Table 2.
Author | Selection bias | Measurement bias |
Confounding | Overall (0–4)a | |
---|---|---|---|---|---|
Retirement | Physical activity | ||||
Berger (2005)22 | No | Yes | Yes | ? | 1 |
Brown (2009)23 | Yes | No | No | Yes | 2 |
Chung (2009)24 | No | No | Yes | No | 3 |
Evenson (2002)25 | Yes | No | No | Yes | 2 |
Fonseca (2004)26 | Yes | No | Yes | Yes | 1 |
Glamser (1985)27 | Yes | No | Yes | Yes | 1 |
Henkens (2008)28 | Yes | No | Yes | ? | 1 |
Mein (2005)29 | No | No | Yes | No | 3 |
Midanik (1995)30 | No | No | Yes | Yes | 2 |
Parnes (1985)31 | Yes | No | Yes | No | 2 |
Patrick (1985)32 | Yes | ? | No | Yes | 1 |
Slingerland (2007)33 | Yes | No | Yes | No | 2 |
Touvier (2010)34 | Yes | No | No | Yes | 2 |
Wells (1999)35 | No | No | Yes | Yes | 2 |
Wister (1996)36 | No | No | Yes | ? | 2 |
Chung (2005)37 | No | No | Yes | No | 3 |
Nekuda (2007)38 | ? | ? | Yes | No | 1 |
Parise (2006)39 | ? | Yes | Yes | Yes | 0 |
Zheng (2008)40 | ? | No | Yes | No | 2 |
Note: ?, not reported; studies ordered by type (published articles followed by PhD theses) and then alphabetically by first author
0–1 indicates low quality; 2 indicates modest quality; 3–4 indicates high quality.