Table 2.
Multiple linear regression models assessing the relationship between meal and sleep timing variables and body mass index: cross-sectional analysis (GCAT study 2018)
| Exposures1 | ||||||
|---|---|---|---|---|---|---|
| Time of first meal | Fasting hours | Eating occasions | Sleep dissatisfaction | Sleep time | ||
| Overall (n = 7074) | ||||||
| Model 02 | 0.06 [0.00, 0.12] | 0.00 [-0.06, 0.06] | 0.16 [-0.02, 0.33] | |||
| Model 13 | 0.32 [0.18, 0.47] | -0.27 [-0.41, -0.13] | 0.14 [-0.05, 0.32] | |||
| Model 24 | 0.24 [0.08, 0.40] | -0.19 [-0.35, -0.04] | 0.15 [-0.04, 0.33] | 0.19 [-0.11, 0.50] | 0.16 [0.02, 0.29] | |
| Men (n = 2908) | ||||||
| Model 13 | 0.23 [0.04, 0.43] | -0.18 [-0.37, 0.01] | -0.01 [-0.27, 0.26] | |||
| Model 24 | 0.18 [-0.03, 0.4] | -0.14 [-0.34, 0.07] | 0.00 [-0.26, 0.27] | 0.39 [-0.06, 0.84] | 0.1 [-0.09, 0.28] | |
| Women (n = 4166) | ||||||
| Model 13 | 0.39 [0.19, 0.59] | -0.33 [-0.53, -0.13] | 0.24 [-0.01, 0.49] | |||
| Model 24 | 0.28 [0.05, 0.50] | -0.24 [-0.45, -0.02] | 0.25 [0.00, 0.51] | 0.10 [-0.30, 0.50] | 0.20 [0.01, 0.39] | |
| Premenopausal women (n = 1693) | ||||||
| Model 13 | 0.58 [0.25, 0.90] | -0.59 [-0.91, -0.28] | -0.09 [-0.50, 0.33] | |||
| Model 24 | 0.40 [0.04, 0.77] | -0.45 [-0.80, -0.11] | -0.07 [-0.49, 0.34] | 0.77 [0.09, 1.45] | 0.31 [-0.01, 0.62] | |
| Postmenopausal women (n = 2473) | ||||||
| Model 13 | 0.25 [-0.01, 0.51] | -0.12 [-0.37, 0.13] | 0.47 [0.16, 0.78] | |||
| Model 24 | 0.18 [-0.1, 0.47] | -0.06 [-0.34, 0.21] | 0.48 [0.18, 0.80] | -0.25 [-0.75, 0.25] | 0.13 [-0.11, 0.37] | |
N: sample size. Bold: statistically significant
1. Values are regression coefficients [95% confidence interval]. Coefficients should be interpreted for 1-hour increase for each time exposure, for eating occasions it should be interpreted as an increase in 1 eating occasion, for sleep dissatisfaction, it should be interpreted as a higher degree of dissatisfaction. Outcome: BMI. Multicollinearity test was performed, all GVIF < 10.
2. Model 0: Meal timing variables not mutually adjusted, i.e. three separate models. Adjustment variables: age, sex, adherence to Mediterranean diet, education, employment, mental health, smoking habit, physical activity, time of first meal, fasting hours, number of eating occasions, and energy intake
3. Model 1 same adjustment variables. Meal timing variables mutually adjusted in one sole model
4. Model 2 includes all the variables in Model 1 + sleep dissatisfaction and sleep time