Table 1.
Characteristics of the included studies.
| Reference | Country | Study Design and Follow-Up Period | Participants | Exposure vs. Comparison, Measurement | Outcome, Measurement | Results | |
|---|---|---|---|---|---|---|---|
|
n Total n Cases * |
Sex ** Age *** |
Estimated Effect Sizes (95%CI LL; UL) |
|||||
| Goto et al., 2008 [36] | Japan | retrospective, check-up data, 2000–2007 | 4634 598 |
100% 21.5 |
Skipping ≥ 2 vs. ≤ 1 d/wk, self-administered questionnaire |
>5% increased BMI, weight and height measurement |
OR = 1.34 (1.12; 1.62) |
| Guinter et al., 2020 [90] | USA, Puerto Rico | Sisters Study, prospective cohort, 2003–2015 | 46,037 2797, 2383, 6807 |
0% 55.3 |
Eating 3–4 d/wk vs. 0, 1–2, 5–6, 7 d/wk, FFQ |
5-year incident BMI ≥ 25 kg/2, ≥ 30 kg/2, ≥ 5 kg weight gain, weight and height measurement and self-reported weight |
5-yr incident BMI ≥ 25kg/2: 0 d/wk RR = 0.74 (0.62; 0.89), 1–2 d/wk RR = 0.91 (0.78; 1.07), 5–6 d/wk RR = 0.97 (0.85; 1.09), 7 d/wk RR = 0.88 (0.78; 0.99) 5-yr incident BMI ≥ 30kg/2: 0 d/wk RR = 0.72 (0.59; 0.87), 1–2 d/wk RR = 0.75 (0.62; 0.89), 5–6 d/wk RR = 0.91 (0.80; 1.04), 7 d/wk RR = 0.79 (0.70; 0.90) 5-yr incident ≥ 5 kg weight gain: 0 d/wk RR = 1.00 (0.90; 1.11), 1–2 d/wk RR = 0.98 (0.89; 1.08), 5–6 d/wk RR = 0.99 (0.92; 1.06), 7 d/wk RR = 0.97 (0.91; 1.04) |
| Hurst and Fukuda, 2018 [37] | Japan | secondary analysis of insurance and health check-up data, 2008–2013 | 59,717 20,671 |
66% 47.4 |
Skipping ≤ 2 vs. ≥ 3 d/wk, Health check-up question |
BMI ≥ 25 kg/2, BMI and WC change, BMI and WC data from check-up |
BMI ≥ 25kg/2: OR = 0.92 (0.87; 0.97) BMI change (in kg/2): β = 0.00 (−0.03; 0.04) WC change (in cm): β = 0.03 (−0.11; 0.16) |
| Kahleova et al., 2017 [93] | North America Canada | AHS-2, prospective cohort, 2002–2010 | 50,660 n.g. ++ |
36% 58 |
Eating vs. skipping, Hospital History Form |
BMI change/year, weight and height measurement and self-report |
BMI change (in kg/2): β = −0.03 (−0.04; -0.01) |
| Kito et al., 2019 [38] | Japan | retrospective cohort, 2008/09–2012 | 45,524 5093 |
100% 34 |
Skipping ≥ 3 vs. ≤ 2 d/wk, Health check-up question |
BMI ≥ 25 kg/2, weight and height measurement |
OR = 1.18 (1.04; 1.33) |
| Nooyens et al., 2005 [92] | The Netherlands | Doetinchem Cohort Study, prospective, 1987–2002 | 288 n.g. ++ |
100% 54.9 |
Eating 0–7 d/wk Dutch version of EPIC FFQ |
Weight and WC change/year, weight, height and WC measurement |
Weight change (in kg): β = 0.04 (n.g. ++) WC change (in cm): β = 0.10 (n.g. ++) |
| Odegaard et al., 2013 [91] | USA | CARDIA Study, prospective cohort, 1992/93–2011 | 3598 972 WC 783 BMI |
44% 32.1 |
Eating ≤ 3 vs. 4–6, 7 d/wk, interviewer-administered CARDIA DHQ |
BMI ≥ 30 kg/2, WC > 88 cm for women and > 102 cm for men, weight, height and WC measurement |
BMI ≥ 30 kg/2: 4–6 d/wk HR = 0.75 (0.62; 0.90), 7 d/wk HR = 0.57 (0.47; 0.68) WC > 88 or 120 cm: 4–6 d/wk HR = 0.84 (0.70; 0.99), 7 d/wk HR = 0.78 (0.66; 0.91) |
| Smith et al., 2017 [39] | Australia | CDAH Study, prospective cohort, baseline 2002/04–2011 | 1155 410 |
43% 31.5 |
Met guidelines # consistently vs. met not, postal questionnaire |
5-year weight change, weight and height measurement and self-report |
5-yr weight change (in kg): β = 1.5 (0.5; 2.8) |
| van der Heijden et al., 2007 [94] | USA | HPFS, prospective cohort, 1992–2002 | 20,064 5857 |
100% 57.3 |
Eating vs. skipping, semi-quantitative FFQ |
≥ 5 kg weight gain, self-reported weight |
HR = 0.87 (0.82; 0.93) |
* participants, in which the analyzed outcome occurred; ** male sex in percentage; *** mean age in years; 95% CI LL; UL = 95% confidence interval: lower limit; upper limit; d/wk = days per week; ++ information not given; AHS = Adventist Health Study; CARDIA = Coronary Artery Risk Development in Young Adults; CDAH = Childhood Determinants of Adult Health; HPFS = Health Professionals Follow-Up Study; FFQ = Food Frequency Questionnaire; EPIC = European Prospective Investigation into Cancer and Nutrition; BMI = body mass index; WC = waist circumference, kg = kilograms; m2 = square meters; yr = year; OR = odds ratio; RR = relative risk; HR = hazard ratio; # referring to the 2010 Dietary Guidelines for Americans: eating a nutrient-dent breakfast vs. not eating breakfast [35].