Abstract
Obesity is arguably the most important public health problem facing the United States. 1 A total of 34% of the US population is obese, and an additional 32% is overweight. 2 Differing opinions exist about the usefulness of the primary care setting in the management of obesity. 3 This pilot study sought to compare a popular commercial weight loss program with a clinic-based intervention. A secondary goal was to assess the results of offering intensive weight loss interventions to a low-socioeconomic-status, predominantly ethnic minority patient population.
Methods
The setting was an internal medicine clinic that is 1 of 8 primary care sites affiliated with Denver Health, the main safety net institution for Denver, Colorado. 4 The majority of our patients live at or below 200% of the US federal poverty level.
Study participants were recruited through flyers and referrals from clinic physicians. Inclusion criteria included body mass index of 30 kg/m2 or greater and no life-threatening medical conditions (eg, cancer, end-stage renal disease). Exclusions included use of medications associated with large fluctuations in weight (eg, steroids, second generation antipsychotics) or untreated major depression, assessed with the Patient Health Questionnaire. 5 To enroll in the trial, individuals had to complete 2 visits, 1 for informed consent and 1 for a baseline assessment. If they completed both visits, they were randomized to 1 of the 2 treatment groups.
Participants assigned to the commercial program were given a voucher to attend Weight Watchers for 17 weeks at no cost. Weight Watchers is a well-known commercial program, founded over 40 years ago. Weight Watchers was the first commercial program to study its effectiveness in a randomized trial. 6 Currently, there are 33 Weight Watchers chapters in the city of Denver. If participants did not use the Internet, study team members helped them to find a chapter convenient for them.
Participants assigned to the clinic arm were offered 12 weight loss counseling visits over 17 weeks with a counselor who was either a registered nurse (E.R.) or a clinical registered nurse practitioner (J.C.). Individual weight loss counseling visits lasted 20-30 minutes. The counseling materials we used were downloaded directly from the Diabetes Prevention Program, which showed that a weight loss of 7% of initial weight with an increase in physical activity reduced the risk of progression from impaired glucose tolerance to type 2 diabetes by 58%.7,8 The materials are publicly available (http://www.bsc.gwu.edu/dpp/index.htmlvdoc). Participants in the clinic arm also were offered a choice of either meal replacement shakes (eg, Slim-Fast, Glucerna) or weight loss medication (phentermine) to augment their weight losses.9-12 Participants with contraindications to phentermine (eg, uncontrolled hypertension) were only offered meal replacements. Phentermine was used for only 12 weeks in each patient to stay within the Food and Drug Administration label for short-term use.
Data were analyzed using analysis of covariance to compare the between-group difference for changes in weight, blood pressure, waist circumference, and health-related quality of life, while controlling for sociodemographic characteristics of study participants and for factors in which there were slight baseline differences (weight, sex). For participants that dropped out, we assumed weight regain of 0.3 kg per month until the 4-month time point or until the individual reached his or her starting weight. 13
The main study outcome was weight loss. We estimated that with 25 participants per group and a standard deviation of 3.7 kg for weight loss after 4 months, 14 we had 84% power to detect a difference of 3.1 kg between groups, using a 2-sided P value of 0.05. Recruitment was stopped after 46 patients had been randomized, based on the study’s timeline and budget. Secondary outcomes were changes in health parameters (waist circumference, blood pressure), attendance, health-related quality of life, and self-reported measures of dietary intake and physical activity.
Results
Baseline characteristics of the study participants are shown in Table 1. Individuals assigned to the clinic group had a higher percentage of men, higher weight and waist circumference, and lower systolic blood pressure, but none of these differences were statistically significant.
Baseline Characteristics of Study Participants a
Mean ± SE.
The flow of study patients is shown in Figure 1. Of the 138 patients that expressed interest in the study, 27 were excluded for medical reasons, and 59 patients did not complete enrollment. A total of 52 individuals signed a consent form, of whom 6 were excluded prior to randomization, either because of untreated depression or because they did not return for the second screening visit (baseline assessment). Thus, 46 individuals completed enrollment and were assigned to a treatment arm: 23 in the commercial program and 23 in the clinic arm. Of these, 16 patients in each group completed their follow-up assessment at 4 months. Of the 23 individuals in the clinic arm, 10 used phentermine, and 13 used meal replacements. Of the 13 that used meal replacements, 6 had requested to use phentermine but were not eligible to take it. Exclusions from phentermine included uncontrolled blood pressure (n = 4), resting tachycardia (n = 1), pulmonary hypertension (n = 2), and glaucoma (n =1). Two participants had 2 reasons each why they could not take phentermine.
Flow of Study Participants
Study outcomes are shown in Table 2. Those assigned to the commercial program lost 4.0 ± 1.2 kg, while those assigned to the commercial program lost 0.4 ± 1.1 kg (P = .037 for difference). Among those in the clinic group, individuals that selected phentermine lost 5.0 kg, compared to 3.2 kg for those using meal replacements.
Changes in Weight and Related Health Parameters a
Mean ± SD. All outcomes adjusted for age, sex, race/ethnicity, educational level, as well as baseline weight.
Reductions indicate improvement.
Medical Outcomes Study, Short Form 12: PCS, physical component summary; MCS, mental component summary. Increases indicate improvement.
Participants assigned to the clinic arm attended a mean of 8.4 counseling visits, equal to 70% of their possible maximum of 12 visits per person. Those assigned to the commercial program attended 7.6 group sessions, equal to 44.7% out of a possible maximum of 17 sessions per person.
Changes in blood pressure, waist circumference, and health-related quality of life were similar in the 2 groups. Changes in self-reported measures of dietary intake and physical activity also were similar in the 2 groups (data not shown).
A total of 21 intervention-related adverse events occurred among 18 participants. Among those using phentermine, 3 patients had clinically significant increases in blood pressure, and 1 had a pulse increase. Of these 4 individuals, 2 had resolution of their blood pressure/pulse increase and stayed on phentermine for the full 3 months; 1 participant stopped the drug and switched to meal replacements; and 1 stopped but later restarted.
Discussion
In this pilot randomized trial, we found that primary care patients with obesity lost more weight when assigned to a clinic-based intervention, as compared to attending a popular commercial program. Changes in blood pressure, waist circumference, health-related quality of life, diet, and physical activity were similar in the 2 groups, but the study was not powered to detect these differences. Attendance was better among participants assigned to the clinic arm, which is likely part of the explanation for the greater weight losses in that group.
Weight losses in the study were smaller than expected in both arms. The US randomized trial of Weight Watchers reported a weight loss of 4.3 kg after 1 year, with 2.9 kg maintained after 2 years 6 ; participants assigned to Weight Watches in the current study lost only 0.4 kg after 4 months. Studies that combined intensive individual counseling with meal replacements or weight loss medication demonstrated weight losses of 7-9 kg.15,16 The weight loss of 4.0 kg in clinic-based arm in this study is closer to studies of counseling interventions without meal replacements or medication.17,18 We believe that the smaller weight losses observed in both groups in the current study demonstrate the challenges involved with conducting an intensive lifestyle intervention in a low-income patient population. In our clinic, we offer a 5-week group weight loss visit that is open to all patients. We have experienced a high no-show rate for these visits. The extent of participation in the study intervention (44% to 70% of visits completed) was greater than in routine practice, most likely because (1) patients were screened more carefully for the study and (2) visits were more flexible in the study. All these facts demonstrate the difficulty of meaningfully affecting obesity for an entire population of primary care patients—the practical reality is that intensive interventions affect a subset of these individuals.
Four main limitations of our study should be noted. First, the results are short-term only; thus, we do not know if weight losses were sustained. Second, this was a single-site study with a small sample size, so the implications are limited. Third, although this was a randomized trial, the 2 interventions were different. Weight Watchers is a group program, while in the clinic arm, we provided individual counseling. We would have preferred to offer group counseling (which has been shown to be more effective), but this was not feasible. The choice of meal replacements versus weight loss medication in the clinic arm also was not randomized—this design decision reflects the clinical reality that not all patients are appropriate for weight loss medication. 19 Fourth, although interest in the study was strong, only a minority of interested clinic patients actually completed the entire enrollment process to participate.
There is increasing pressure on primary care providers and health systems to address the problem of obesity in clinical settings.20-22 This trial, although small, begins to address that concern. The results suggest that providing weight management in a clinical setting was more effective than “outsourcing” it to a popular commercial program. We believe that larger comparative studies like this should be done to more definitively answer this question. Such trials should include economic analysis, as clinic-based interventions are more expensive. Ultimately, the goal is to provide guidance to policy makers and health care payers on what types of disease management interventions should be reimbursed.23,24 The lack of consistent reimbursement for treatment of obesity has been considered to be a major barrier to the provision of such treatment.25,26 However, the recent decision from Medicare to reimburse primary care practitioners for intensive weight loss counseling should change how providers view their role in addressing weight management. 27
Conclusions
A clinic-based intervention was more effective than a commercial program for weight loss. We believe that the primary care clinic can be an important route to addressing the epidemic of obesity.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.AQ:Please verify that this statement is accurate and correct.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by a pilot grant from the Nutrition Obesity Research Center at the University of Colorado School of Medicine (5P30DK048520-14S1).