Table 3.
Subgroup analysis assessing the effect of pre‐selected moderators on the prevalence of weight loss attempts in ethnic‐minority populations
Moderators | N analyses | Prevalence (%) [95% CI] | Q 1 | p 1 | I 2 (%) |
---|---|---|---|---|---|
Decade of survey | 0.415 | 0.813 | |||
1980–1989 | 1 | 40.0 [35.8, 44.4] | 99.7 | ||
1990–1999 | 3 | 48.5 [46.3, 50.7] | 99.6 | ||
2000–2009 | 6 | 44.9 [28.1, 62.8] | 99.8 | ||
Geographic region | 0.305 | 0.581 | |||
Europe and Central Asia | 2 | 35.6 [33.5, 37.7] | 95.9 | ||
North America | 10 | 41.0 [29.5, 53.5] | 98.8 | ||
Meta‐regressions | N analyses | Slope, SE [95% CI] | Z | p | |
% Women | 12 | −0.009, 0.011 [−0.030, 0.013] | −0.771 | 0.440 | |
Mean age | 11 | 0.042, 0.027 [−0.011, 0.095] | 1.561 | 0.118 |
Cochran's Q statistic and p‐values correspond to subgroup differences in effects.