Table 2.
Results of the main and secondary/sensitivity analyses.
| N cases | N controls | Odds Ratio (95%CI) | P-value | |
|---|---|---|---|---|
| Composite outcome | ||||
| LTL (age-adjusted) (per 1 SD shorter) | 914 | 5861 | 1·17 (1·05, 1·30) | 0·004 |
| Age at COVID-19 test (per 5 yrs older) | 1·58 (1·51, 1·65) | <0·001 | ||
| Sex (male vs female) | 1·88 (1·62, 2·19) | <0·001 | ||
| Ethnicity (non-White vs White) | 1·80 (1·39, 2·34) | <0·001 | ||
| Separate components as outcome* | ||||
| Hospitalisation | 672 | 5861 | 1·17 (1·03, 1·32) | 0·013 |
| Critical care support | 383 | 1·31 (1·12, 1·53) | <0·001 | |
| Respiratory support | 279 | 1·36 (1·13, 1·64) | <0·001 | |
| Death | 157 | 1·36 (1·07, 1·72) | 0·013 | |
| Population controls | ||||
| LTL (age-adjusted) (per 1 SD shorter) | 914 | 465,946 | 1·19 (1·08, 1·31) | <0·001 |
| Excluding participants with recent hospitalisation | ||||
| LTL (age-adjusted) (per 1 SD shorter) | 732 | 5861 | 1·15 (1·02, 1·30) | 0·019 |
| Mendelian Randomisation | ||||
| MR IVW | 914 | 5861 | 1·30 (0·85, 2·00) | 0·224 |
| MR-median | 1·25 (0·62, 2·50) | 0·537 | ||
The main analysis is based on our composite outcome and the full multivariable model estimates are shown for each risk factor. *For each component of the composite outcome analysed separately, the results shown for these are labelled by outcome component but represent the LTL (age-adjusted) estimate (per 1 SD shorter). For each analysis, the numbers of cases and controls are given alongside the odds ratio, 95% confidence interval and P-value (from logistic regression models or MR). MR IVW: Mendelian randomisation inverse-variance weighted method. MR-median: Mendelian randomisation weighted median method.