Skip to main content
. 2017 Jun 26;19(6):e229. doi: 10.2196/jmir.6972

Table 4.

A summary of meta-analyses.

Review and comparisona Number of included articles Outcome (units) and follow-up N Heterogeneityb Effect sizec (95% CI) P





χ2 (df) P T2 I2

Neve et al, 2009 [14]









Web vs control 3 Weight loss (kg) 151 12.8 (2) .002 84.4% 0.73 (–0.6, 1.51) WMD .07

Enhanced Web vs basic Web 3 Weight loss (kg) 217 3.8 (3) .28 21% 2.24 (1.27, 3.21)d SMD <.001

Web vs control 2 Weight loss maintenance (kg) 409 0.02 (1) .90
–0.30 (–0.34,–0.26)d WMD <.001

Web vs face-to-face 2 Weight loss maintenance (kg) 182 12.2 (3) .007 76% 1.80 (–1.18, 4.79) WMD .24
Kodama et al, 2012 [15] 23 Weight loss, weight loss maintenance (kg) 8697 <.001 84% –0.68 (–1.29,–0.08) WMD .03
Reed et al, 2012 [17]









Intervention with Web vs without Web 5 Weight loss (kg) 336 0.7 (5) .98 0.00 0% –1.48 (–2.52,–0.43) WMD .006

Web vs non-Web 5 Weight loss (kg) 544 14.2 (5) .01 3.61 65% 0.36 (–1.80, 2.53) WMD .74

Web vs non-Web 4 (articles published in 1995 or later) Weight loss (kg) 538 1.7 (4) .78 0.00 0% 1.47 (0.13, 2.81) WMD .03

Intervention with Web vs without Web 4 Weight loss (kg); short-term follow-up 100 0.2 (4) >.99 0.00 0% –1.89 (–3.41,–0.38) WMD .01

Intervention with Web vs without Web 1 Weight loss (kg); long-term follow-up 236 NA NA NA NA –1.10 (–2.55, 0.35) WMD .14

Intervention with Web vs without Web 2 Weight loss (kg); short-term follow-up 53 0.04 (1) .85 0.00 0% –1.95 (–3.50,–0.40) WMD .01

Intervention with Web vs without Web 3 Weight loss (kg) long-term follow-up 283 0.03 (3) >.99 0.00 0% –1.08 (–2.50,–0.34) WMD .14

Intervention with Web vs without Web 2 (articles published prior to 1995) Weight loss (kg) 47 0.02 (2) .99 0.00 0% –0.63 (–7.91, 6.66) WMD .87

Intervention with Web vs without Web 3 (articles published in 1995 or later) Weight loss (kg) 289 0.6 (2) .72 0.00 0% –1.50 (–2.55, 0.44) WMD .006

Intervention with Web vs without Web 3 Weight loss (BMI) 380 0.8 (2) .67 0.00 0% –0.43 (–0.83,–0.03) WMD .04

Web vs non-Web 2 Weight loss (BMI) 51 0.3 (2) .88 0.00 0% 0.44 (–1.15, 2.03)c WMD .59

Web vs control 2 Weight loss (kg) 511 0.04 (1) .84 0.00 0% −1.5 (−2.1, −0.9) MD <.001
Wieland et al, 2012 [18]









Web vs control 2 Weight loss maintenance (kg) 897 0.7 (1) .41 0.00 0% −0.7 (−1.2, −0.2) MD .004

Web vs face-to-face 2 Weight loss maintenance (kg) 897 2.9 (1) .09 0.41 66% 0.5 (−0.5, 1.6) MD .32
Grunenberg et al, 2013 [21]









Web vs control 5 Weight loss (BMI) 727 10.5 (4) .03 0.15 62% –0.49 (–0.95,–0.03) MD .04

Web vs control 5 Weight loss (kg) 727 16.7 (4) .002 1.46 76% –1.32 (–2.59,–0.06) MD .04

a Web vs control: Web-based intervention vs control condition (minimal intervention); enhanced Web vs basic Web: enhanced Web-based interventions vs basic Web-based interventions; Web vs face-to-face: Web-based intervention vs face-to-face intervention; intervention with Web vs intervention without Web: adding a Web-based component to an intervention vs the same intervention without the Web-based component; Web vs non-Web: Web-based interventions vs non-Web-based comparable interventions.

b Ι2: Percentage of the variation across studies attributable to study heterogeneity rather than chance, indicating the level of inconsistency across study results; Τ2: between-study variance.

c Effect sizes were retrieved from original articles reporting a statistically significant pooled effect estimated from at least two trials. All studies except for those indicated used a random effects model to calculate the aggregated effect size. MD: mean difference; SMD: standardized mean difference (Cohen d; standardized weighted aggregated average difference score between conditions across primary studies that use different outcome measures/metrics; to facilitate aggregation across measures/metrics, the between-condition difference for each primary study is converted to standard deviation units that are then weighted with primary studies with more precise estimates carrying more weight in aggregation); WMD: weighted mean difference (unstandardized weighted aggregated average difference score between conditions across primary studies that use the same outcome measure/metric; the between-condition difference for each primary study is weighted with primary studies with more precise estimates carrying more weight in aggregation).

d A fixed effect model was used to calculate the aggregated effect size.