P2502Intensity and duration of lifestyle interventions for long term weight loss and association with mortality: a meta-analysis of randomised trials

Abstract

Abstract Background Lifestyle interventions are widely recommended for weight management, particularly in patients with diabetes and cardiovascular disease. The intensity and duration of interventions needed to achieve sustained weight reduction are uncertain. Objectives To investigate the importance of the frequency and duration of lifestyle interventions for achieving weight loss over ≥ one year, and to evaluate associations between weight loss and all-cause mortality. Method Meta-analysis of randomised trials published in English-language journals from 1980 to June 2018 that assessed lifestyle compared to control interventions on weight loss in ≥100 subjects, and which reported weight change and mortality for ≥ one year. Results 31 randomised trials with a total of 20 816 overweight or obese participants, 70% of whom had cardiometabolic risk factors, were included. The weighted mean difference in body weight between diet and lifestyle intervention and control arms at 1 year was −3.65, 95% confidence interval (CI) −4.66, −2.65 kg, and at 3 years was −2.45, 95% CI: −3.73, −1.17kg. Weight loss at one year was greater in studies with more compared to fewer interventions (>vs ≤28/year (−4.53 [IQR −5.93, −3.12] kg vs −2.38 [−3.98, −0.78] kg, p=0.001). There were 593 deaths (rate ∼0.3%/year), and the odds ratio for mortality for weight loss interventions compared to the controls was 0.86 (95% CI 0.74, 1.01), p=0.09. Table 1. Association with intervention intensity with weight loss achieved and mortality Number of interventions in first year N studies# Number of patients Mean difference p Total deaths Odds ratio p Random effect model (kg) (95% CI) Fixed effect model (95% CI) ≤6 3 510 −0.84 [−1.40, −0.28] <0.0001 5 1.45 [0.22, 9.40] 0.7 7–12 6 2022 −2.04 [−3.24, −0.84] <0.0001 10 1.34 [0.35, 5.16] 0.67 13–24 6* 3490 −2.46 [−5.59, 0.67] <0.0001 23 1.20 [0.49, 2.96] 0.69 ≥25 18 13717 −3.53 [−4.13, −2.92] <0.0001 555 0.84 [0.71, 1.00] 0.05 Total 31 100 −2.98 Total 31 *2 studies reported mortality but not weight loss. Conclusion Lifestyle programs with frequent patient interactions sustained over a year or more achieve modest weight loss. The level of intervention needed for clinically meaningful weight reduction may be unrealistic for most medical practices. There is a modest non-significant reduction in mortality with lifestyle programs, but limited data on whether lifestyle interventions for obesity decrease mortality in persons at higher risk because of cancer, heart failure or ischaemic heart disease. Acknowledgement/Funding None