Participants in the Intensive Lifestyle Intervention achieved a 4.7% reduction in initial weight at year 4. This loss is among the largest reported at this length of follow-up for individuals in a randomized controlled trial who were treated by a lifestyle intervention. The results are consistent with those from the DPP (17
) and demonstrate that, with long-term participant support, weight loss achieved with a behavioral intervention is not invariably followed by a return to baseline weight (38
). Nearly 25% of the ILI participants achieved a loss ≥ 10% of initial weight at year 4. Fully 46% had a loss ≥ 5%, an amount widely agreed to produce clinically significant improvements in cardiovascular disease (CVD) risk factors (41
). ILI participants were significantly more successful than their DSE counterparts in reaching all of the categorical weight losses examined and in achieving improvements in glycemic control and several markers of CVD risk, as reported by Wing et al (14
). Additional follow-up, through 2014, will reveal whether the improvements described here are sufficient to significantly reduce incident cardiovascular morbidity and mortality.
The ILI was effective in achieving weight loss at year 4 across a highly diverse sample of participants, demonstrating the potential clinical significance of the intervention. Men lost significantly more weight than women during the first 3 years but not at year 4. Regardless of the statistical findings, differences in weight loss between men and women were not clinically meaningful (0.8%–1.3%) at any time. Consistent with prior findings (43
), the study’s oldest participants lost significantly more weight than their younger counterparts at all 4 years. Older individuals attended significantly more treatment sessions the first year and had significantly more treatment contacts in years 2 to 4 than did younger individuals. Greater treatment participation, as well as a lower self-reported daily calorie intake, likely contributed to the oldest participants’ larger weight loss at year 4. The oldest participants’ superior behavioral adherence could be attributable to their simply having more time to devote to the lifestyle intervention but could reflect other factors, including differences among age groups in motivation to improve health.
Non-Hispanic white participants lost significantly more weight than participants from the three other racial/ethnic subgroups at the end of the first and second years, as observed in previous studies (17
). At years 3 and 4, however, there were no statistically significant differences among the four sub-groups, after adjustment for multiple comparisons. The relative equivalence in outcomes in these later years was attributable to smaller weight regain in African-American, Hispanic, and American-Indian participants than in non-Hispanic whites. A similar convergence in long-term weight loss among racial/ethnic subgroups was reported in the DPP (43
) and in the Trials of Hypertension Prevention (TOHP) (44
). In the present study, the mean loss in African-American women of 4.4 ± 0.5% at year 4 was approximately double that reported in other long-term trials (17
). This loss may reflect interventionists’ efforts to tailor treatment to participants’ potential cultural differences (33
) but also may be attributable to the study’s strong behavioral protocol that recommended, among other components, recording food and calorie intake and using meal replacements, the latter of which has been shown to improve the induction of weight loss (28
). Results of the step-wise regression analysis, which revealed that gender, age, and race/ethnicity accounted for only 2.5% of the variance in weight loss at year 4, again suggest the effectiveness of the lifestyle intervention across a diverse sample of participants.
Losing a large amount of weight the first year was by far the strongest determinant of achieving a large loss at year 4. The step-wise regression analysis revealed that weight loss at 1 year accounted for an additional 22% of the variance in year-4 weight loss, beyond the 6.5% attributable to demographic characteristics and treatment attendance. Additional analyses showed that the odds of achieving a loss ≥ 10% at year 4 were 10.4 times greater in persons who had lost ≥ 10% at 1 year as compared to individuals who had lost < 5% at 1 year. In this latter group, only 7.1% of participants eventually achieved a 10% loss, the individual goal prescribed for participants. Losing 10% of initial weight the first year similarly improved the odds of having a loss ≥ 5% at year 4. The present results extend in a far larger sample – and over a longer period of follow-up – findings from several studies that suggest the importance of large initial weight loss for maintaining a clinically significant long-term reduction (19
). Further study, however, as will be provided by Look AHEAD, is needed of the long-term health consequences of initially losing a large amount of weight and then potentially regaining some or all of it. Are there, for example, different health consequences of losing 10% of initial weight in the first year and regaining half this amount at follow-up, as compared with losing only 5% the first year but maintaining this full loss at follow-up?
The present study provided an exceptional opportunity to prospectively examine the maintenance of weight loss in 887 participants who lost ≥ 10% of initial weight in the first year and were followed through year 4. As shown in , 42.2% (N = 374) of this subgroup maintained a loss ≥ 10% at year 4, and an additional 28.3% (N = 251) kept off 5.0 to 9.9%. We believe that the successful maintenance of weight loss in ILI participants was attributable to their being provided twice-monthly counseling contacts with their lifestyle interventionist to facilitate continued adherence to the study’s diet and activity goals. Several randomized trials have demonstrated the benefit of such long-term participant-provider contact (7
). In the 887 participants who lost ≥ 10% at 1 year, those who maintained this loss at year 4 completed significantly more treatment contacts per year in years 2–4 than did individuals who regained to their baseline weight (N = 88) or who maintained a loss of only 0 to 4.9% (N = 174). Similarly, the 10% maintainers reported significantly greater physical activity at year 4 than did participants in the three other weight categories. They also reported a significantly lower calorie intake at year 4 than participants who regained all of their lost weight. Thus, at year 4, successful maintainers in Look AHEAD appeared to have taken greater advantage of the treatment sessions provided and displayed eating and activity behaviors similar to those of individuals in the National Weight Control Registry (who have lost at least 30 lb and kept off the weight for at least 1 year) (23
Look AHEAD’s study design prevents us from definitively determining the contribution to long-term weight loss of the lifestyle intervention’s different treatment components --including the prescription of meal replacements and high levels of physical activity, use of a treatment toolbox, and the provision of Refresher Groups and National Campaigns. A separate report will examine the weight losses associated with several of the Refresher Groups and National Campaigns. We also were unable to fully evaluate the effects of taking insulin on weight loss, in either the ILI or DSE groups. Persons in both groups who took insulin at baseline and year 4 lost significantly less weight than those who were free of insulin at both times. However, differences in weight loss between insulin users and non-users were not clinically meaningful and should mitigate concerns that insulin users cannot lose weight. Persons who initiated insulin after the study began generally had suboptimal weight loss, but insulin may have been introduced because of lack of weight loss and suboptimal glycemic control.
Despite the lifestyle intervention’s strengths in producing a weight loss of 4.7% at year 4 and in facilitating the maintenance of weight loss in a substantial subset of participants, the intervention clearly had some limitations. Perhaps foremost among them was the inability to induce a loss of 5% of initial weight in the first year in 729 participants, as shown in . Only 21.8% of participants (N = 159) in this subset achieved a weight loss of this size at year 4. This finding underscores the need for further research on behavioral and other methods to induce weight loss in persons with documented treatment-resistant obesity, similar to research on individuals with treatment-resistant depression in which additional therapies are introduced if the first intervention is not successful (45
). While pharmacotherapy and bariatric surgery are options for some individuals (46
), they are not medically appropriate for (or acceptable to) many overweight and obese patients. Additional options are needed (47
Look AHEAD’s treatment protocols for the first 4 years of intervention are now available to practitioners (and researchers) and could help them improve weight management in their own overweight/obese patients. Efforts, however, undoubtedly will be needed to translate Look AHEAD’s intensive and costly program of lifestyle modification (delivered by experienced interventionists in academic medical centers) for use with patients treated in primary care practice and community settings. DPP investigators have already begun to address this efficacy-to-effectiveness translation. Ackermann et al tested a modified version of the first-year DPP protocol, which used group rather than individual counseling sessions and was delivered at YMCA sites by staff health counselors (48
). Intervention participants lost 6 kg in the first 6 months and maintained the loss at 1 year. The cost of the intervention was reduced to approximately $300 per participant, approximately one-fifth of the expense for the first year of treatment in the DPP study (49
). McTigue et al similarly reported a 1-year weight loss of 4.8 kg, achieved with an Internet-delivered version of the DPP, although this was an uncontrolled trial (50
). If, as expected, the Look AHEAD lifestyle intervention is found to reduce cardiovascular morbidity and mortality, substantial effort will need to be devoted to identifying cost-effective methods of delivering the intervention to the millions of individuals who would benefit from it.