Table 1.
Study characteristics and baseline data.
| Study, Place | No. of Participants | Age, mean (range), y | Women, No. (%) | Original depression scale, (range) | Dementia, No. (%) | Thyroid medication, No. (%) | Depression medication, no. (%) | Median TSH, mIU/l | Normal range FT4, pmol/l | First/last available follow-up (mean), y |
|---|---|---|---|---|---|---|---|---|---|---|
| Studies included in the main analyses | ||||||||||
| Leiden 85 + Study, The Netherlands | 281 | 85 (85–85) | 185 (66) | GDS-15 (0–15) | 37 (13) | 11 (4) | 8 (3) | 1.57 | 13–23 | 1.0/3.7 |
| PROSPER Study, The Netherlands | 405 | 75 (70–83) | 190 (47) | GDS-15 (0–15) | 0 (0) | 10 (2) | 12 (3) | 2.04 | 12–18 | 3.0/3.0 |
| InChianti Study, Italy | 952 | 66 (21–98) | 515 (54) | CESD-20 (0–60) | N.A | 24 (2) | 29 (3) | 1.34 | 10–27 | 3.3/3.6 |
| Health ABC Study, United States | 2250 | 75 (69–81) | 1139 (51) | CESD-20 (0–60) | 481 (21) | 212 (9) | 7 (< 1) | 2.15 | 10–23 | 2.9/7.8 |
| CHS, United States | 3419 | 75 (64–98) | 2013 (59) | CESD-10 (0–30) | 451 (13) | 280 (8) | 57 (2) | 2.14 | 9–22 | 1.1/8.2 |
| HUNT, Norway | 15,731 | 53 (19–86) | 10,963 (70) | HADS (0–21) | N.A | 684 (4) | N.A | 1.50 | 8–20 | 11.2/11.2 |
| Overall | 23,038 | 60 (19–98) | 15,005 (65) | N.A | 969 (4) | 1221 (5) | 113 (< 1) | 1.63 | N.A | 8.2/9.9 |
| Studies only included in sensitivity analyses or secondary outcomes | ||||||||||
| MrOs*, United States | 1356 | 73 (65–92) | 0 (0) | GDS-15 (0–15) | 29 (2) | 93 (7) | 71 (5) | 2.04 | 9–24 | 3.9/** |
| PREVEND*, The Netherlands | 2124 | 48 (28–75) | 1086 (51) | N.A.‡ | N.A | 36 (2) | N.A | 1.37 | 12–22 | 4.2/** |
| SHIP*, Germany | 2139 | 46 (20–80) | 1133 (53) | BDI (0–63) | 46 (12) | 105 (5) | 31 (1) | 0.68§ | 8–19 | 9.5/** |
| HIMS†, Australia | 4032 | 77 (71–89) | 0 (0) | N.A | 212 (5) | 126 (3) | 261 (7) | 2.00 | 10–23 | N.A¶ |
Leiden 85 + Study13 Leiden 85 plus Study; PROSPER11 prospective study of Pravastatin in the elderly at risk; Health ABC Study46 The health, aging and body composition study; CHS32 cardiovascular health study; InChianti Study47 Invecchiare in Chianti Study; HUNT48 Nord-Trøndelag Health Study; PREVEND49 prevention of renal and vascular end-stage disease; MrOS50 osteoporotic fractures in men study; SHIP51 study of health in pomerania, HIMS36 health in men study; y year; GDS-15 geriatric depression scale 15-item; CESD-10/-20 center for epidemiologic studies depression 10/20-item scale; HADS hospital anxiety and depression scale; BDI beck depression inventory scale; TSH thyroid stimulating hormone; FT4 free thyroxine; N.A., not available; No. number.
*Not included in the main analysis because no data on depressive symptoms at baseline available.
†Not included in the main analysis because no continuous scale used to measure depressive symptoms during follow up. Incidence of depression available (via Data Linkage).
‡No validated depression scale used.
§SHIP includes participants from Pomerania, where an iodine supplementation program began in the mid-1990s. This shifted the distribution of TSH values towards the left in its first years, which lowered TSH values in the population of the SHIP Study during baseline examinations in 1997–2001.
¶follow-up via data-linkage from baseline 2001–2004, censor date 2013.
**In the sensitivity analyses and subgroup analyses we only assessed the effect estimate at the first available follow-up.