Effectiveness of Web-Based Interventions for Weight Loss and Weight Loss Maintenance Among Middle-Aged Adults: A Systematic Review and Meta-Analysis

Background/Purpose:Obesity is a continuing public health concern in the US. In 2012, every state has more than 20% of obesity prevalence (CDC, 2013). Web-based health management programs are a feasible way to treat large number of individuals with overweight and obesity. The purpose of this systematic review study is to provide an update of web-based intervention studies for weight loss and weight loss maintenance. In order to achieve the purpose, the objectives of this systematic review are to 1) evaluate the methodological quality and feature of the selected web-based intervention studies, and 2) examine the associations between the quality and features of web-based interventions, and the effectiveness of interventions.

Methods:An extensive literature search from January 2003 and September 2013 was conducted through six databases: PubMed, Medline, Cochrane CENTRAL, CINAHL, and PsycINFO. Using the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) flow diagram (Moher et al., 2009), studies were eligible for the following inclusion criteria: a web-based intervention with the primary aim of weight loss or maintenance, middle aged adults, longer than 6 month follow-up, a randomized clinical trial, and greater than 70% of retention rates. Twelve studies met the inclusion criteria. The total number of participants across the 12 selected studies was 4,669, of which 41.4% was female and the average age was 46.66. In order to assess methodological quality of the selected intervention articles, we used Miller et al (2003)’s 11 criteria (e.g., quality control, follow-up rate etc.). We also used Neve et al. (2010)’s 11 criteria to assess the features of each intervention (e.g., interactions with professionals, buddy program, etc.) RevMan 5.2 software was used to synthesize the data for meta-analysis. Heterogeneity was assessed using chi-squared across the selected studies. The fixed effect model was used because all meta-analysis models reported in this study were homogenous.

Results:Overall, the range of the methodological quality rating of the selected studies is 11 to 15 points (out of 16), whereas the range of the feature rating is 6 to 10 points (out of 11). The results showed that higher scores of the quality and feature were significantly more effective in weight loss and weight loss maintenance than those studies with lower scores. Regarding comparisons of web-based interventions, and other types of interventions, meta-analyses showed that 1) web-based interventions were more effective than self-directed methods or usual care; 2) face-to-face interventions were more effective than web-based interventions; 3) there were no significant differences in the effectiveness between web-based interventions and combined methods of web-based interventions and in-person interventions.

Conclusions and Implications: Largely, the results are consistent with the assumption of cognitive behavior theory. These results may contribute to improve the effectiveness of the existing web-based interventions for weight loss and weight loss maintenance. The multiple aspects of quality and feature should be incorporated into designing a web-based intervention to improve effectiveness. Further implications will be discussed.