Variables Associated with Weight Gain or Reduced Weight Loss Success from Observational
Cohort or Weight Loss Intervention Studies Among Health Plan Members—A Preliminary Analysis
Partners for Better Health/Weigh-To-Be
Nicolaas P. Pronk, PhD, Nancy E. Sherwood, PhD, Brian Martinson, PhD
Health Partners, Minneapolis, MN
Purpose: This analysis involved a preliminary investigation on the identification of variables to:
- Predict significant weight gain among a random sample of older (40 years and older) health plan members who completed surveys in 1995 and 1997 as part of the Partners for Better Health (PBH) Cohort study, and
- Predict weight loss or change variables associated with weight loss among a sample of health plan members enrolled in a two-year weight loss intervention called the Weigh-To-Be study.
Methods: Study 1-PBH: The PBH survey was sent in 1995 and 1997 to a cohort of randomly selected health plan members age 40 and older. All members age 40 and over who were enrolled in the health plan on December 15, 1994, were identified by whether they had a diagnosis of heart disease, diabetes, hypertension, or dyslipidemia. Subsequently, they were classified as having none, one, or more of these four conditions. A random sample of 3,000 was selected from those who had none of four diagnosed conditions, a random sample of 2,500 was selected form those who had one of the conditions, and a sample of 2,500 was selected form those who had two or more of the four conditions. A 60-question, 12-page survey was sent to the subjects in 1995 and again in 1997 (number of respondents with answers for both years was 4,022) that included items on demographics, health status, use of preventive services, dental health, health behaviors, and readiness to change health behaviors. Analyses were conducted using SAS General Linear Models procedure and all results are adjusted for baseline body weight. Study 2--WTB: The WTB study is a two-year RCT on the implementation and evaluation of mail- and telephone-based interventions for weight management in a managed-care setting. Study recruitment was conducted through four MCO clinics over a period of 12 months. Eligibility requirements, intentionally broad, included only age greater than or equal to 18 years and body mass index (BMI) greater than or equal to 27.0 based on reported height and weight. 1801 subjects enrolled in the study. Participants were randomized to one of three experimental conditions: usual care, telephone intervention, or mail intervention. All subjects were asked to complete a survey at baseline and 24 months and were measured for height and weight. Data in the current analysis is pooled and all results are adjusted for baseline body weight and treatment group.
Results: The results of the Study 1-PBH analysis indicate that the following variables were associated with weight gain among older adult health plan members (p<0.05): self-perceived general health status (weight gain among healthier members), feeling worried tense or anxious, not receiving advice on exercise or physical activity (PA) from a health professional other than a RN or MD, receiving advice on exercise or PA from a RN or MD, not receiving advice to eat more fruits, vegetable or fiber, being in an action or maintenance stage of readiness to change for eating less fat, defining PA at work as “mainly sitting”, being classified with “heavy”, “moderate” or “mild” PA at baseline, trying to lose weight at baseline by eating less fat, and having completed a college degree.
The results of the Study 2-WTB analysis indicate that several variables are associated with less weight loss among overweight health plan members who enrolled in a weight loss study, regardless of treatment approach. Significantly (p<0.05) less weight was lost by females, subjects who were on depression medication at baseline, subjects who had previously participated in a formal weight loss program, and subjects who stopped weighing themselves (those who started self-weighing lost weight). In addition, several change scores were predictive of weight loss, namely, an increase in the number of fruit/vegetable/fiber servings, a reduction in the number of servings of fat, an increase in recent exercise, an increase in the mental health score (SF-12 score), and an increase in the physical health score (SF-12 score).
Discussion: Using different types of studies, several predictors of weight gain and less successful weight loss were identified. Among a sample of the general health plan membership age 40 and over, variables associated with weight gain included high self-rated health status, sedentary nature of work, a perception of a physically active lifestyle, and a college degree. In addition, whereas advice from an allied health professional on PA or exercise was related to weight loss, if advice to lose weight was received from a clinician, such advice was actually related to weight gain. On the other hand, among those who were overweight and had enrolled in a weight loss study, mental health status (depression medication use) predicted less weight loss success and more successful weight loss was associated with an increase in the SF-12 mental health score (as well as the physical health score). Self-weighing was identified as a strong predictor of success as those who stopped gained weight and those who started lost weight. Diet (increased fruits and vegetables and reduced fat intake) and recent exercise were also associated with weight loss success. Important learning’s of this analysis using two types of studies include the findings that advice from clinicians regarding weight loss is not necessarily successful when considering the results of the cohort study, but among those who are overweight several behaviors that should be complemented with clinical advice are identified in the intervention study. More specific studies on communication and targeting of messaging should be conducted in this area. In addition, both analyses reported evidence on the important role of mental health. Anxiety and depression were associated with weight gain and less weight loss success and improvement in mental health status was associated with improvement ion weight loss success. Additional research is warranted in this area as well.