Predicting attrition in weight control among older adults

Abstract

The present study hypothesized that a combination of variables would predict adherence in older individuals enrolled in a weight control program, "Dietary Interventions: Evaluation of Technology" (D.I.E.T.). A sample of 150 older adults (mean age = 71.3 years) in the program were characterized as: refusing treatment ("refusers"), dropping out in the intensive 10-week phase of treatment ("early drop-outs"), dropping out in the follow-up phase of treatment ("late drop-outs") and completing treatment ("completers"). Stepwise discriminant function analysis yielded five psychological predictors of leaving treatment (in order of magnitude): perceived health, exercise self-efficacy, concern with dieting, psychological well-being and weight fluctuation. Perceived health and exercise self-efficacy were the most influential discriminators, and early and late drop-outs were the best discriminated. Refusers and completers were found to be more similar than different. Moderate, not high scores were predictive of completing treatment. Early drop-outs reported high perceived health, low exercise self-efficacy, a high concern for dieting, high psychological well-being and high weight fluctuation. In addition, early drop-outs reported the longest history of dieting, while late drop-outs reported the shortest. Post-treatment, noncompleter groups reported similar reasons for leaving. Despite this, baseline psychological predictors were found to be useful in discriminating adherence status, particularly that of drop-outs. Interpretations of each group's profile are provided. In addition, the clinical utility of these findings as well as suggestions for treatment are offered.