661.6 kcals vs. 1142.7 6 621.6 kcals; p 5 .005), and also had greater intake of fat, carbohydrates, and protein than those on the SLDm diet (all p ! .05).
Discussion
Fig. 2. Weight change from baseline by intervention type and time period.
Satisfaction survey Fifty-five (56%) participants returned satisfaction surveys, 29 from the SLDm diet and 26 from the UC diet. More than 86% of each group scored the diet favorably. Those who had unfavorable comments in the open-ended section most frequently cited the ‘‘cost of the foods’’ as what was
The results of this study suggest that the Meal Replacement diet can induce weight loss among individuals with mobility impairments, as it did for individuals with IDD.25 Overall, the SLDm diet group lost more weight than the UC diet group during the intervention phase (median of 13.0 vs. 7.0 pounds) and was able to maintain the loss, or lose more, during the 12-month follow up. In addition, among those who lost weight, the SLDm diet group lost significantly more than the UC diet group, at both measurement points. Weight reduction of at least 5% has been shown to effect significant clinical outcomes such as cardiovascular disease and prevention of diabetes (Perez et al, 2007). In this study, more than 56% of those on the SLDm diet who lost weight lost at least 7% of their baseline weight at the end of the intervention, and more than 51% had maintained or lost at least 7% at the 6 month follow up. These results suggest that, for this population, similar to the general population26,27 controlling portion size and reducing the consumption of high-calorie foods by using
Table 2 Weight measures after intervention and at 12-month follow-up, among completers Post-intervention
Median change in weighta (mean) Median change in BMIa (mean) Number who lost weight n (%) Mean weight loss, among those who lost weightb Lost at least 5%, n (%) At least 7% At least 10% At least 15%
12-Month follow-up
SLDm diet
UC diet
SLDm diet
UC diet
13.00 (10.87)c 2.27 (2.64)c 32 (72.2%) 18.97e 23 (52.3%) 18 (40.9%) 11 (25.0%) 3 (6.8%)
7.00 (5.96)d 1.31 (1.08)d 40 (74.1%) 11.86 14 (27.5%) 8 (12.9%) 3 (5.9%) e
17.00 2.58 21 22.9 24 15 9 5
4.00 (1.39) 0.65 (0.27) 17 (54.8%) 13.76 8 (25.8%) 4 (12.7%) 3 (9.7%) 1 (3.2%)
(14.6)c,f (2.64)c,f (72.4%) (40.0%) (51.7%) (31.0%) (17.2%)
Note: Intention-to-treat analysis demonstrated similar changes in weight post-intervention and at 12-month follow-up. Post intervention and 12-month followup values were 13.1 and 12.5, respectively, for the SLDm diet group and 6.95 and 4.89 for the UC diet group. a Wilcoxon rank-sum test comparing baseline to post-intervention, and baseline to 6-month follow-up. b Fisher’s Exact test comparing MR diet group values to UC Diet values at post-intervention and at 6-month follow-up. c p 5 .000. d p ! .01. e p ! .05. f SLDm dieters significantly different UC dieters ( p ! .05).
68
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69
portion-controlled, healthy meals may be more effective than limiting portions when preparing one’s own meals. Some participants on the SLDm diet plan indicated that they would likely continue to follow the meal replacement diet because they do not have assistance in meal preparation and are looking for an easy method of preparing meals themselves. Others indicated they were not likely to use prepackaged meals, but they learned portion sizes using prepackaged meals and learned to identify high-calorie food items for portion control through use of the Stop Light Chart. Notably, however, the UC diet group also lost weight, on average, and maintained the loss at the 6 month follow up. The mean amount of weight loss was modest and the dropout rate by 6-month follow up was high; however, the finding that more than 73% of all participants lost weight despite the large obstacles (e.g., very high BMI, poverty, chronic health conditions, etc.) indicates that this methodology of combining education about approaches to healthy eating and monthly support meetings, regardless of the diet, is promising for people with mobility impairments who have low income. Although most of the personal characteristics of the participants did not predict weight loss success in either group, there were two exceptions. First, in the SLDm diet group and the total group of participants, more non-wheelchair users than wheelchair users lost weight. Interestingly, however, a greater proportion (62.5 vs. 51.7%, not significant) of wheelchair users than non-users reported exercising. Surprisingly, exercise did not significantly influence weight loss success overall. Additional research is needed to tease out the cause of this, to determine the ways in which those who exercise differ from those who do not. For example, did those who exercise replace more kcals? Was exercise new or was it a continuation? Second, although diabetes management was not the focus of the project, more diabetics than non-diabetics in the UC diet group lost weight. Most of the participants were enrolled in Kansas Medicaid for health insurance, which does not cover the costs of diabetes management. Thus, we hypothesize that the relationship between weight loss and diabetes likely stems from the education participants received about the relationship between what they eat, how it impacts their blood sugar, and how this influences weight loss or gain. This finding points to a need to further investigate the effectiveness of diet and diabetes education on the control of blood sugar and weight loss. Moreover, a greater number of participants in the SLDm diet who prepared their own food than who had it prepared by others lost weight. A better understanding of what underlies this difference could benefit future interventions. Information to tease out includes whether those who have to have their food prepared for them have less control over the food selection? Are those who prepare the food for the participants not supportive of the diet and therefore unwilling to make the recommended healthy dietary changes? Do
those who are able and do prepare their own meals have a greater mental investment in weight loss in some way? Results of this intervention should be viewed as preliminary due to the small sample size and diversity of the sample population. Specifically, the type and severity of the mobility impairment varied widely. In addition, there was a wide range of co-existing disabilities. The age of the participants also was diverse, ranging from 23 to 78 years. The range in BMI scores at baseline also varied from 25.1 to 65.2. Overall, both groups had an average BMI score in the extreme obese range, which may have made weight loss and exercise more difficult to achieve. Thus, future studies of weight management should control for differences in age, BMI, type and severity of mobility impairment, and extent of co-existing disabilities or chronic health conditions. We experienced several barriers in conducting this study. Retention may have been a problem for those with a bias against portion-controlled meals. Eight participants (6%) discontinued participation after learning they had been assigned to the SLDm diet. Only one participant discontinued immediately after assignment to the UC diet. Retention and meeting times were also affected by issues related to poor health and poverty. Ten participants (8%) were discontinued simply because staff was unable to schedule a meeting, despite numerous attempts. Another barrier was in finding portable scales for weighing wheelchairs users, especially those with electric wheelchairs. Our portable scales only weighed reliably up to 450 pounds and many motorized wheelchairs are above that weight. Also, portable scales typically use runners instead of a platform for measuring weight. Wheelchair widths did not always conform to the width of chairs. Thus, for some of our participants, platform scales suited for wheelchair use were the only option. A potential problem for individuals to continue to maintain weight loss or continue to lose weight by following what they learned in this intervention is costs of healthy food.28,29 Most of our participants were on very limited incomes. Many rely on food stamps and community food banks to obtain adequate nourishment. Thus, maintaining a diet that is high in fruits and vegetables and low in high carbohydrates and fat presents a challenge. Monetary incentives were used in this study, in part because food cost was a known obstacle and therefore, payments have augmented how much participants could spend on foods.
Conclusion People with mobility impairments face disparities in obesity compared to the general population. Moreover, they frequently encounter many obstacles for losing weight or maintaining a healthy weight. This study demonstrated that a program of diet and exercise education and monthly counseling with trained professionals can help overweight/obese people with mobility impairments lose weight. In addition,
A. Reichard et al. / Disability and Health Journal 8 (2015) 61e69
this intervention highlights the need to continue this line of research, especially to investigate how to better address obesity among those with diabetes and wheelchair users. References 1. Centers for Disease Control and Prevention. Disability and Functioning (Adults) [2/24/2012]. Available from: http://www.cdc.gov/ nchs/fastats/disable.htm; 2011. 2. Centers for Disease Control and Prevention CDC. State-specific prevalence of obesity among adults with disabilitieseeight states and the District of Columbia, 1998-1999. MMWR Morb Mortal Wkly Rep. 2002;51(36):805e808. 3. Reichard A, Stolzle H, Fox MH. Health disparities among adults with physical disabilities or cognitive limitations compared to individuals with no disabilities in the United States. Disabil Health J. 2011;4(2): 59e67. 4. Fontaine KR, Redden DT, Wang C, Westfall AO, Allison DB. Years of life lost due to obesity. J Am Med Assoc. 2003;289(2): 187e193. 5. Grotto I, Huerta M, Kark JD, Shpilberg O, Meyerovitch J. Relation of parental history of coronary heart disease to obesity in young adults. Int J Obes Relat Metab Disord. 2003;27(3):362e368. 6. Woloshin S, Schwartz LM. Press releases: translating research into news. J Am Med Assoc. 2002;287(21):2856e2858. 7. Carnethon MR, Gidding SS, Nehgme R, Sidney S, Jacobs DR Jr, Liu K. Cardiorespiratory fitness in young adulthood and the development of cardiovascular disease risk factors. J Am Med Assoc. 2003;290(23):3092e3100. 8. Neter JE, Stam BE, Kok FJ, Grobbee DE, Geleijnse JM. Influence of weight reduction on blood pressure: a meta-analysis of randomized controlled trials. Hypertension. 2003;42(5):878e884. 9. Evenson KR, Stevens J, Cai J, Thomas R, Thomas O. The effect of cardiorespiratory fitness and obesity on cancer mortality in women and men. Med Sci Sports Exerc. 2003;35(2):270e277. 10. Parker ED, Folsom AR. Intentional weight loss and incidence of obesity-related cancers: the Iowa Women’s Health Study. Int J Obes Relat Metab Disord. 2003;27(12):1447e1452. 11. Doshi JA, Polsky D, Chang VW. Prevalence and trends in obesity among aged and disabled U.S. Medicare beneficiaries, 1997-2002. Health Aff (Millwood). 2007;26(4):1111e1117. 12. Liou T, Pi-Sunyer FX, Laferrere B. Physical disability and obesity. Nutr Rev. 2005;63(10):321e331. 13. Must A, Spadano J, Coakley EH, Field AE, Colditz G, Dietz WH. The disease burden associated with overweight and obesity. J Am Med Assoc. 1999;282:1523e1529.
69
14. Okoro CA, Hootman JM, Strine TW, Balluz LS, Mokdad AH. Disability, arthritis and body weight among adults 45 years and older. Obes Res. 2004;12:854e861. 15. Kinne S, Patrick DL, Doyle DL. Prevalence of secondary conditions among people with disabilities. Am J Public Health. 2004;94(3):443e445. 16. Harris BH, Hendershot G, Stapleton DC. Disability Statistics User Guide Series: A Guide to Disability Statistics from the National Health Interview Survey. Ithaca, NY: Rehabilitation Research and Training Center, Cornell University; October 2005. 17. Weil E, Wachterman M, McCarthy EP, et al. Obesity among adults with disabling conditions. J Am Med Assoc. 2002;288(10):1265e1268. 18. Rimmer JH, Wang E, Smith DS. Barriers associated with exercise and community access for individuals with stroke. J Rehabil Res Dev. 2008;45:315e322. 19. Iezzoni LI, Killeen MB, O’Day BL. Rural residents with disabilities confront substantial barriers to obtaining primary care. Health Serv Res. 2006;41(4):1258e1275. 20. Nary DE, Froehlich K, White GW. Accessibility of fitness facilities for persons with physical disabilities using wheelchairs. Top Spinal Cord Inj Rehabil. 2000;6:87e98. 21. Rimmer JH, Riley B, Wang E, Rauworth A. Accessibility of health clubs for people with mobility disabilities and visual impairments. Am J Public Health. 2005;95:2022e2028. 22. Center for Research on Women With Disabilities. In: Improving the Health and Wellness of Women With Disabilities. Symposium to Establish a Research Agenda. Baylor College of Medicine: Centers for Disease Control and Prevention; 2003. 23. Laino C. Is Your Medicine Cabinet Making You Fat?: MedicineNet. com [January 23, 2009]. Available from: http://www.medicinenet. com/script/main/art.asp?articlekey556339; 2005. 24. Rimmer J, Wang E, Pellegrini CA, Lullo C, Gerber BS. Telehealth weight management intervention for adults with physical disabilities: a randomized controlled trial. Am J Phys Med Rehabil. 2013;92(12): 1084e1094. 25. Saunders RR, Saunders MD, Donnely JE, et al. Evaluation of an approach to weight loss in adults with intellectual or developmental disabilities. Intellect Dev Disabil. 2011;49(2):103e112. 26. Anderson JW, Reynolds LR, Bush HM, Rinsky JL, Washnock C. Effect of a behavioral/nutritional intervention program on weight loss in obese adults. Postgrad Med. 2011;123(5). 27. Houston KA. Systematic review of literature on the effectiveness of behavioral weight loss programs to achieve weight reduction. J Obes Weight Loss Ther. 2012;3(156):2. 28. Drewnowski A, Darmon N. Food choices and diet costs: an economic analysis. J Nutr. 2005;135(4):900e904. 29. Drewnowski A, Specter SE. Poverty and obesity: the role of energy density and energy costs. Am J Clin Nutr. 2004;79(1):6e16.
