Table 1:
Main characteristics of the 121 studies reporting assessment of the effect of workplace wellness programmes
| All studies (n=121) | Randomised controlled trials (n=82) | Quasi-experimental studies (n=39) | |
|---|---|---|---|
|
| |||
| Study details | |||
| Publication date | |||
| Before 2000 | 18 (15%) | 13 (16%) | 5 (13%) |
| 2000–09 | 31 (26%) | 14 (17%) | 17 (44%) |
| 2010 or later | 72 (60%) | 55 (67%) | 17 (44%) |
| Country | |||
| North America (US and Canada) | 57 (47%) | 46 (56%) | 11 (28%) |
| Europe*, Australia, New Zealand | 36 (30%) | 22 (27%) | 14 (36%) |
| Asia† | 22 (18%) | 9 (11%) | 13 (33%) |
| Other‡ | 6 (5%) | 5 (6%) | 1 (3%) |
| Unit of randomisation | |||
| Individual | ·· | 24 (29%) | ·· |
| Cluster | ·· | 58 (71%) | ·· |
| Lost to follow-up (%) | 26·0% (20·9) | 24·4% (20·2) | 29·6% (22·2) |
| Median (IQR) | 20·9% (12·0–37·4) | 19·8% (12·0–34·6) | 27·9% (12·2–43·3) |
| Bias score§ | 3·4 (1·0) | 3·8 (0·7) | 2·4 (0·9) |
| Median (IQR) | 4·0 (3·0–4·0) | 4·0 (3·0–4·0) | 2·0 (2·0–3·0) |
| Workplace settings | |||
| Type¶ | |||
| Office | 19 (16%) | 13 (16%) | 6 (16%) |
| Hospital | 15 (13%) | 14 (18%) | 1 (3%) |
| School | 5 (4%) | 5 (6%) | 0 |
| Factory | 22 (19%) | 11 (14%) | 11 (30%) |
| Mixed or other | 56 (48%) | 37 (46%) | 19 (51%) |
| Number of sites | 9·4 (14·2) | 10·5 (15·3) | 7·2 (11·2) |
| Median (IQR) | 4·0 (2·0–13·0) | 4·0 (2·0–16·0) | 3·0 (2·0–7·5) |
| Employee characteristics | |||
| Number of participating employees | 1231 (3147) | 886 (1532||) | 1964 (4870) |
| Median (IQR) | 413·0 (124·0–904·0) | 314·0 (98·0–782·0) | 547·0 (253·0–1749·0) |
| Age, years | 42·7 (5·7) | 43·0 (5·8) | 41·8 (5·4) |
| Median (IQR) | 43·0 (38·9–46·0) | 43·6 (39·9–46·5) | 42·0 (37·3–44·6) |
| Sex, percentage male** | 51·4% (30·8) | 48·4% (31·4) | 58·0% (29·1) |
| Median (IQR) | 49·2% (24·6–76·1) | 46·8% (22·6–74·4) | 60·1% (37·7–77·1) |
| Race or ethnicity, percentage white†† | 56·0% (34·2) | 62·9% (30·8) | 36·3% (36·7) |
| Median (IQR) | 61·3% (32·4–89·8) | 73·1% (42·4–90·8) | 38·1% (0·0–68·3) |
| BMI (kg/m2) | 28·0 (3·3) | 28·7 (3·5) | 26·5 (2·6) |
| Median (IQR) | 27·5 (25·5–29·7) | 28·2 (25·7–31·6) | 26·4 (25·2–28·1) |
| Intervention characteristics | |||
| Duration, months | 13·3 (13·8) | 10·1 (8·6) | 20·0 (19·3) |
| Median (IQR) | 9·0 (4·5–18·0) | 6·0 (3·0–15·0) | 12·0 (8·0–24·0) |
| Primary specified target‡‡ | |||
| Diet | 94 (78%) | 60 (73%) | 34 (87%) |
| Physical activity | 81 (67%) | 51 (62%) | 30 (77%) |
| Weight management | 53 (44%) | 36 (44%) | 17 (44%) |
| Other | 76 (63%) | 52 (63%) | 24 (62%) |
| Number of components | |||
| 2 | 12 (10%) | 6 (7%) | 6 (15%) |
| 3 | 31 (26%) | 16 (20%) | 15 (38%) |
| 4 | 22 (18%) | 15 (18%) | 7 (18%) |
| 5 | 14 (12%) | 11 (13%) | 3 (8%) |
| 6 | 10 (8%) | 9 (11%) | 1 (3%) |
| 7 | 5 (4%) | 5 (6%) | 0 |
| 8 | 8 (7%) | 6 (7%) | 2 (5%) |
| 9 | 9 (7%) | 6 (7%) | 3 (8%) |
| 10 | 10 (8%) | 8 (10%) | 2 (5%) |
Data are n (%) or mean (SD), unless otherwise specified.
=not applicable. BMI=body-mass index.
Europe includes the Netherlands, Norway, Denmark, Belgium, the UK, Finland, Germany, Iceland, Ireland, and Switzerland.
Asia includes Japan, Taiwan, South Korea, Malaysia, Singapore, India, China, Jordan, and Bangladesh.
Other countries are Brazil, Tunisia, and South Africa.
The bias score assigned to each study ranged from 1 to 5. This value represents the mean (SD) of the score and the corresponding median (IQR).
Denominator is 117; four studies did not report the type of workplace. Workplace settings were classified into the following types: office, hospital, school, factory, mixed, or others. Mixed is defined as a combination of the prespecified workplaces, whereas others differ from the prespecified categories.
A mean of 1188 participants (SD 1739) were cluster randomised and 171 (183) participants were individually randomised.
Data on number of women were not extracted and therefore are not available.
When available, data on race or ethnicity usually referred to white people, and so this group was chosen to represent this variable in the meta-analyses.
Primary specified target indicates the most frequent, but not limited to, intervention focus per outcome identified subjectively by the investigators (diet quality, weight loss, physical activity, and other such as reduction in cardiovascular disease risk factors, smoking cessation, stress reduction, diabetes, or cancer prevention); for instance the primary specified target for the interventions reporting fruit intake was diet quality (93%), meaning diet quality was the most frequent target, but not exclusively because the intervention could also target physical activity or weight loss less frequently.