Table 4. Subgroup analysis on HRs of mortality in older population (mean age≥70 years).
| Group | Number of data clusters(n) | HRa(95%CI) | I2 (%) | P value for heterogeneity in subgroups | P value for heterogeneity between subgroups |
| All-cause mortality | |||||
| Comparisonb | |||||
| Up than median/below median ALT | 3 | 0.60(0.44–0.76) | 0 | 0.534 | |
| Highest/lowest ALT | 4 | 0.78(0.71–0.86) | 66.8 | 0.029 | 0.041 |
| Follow-up (mean year) | |||||
| <10 | 4 | 0.62(0.52–0.73) | 0 | 0.706 | |
| ≥10 | 3 | 0.85(0.76–0.95) | 32.4 | 0.228 | 0.001 |
| Sample size (n) | |||||
| <1000 | 4 | 0.62(0.50–0.75) | 0 | 0.681 | |
| ≥1000 | 3 | 0.81(0.72–0.90) | 71.3 | 0.031 | 0.014 |
| Statin medication | |||||
| No | 6 | 0.78(0.70–0.85) | 58.9 | 0.033 | |
| Yes | 1 | 0.64(0.48–0.80) | none | none | 0.127 |
| Adjustments of MetS componentsc | |||||
| No | 4 | 0.76(0.66–0.86) | 62.3 | 0.047 | |
| Yes | 3 | 0.73(0.64–0.83) | 68.6 | 0.041 | 0.702 |
| Adjustment of alcohol intake | |||||
| No | 3 | 0.66(0.52–0.79) | 0 | 0.966 | |
| Yes | 4 | 0.78(0.70–0.87) | 74.5 | 0.008 | 0.105 |
| Subjects mainly in normal ranged | |||||
| No | 4 | 0.82(0.73–0.92) | 66.5 | 0.030 | |
| Yes | 3 | 0.65(0.55–0.76) | 0 | 0.974 | 0.019 |
| CV-mortality | |||||
| Follow-up (mean year) | |||||
| <10 | 2 | 0.60(0.37–0.83) | 0 | 0.593 | |
| ≥10 | 2 | 0.82(0.69–0.95) | 0 | 0.761 | 0.104 |
| Sample size (n) | |||||
| <1000 | 1 | 0.80(0.30–1.57) | none | none | |
| ≥1000 | 3 | 0.76(0.65–0.88) | 33.5 | 0.223 | 0.927 |
| Statin medication | |||||
| No | 3 | 0.82(0.69–0.95) | 0 | 0.954 | |
| Yes | 1 | 0.58(0.34–0.82) | none | none | 0.088 |
| Adjustments of MetS componentsc | |||||
| No | 1 | 0.81(0.67–0.95) | none | none | |
| Yes | 3 | 0.68(0.48–0.87) | 0 | 0.401 | 0.276 |
| Adjustment of alcohol intake | |||||
| No | 1 | 0.80(0.34–1.88) | none | none | |
| Yes | 3 | 0.76(0.65–0.88) | 33.5 | 0.223 | 0.927 |
| Subjects mainly in normal ranged | |||||
| No | 2 | 0.82(0.69–0.95) | 0 | 0.761 | |
| Yes | 2 | 0.60(0.37–0.83) | 0 | 0.593 | 0.104 |
HR was evaluated between higher and lower ALT categories.
“Up than median/below median ALT” represented comparison between population with ALT value higher than median level and population with ALT value below the median level (ALT level was continuous); “Highest/lowest ALT” represented comparison between population with ALT value in the highest category and population with ALT value in the lowest category (ALT level was incontinuous).
the adjusted MetS components were defined by NCEP-ATP-III criteria [35]; “Yes” represented four or more MetS covariates adjusted; “No” represented less than four MetS covariates adjusted.
“No” represented the study not defined the “normal range” of ALT level or ≤95% of subjects were in the self-defined “normal range ” of ALT level; “Yes” represented the study defined the “normal range” of ALT level and >95% (or large part) of subjects were in the self-defined “normal range” of ALT level.
Abbreviations: ALT: alanine aminotransferase; CI: confidence interval; CV: cardiovascular; HR: hazard ratio.