Table 5 Meta-regression Results
Covariate | Model Fit (AIC) | Heterogeneity I2 (%) | Effect of Covariate | Significance (p value) | Key Observations |
|---|---|---|---|---|---|
Geographic Region | 96.55 | 97.71 | Western Pacific: − 0.666 | 0.0888 | Marginally significant effect; Western Pacific prevalence slightly lower |
Study Setting | 98.76 | 97.99 | Hospital-Based: − 0.267 | 0.4449 | There is no significant difference between hospital and community-based studies |
Measurement Tool | 99.19 | 98.02 | DXA: + 0.102 | 0.7121 | No significant difference between BIA and DXA tools |
Study Period | 95.50 | 97.92 | Pre- 2022: − 0.564 | 0.0440 | Pre- 2022 studies reported significantly lower prevalence than 2023–2024 studies |
Sample Size | 84.12 | 97.17 |  > 500: − 1.281 |  < 0.0001 | Larger sample sizes (> 500) are associated with significantly lower prevalence |
Age Groups | 98.96 | 98.04 |  > 70: − 0.274, 60–70: − 0.407 | 0.6539 | No significant effect of age groups on prevalence |
Study Design Category | 97.10 | 97.85 | Cohort: − 0.374, Cross-sectional: − 0.122 | 0.2237 | There is no significant difference between study design types |
Duration of T2DM Years | 94.88 | 97.50 |  > 10 Years: + 0.543 | 0.0572 | Marginally significant; longer duration of diabetes linked to slightly higher prevalence |