Table 3.
Multivariable adjusted odds ratios (95% CI) for obesity per increment of 10 risk alleles by frequency of fried food consumption*
| Consumption/week | P for interaction | |||
|---|---|---|---|---|
| <1 | 1-3 | ≥ | ||
| Total fried food consumption | ||||
| Nurses’ Health Study: | ||||
| No of women (obese/normal weight) | 679/2888 | 484/1666 | 356/776 | — |
| Odds ratio (95% CI )† | 1.76 (1.39 to 2.24) | 2.70 (2.00 to 3.65) | 2.54 (1.75 to 3.69) | 0.02 |
| Health Professionals Follow-up Study: | ||||
| No of men (obese/normal weight) | 194/1051 | 216/830 | 269/686 | — |
| Odds ratio (95% CI)† | 2.00 (1.27 to 3.13) | 2.04 (1.31 to 3.20) | 2.86 (1.88 to 4.36) | 0.16 |
| Women’s Genome Health Study: | ||||
| No of women (obese/normal weight) | 743/11673 | 266/3423 | 120/1234 | — |
| Odds ratio (95% CI)‡ | 1.46 (1.19 to 1.78) | 1.57 (1.11 to 2.21) | 2.88 (1.68 to 4.94) | 0.06 |
| Pooled odds ratio (95% CI)§ | 1.61 (1.40 to 1.87) | 2.12 (1.73 to 2.59) | 2.72 (2.12 to 3.48) | 0.002 |
| Fried food consumed at home | ||||
| Nurses’ Health Study: | ||||
| No of women (obese/normal weight) | 808/3181 | 567/1840 | 138/299 | — |
| Odds ratio (95% CI)† | 1.73 (0.92 to 3.27) | 2.46 (1.86 to 3.24) | 4.52 (2.33 to 8.77) | 0.006 |
| Health Professionals Follow-up Study: | ||||
| No of men (obese/normal weight) | 302/1396 | 303/973 | 71/188 | — |
| Odds ratio (95% CI)† | 2.12 (1.47 to 3.05) | 2.57 (1.77 to 3.74) | 1.96 (0.79 to 4.90) | 0.69 |
| Women’s Genome Health Study: | ||||
| No of women (obese/normal weight) | 939/13404 | 208/2778 | 17/211 | — |
| Odds ratio (95% CI)‡ | 1.44 (1.21 to 1.73) | 2.25(1.51 to 3.35) | 2.83 (0.55 to 14.52) | 0.06 |
| Pooled odds ratio (95% CI)§ | 1.57 (1.34 to 1.83) | 2.43 (2.00 to 2.96) | 3.33 (2.00 to 5.55) | 0.003 |
| Fried food consumed away from home | ||||
| Nurses’ Health Study: | ||||
| No of women (obese/normal weight) | 1110/4485 | 375/805 | 33/34 | — |
| Odds ratio (95% CI)† | 2.09 (1.73 to 2.52) | 2.68 (1.86 to 3.85)¶ | 0.28 | |
| Health Professionals Follow-up Study: | ||||
| No of men (obese/normal weight) | 351/1647 | 257/810 | 69/104 | — |
| Odds ratio (95% CI)† | 2.08 (1.49 to 2.91) | 2.53 (1.69 to 3.79) | 4.39 (1.26 to 15.25) | 0.30 |
| Women’s Genome Health Study: | ||||
| No of women (obese/normal weight) | 864/13750 | 263/2611 | 11/127 | — |
| Odds ratios (95% CI)‡ | 1.52 (1.27 to 1.83) | 1.84 (1.30 to 2.61) | 5.19 (0.45 to 59.96) | 0.17 |
| Pooled odds ratio (95% CI)§ | 1.81 (1.60 to 2.05) | 2.29 (1.84 to 2.83) | 4.53 (1.49 to 13.79) | 0.02 |
*Derived from logistic regression analyses, using data on fried food consumption assessed at baseline and obesity status assessed four years later in Nurses’ Health Study and Health Professionals Follow-Up Study and three years later in Women’s Genome Health Study.
†Data adjusted for age, source of genotyping data, physical activity, television watching, smoking, alcohol intake, sugar sweetened beverage intake, alternative healthy eating index, and total energy intake.
‡Data adjusted for age, physical activity, smoking, alcohol intake, sugar sweetened beverage intake, alternative healthy eating index, and total energy intake.
§Results for three cohorts pooled by means of fixed effects meta-analyses (if P≥0.05 for heterogeneity between studies) or random effects meta-analyses (if P<0.05 for heterogeneity between studies).
¶For fried food consumed away from home in Nurses’ Health Study, participants in categories of 1-3/week and ≥4/week combined because of small sample size in category of ≥4/week.