Table 3.
Meta-regression model with the hazard ratio (95% CI) of the all-cause mortality as a dependent variablea.
| Variable | Unadjusted meta-regression model | Adjusted meta-regression model | τ 2 | ||||
|---|---|---|---|---|---|---|---|
| Coefficient | 95% CI | P | Coefficient | 95% CI | P | ||
| Without any covariate | 0.034 | ||||||
| Univariate model | |||||||
| Age | 0.99 | (0.97–1.00) | 0.34 | 0.364 | |||
| Year | 0.99 | (0.96–1.03) | 0.98 | 0.040 | |||
| Region | |||||||
| United States | 1.04 | (0.77–1.40) | 0.75 | 0.039 | |||
| Europe | 1.06 | (0.77–1.47) | 0.65 | 0.038 | |||
| Follow-up | 1.00 | (0.97–1.03) | 0.58 | 0.038 | |||
| Sample size | 1.00 | (1.00–1.00) | 0.014 | 0.004 | |||
| Diabetes status | 0.95 | (0.69–1.29) | 0.72 | 0.039 | |||
| Cancer status | 0.92 | (0.67–1.25) | 0.57 | 0.037 | |||
| Model 1 | 0.012 | ||||||
| Sample size | 1.00 | (1.00–1.00) | 0.04 | ||||
| Age | 0.99 | (0.98–1.00) | 0.56 | ||||
| Model 2 | |||||||
| Sample size | 1.00 | (0.99–1.00) | 0.07 | ||||
| Age | 0.99 | (0.97–1.01) | 0.56 | ||||
| Year | 0.99 | (0.96–1.03) | 0.92 | ||||
| Model 3 | 0.025 | ||||||
| Sample size | 1.00 | (0.99–1.00) | 0.10 | ||||
| Age | 0.99 | (0.97–1.01) | 0.49 | ||||
| Year | 0.99 | (0.95–1.03) | 0.83 | ||||
| United States | 0.93 | (0.66–1.32) | 0.69 | ||||
| Model 4 | 0.023 | ||||||
| Sample size | 1.00 | (0.99–1.00) | 0.09 | ||||
| Age | 0.99 | (0.97–1.01) | 0.40 | ||||
| Year | 0.99 | (0.95–1.03) | 0.73 | ||||
| Europe | 1.13 | (0.82–1.57) | 0.38 | ||||
| Model 5 | 0.035 | ||||||
| Sample size | 1.00 | (0.99–1.00) | 0.20 | ||||
| Age | 1.00 | (0.96–1.01) | 0.51 | ||||
| Year | 0.99 | (0.94–1.04) | 0.78 | ||||
| United States | 0.93 | (0.63–1.37) | 0.69 | ||||
| Diabetes status | 0.97 | (0.65–1.45) | 0.88 | ||||
| Model 6 | 0.029 | ||||||
| Sample size | 1.00 | (0.99–1.00) | 0.20 | ||||
| Age | 0.99 | (0.96–1.01) | 0.35 | ||||
| Year | 0.98 | (0.94–1.03) | 0.57 | ||||
| Europe | 1.18 | (0.80–1.75) | 0.33 | ||||
| Diabetes status | 0.91 | (0.60–1.36) | 0.61 | ||||
| Model 7 | 0.034 | ||||||
| Sample size | 1.00 | (0.99–1.00) | 0.16 | ||||
| Age | 0.99 | (0.96–1.01) | 0.50 | ||||
| Year | 0.99 | (0.95–1.04) | 0.80 | ||||
| United States | 0.93 | (0.63–1.37) | 0.69 | ||||
| Cancer status | 1.02 | (0.70–1.47) | 0.89 | ||||
| Model 8 | 0.028 | ||||||
| Sample size | 1.00 | (0.99–1.00) | 0.09 | ||||
| Age | 0.98 | (0.96–1.01) | 0.30 | ||||
| Year | 0.98 | (0.94–1.03) | 0.58 | ||||
| Europe | 1.24 | (0.81–1.88) | 0.26 | ||||
| Cancer status | 1.14 | (0.76–1.71) | 0.46 | ||||
The meta-regression model is based on the ReML estimator method. Initial heterogeneity without any variable assigned in the model was τ2 = 0.034. “Age” indicates the mean age; “Duration” indicates the duration of follow-up in years; “Sample size” indicates the total sample size; “Year” indicates the year of study publication; “Coefficient” indicates the regression coefficient; and “τ2” indicates the between-study variance in a random-effects meta-analysis.