Toward the development of self-help health behaviour change programs: weight loss by correspondence

The goal of this study was to evaluate a correspondence weight control program, and to assess the impact of three program elements (weekly homework, interim weigh-ins, and participation deposits) individually and in combination. All treated participants received 15 weekly standard lessons by mail. Three program features were varied factorially: a) homework assignments, b) interim weigh-ins and c) a deposit refunded contingent on returning homework and/or attending interim weigh-ins. Participants were assigned randomly to active treatment conditions or a delayed treatment control group. Among treated males (N = 14), initial average weight loss and BMI reduction were 9.6 kg and 3.1 respectively; average net weight loss and BMI reduction at one year follow-up were 5.8 kg and 1.9 respectively. Among treated females (N = 128), initial average weight loss and BMI reduction were 3.1 kg and 1.2 respectively; average net weight loss and BMI reduction at one year were 2.3 kg and .88 respectively. Women in all treated groups, except lessons only, showed a greater BMI reduction than untreated controls at the end of treatment. Women in conditions including both homework and interim weigh-ins had greater initial BMI reductions (M = 1.6) than those who received lessons only (M = .76). At one year, net BMI reductions were comparable across all treated groups. Of the 42 women initially registered in conditions that included both homework and weigh-ins, 12 who denied joining other programs lost at least 4.5 kg (M = 7.1) during treatment, and 7 had a net loss of at least 4.5 kg (M = 8.0) at one year without apparent involvement in any other program.