Table 4. Details of prospective studies included in the meta-analysis assessing the relationship of short sleep with incident diabetes.
| Study | Country | Sample size | Sex | Follow-up Duration | Exposure (short sleep /referent) | Outcome | Covariate adjustments * |
|---|---|---|---|---|---|---|---|
| Ayas et al [14] | USA | 70,026 | Women | 10 y | ≤ 5 h vs 8 h | Total diabetes | 1, 2, 9, 14, 18, 19, 21, 28, 34, 36, 37 |
| Beihl et al [17] | USA | 654 | Both | 5 y | ≤ 7 h vs 8 h | Type 2 diabetes | 1, 4, 8, 10, 14, 21, 33, 36 |
| Gangwisch et al [16] | USA | 8,992 | Both | 8-10 y | ≤ 5 h vs 7 h | Total diabetes | 1, 2, 9, 10, 13, 26, 28 |
| Hayashino et al [28] | Japan | 6,509 | Both | Median 4·2 y | < 6 h vs 6-7 h | Total diabetes | 1, 18, 21, 28, 29, 33, 36, 38 |
| Holliday et al (this study) | Australia | 180,891 | Both | Mean 2·3 y | < 6 h vs 7-8 h | Type 2 diabetes | 1, 2, 5, 10, 16, 22, 26, 28, 31, 33, 36 |
| Kita et al [29] | Japan | 3,570 | Both | 3-5 y | ≤ 5 h vs > 7 h | Total diabetes | 1, 2, 6, 10, 15, 23, 28, 30, 33, 34, 36, 41. Notes: effects reported separately for those with and without a family history of diabetes. |
| Mallon et al [18] | Sweden | 1,170 | Both | 12 y | ≤ 5 h vs > 5 h | Total diabetes | 1, 2, 9, 21, 25, 27, 36, 37. Notes: effects reported separately for men and women. |
| Von Ruesten et al [39] | Germany | 23,620 | Both | 9-13 y | < 6 h vs 7-8 h | Type 2 diabetes | 2, 3, 6, 7, 10, 11, 17, 20, 21, 24, 28, 33, 35, 36, 40 |
| Xu et al [15] | USA | 174,542 | Both | 7-10 y | < 5 h vs 7-8 h | Total diabetes | 1, 2, 5, 7, 10, 12, 13, 14, 26, 33, 36. Notes: Diabetes cases diagnosed within the first 3-4 y of follow-up were excluded. Estimate is from an analysis of non-smoking, nonobese, physically active, healthy individuals. |
| Yaggi et al [40] | USA | 1,139 | Men | 15-17 y | ≤ 5 h vs 7 h | Total diabetes | 1, 10, 21, 32, 36, 39 |
Covariate definitions: 1: age, 2: alcohol use, 3: antidepressant intake, 4: baseline glucose tolerance, 5: baseline health status, 6: BMI, 7: caffeine use, 8: clinic, 9: depression, 10: education, 11: employment status, 12: energy intake, 13: ethnicity, 14: family history of diabetes, 15: fasting plasma glucose, 16: health insurance, 17: health satisfaction, 18: high cholesterol, 19: hormone replacement, 20: hyperlipidaemia, 21: hypertension, 22: income, 23: job stress, 24: life satisfaction, 25: living arrangement, 26: marital status, 27: obesity, 28: physical activity, 29: potential diabetes history, 30: rate of sedentary work, 31: residential remoteness, 32: self-rated health, 33: sex, 34: shift work, 35: sleeping disorders, 36: smoking, 37: snoring, 38: study intervention, 39: waist circumference, 40: waist-to-hip ratio, 41: working hours per week.