Table 2. Associations Between Weight Change Categories and All-Cause Mortality.
| Variablea | Weight change, kg | ||
|---|---|---|---|
| Loss (>2.5) | Stable (≤2.5) | Gain (>2.5) | |
| DFTJ cohort | |||
| No. of events/person years | 238/20 234 | 199/30 695 | 86/10 218 |
| HR (95% CI) | 1.56 (1.28-1.91) | 1 [Reference] | 1.30 (1.00-1.69) |
| Kailuan study | |||
| No. of events/person years | 949/83 503 | 1394/164 441 | 1162/116 981 |
| HR (95% CI) | 1.29 (1.18-1.40) | 1 [Reference] | 1.09 (1.00-1.18) |
| Pooled results | |||
| No. of events/person years | 1187/103 737 | 1593/195 136 | 1248/127 199 |
| HR (95% CI) | 1.33 (1.23-1.43) | 1 [Reference] | 1.10 (1.02-1.19) |
Abbreviations: DFTJ, Dongfeng-Tongji; HR, hazard ratio.
a
The multivariable models were adjusted for height and weight at cohort recruitment, waist circumference change (continuous variables), smoking status, alcohol intake status, dietary pattern, educational attainment, physical activity, hypertension, and diabetes, and stratified by age at risk (5-year interval) and sex. We conducted cohort-specific analyses, which were pooled together using fixed-effect meta-analyses. The P for heterogeneity is 0.087 for the weight loss group, and it is 0.21 for the weight gain group.