Table 1.
Baseline characteristics of studies included in the systematic review and meta-analysis
| Study | Country | Study design | Sample size | Age at baseline | Percentage male | Knee OA cases | Follow-up (year) | Covariates in fully adjusted model | NOS score |
|---|---|---|---|---|---|---|---|---|---|
| Grotle et al17 | Norway | Cohort | 1854 | 24–76 | 43.7% | 114 | 10.0 | Age, gender, work type, leisure time activities | 7 |
| Hart et al18 | UK | Cohort | 715 | 54.1 | 0 | 95 | 4.0 | Hysterectomy, ERT, smoking, physical activity, knee pain and social class | 6 |
| Cooper et al19 | UK | Cohort | 346 | >55.0 | 26.3 | 45 | 5.1 | Age and sex | 7 |
| Reijman et al20 | Netherlands | Cohort | 1372 | >55.0 | – | 75 | 6.6 | Age, sex and follow-up time | 8 |
| Järvholm et al21 | Sweden | Cohort | 320 192 | 15–67 | 100% | 502 | 12.0 | Age and smoking | 8 |
| Niu et al22 | USA | Cohort | 2623 | 62.4 | 40% | 163 | 2.5 | Age, gender, race, bone mineral density and knee injury | 7 |
| Liu et al23 | UK | Cohort | 490 532 | 50–69 | 0 | 974 | 2.9 | Age, region of recruitment, deprivation index | 8 |
| Wang et al24 | Australia | Cohort | 41 528 | 25–75 | 41.1% | 541 | 5.0 | Age, gender, country of birth and education | 8 |
| Toivanen et al25 | Finland | Cohort | 823 | >30 | 44.8% | 94 | 22.0 | Age, gender | 9 |
| Lohmander et al8 | Sweden | Cohort | 27 960 | 45–73 | 39.4% | 471 | 11.0 | Age, gender, smoking and physical activity | 9 |
| Felson et al26 | USA | Cohort | 598 | 63.7 | 36.3% | 93 | 8.0 | Age, gender, smoking, physical activity, knee injury and weight change | 7 |
| Hochberg et al27 | USA | Cohort | 437 | >20 | 68.2% | – | 4.0 | Age, gender and smoking | 6 |
| Manninen et al28 | Finland | Cohort | 6647 | 40–64 | – | 126 | 10.0 | None | 7 |
| Shiozaki et al29 | Japan | Cohort | 1191 | 40–65 | 0 | – | 14.0 | Physical exercise and knee injury | 7 |
ERT, estrogen replacement therapy; NOS, Newcastle-Ottawa-Scale; OA, osteoarthritis.