Table 3.
Pooled, Multivariable-Adjusted Results for the Relationships between Changes in Lifestyle Habits and Weight Change.*
| Lifestyle Habit | Weight Change Every 4 Years (95% Confidence Interval) | |||
|---|---|---|---|---|
| Age-Adjusted Change | P Value | Multivariable-Adjusted Change† | P Value | |
| lb | lb | |||
| Dietary change, median score‡
| ||||
| Quintile 1: 42 | 3.97 (2.85 to 5.10) | <0.001 | 3.93 (2.88 to 4.98) | <0.001 |
|
| ||||
| Quintile 2: 47 | 2.32 (1.99 to 2.66) | <0.001 | 2.41 (1.96 to 2.87) | <0.001 |
|
| ||||
| Quintile 3: 50 | 2.08 (1.35 to 2.80) | <0.001 | 2.10 (1.37 to 2.83) | <0.001 |
|
| ||||
| Quintile 4: 54 | 1.64 (1.22 to 2.06) | <0.001 | 1.64 (1.22 to 2.05) | <0.001 |
|
| ||||
| Quintile 5: 60 | Reference | Reference | ||
|
| ||||
| Physical-activity change, median change in MET-hr/wk
| ||||
| Quintile 1: −16.3 | Reference | Reference | ||
|
| ||||
| Quintile 2: −2.59 | −0.07 (−0.36 to 0.22) | 0.64 | −0.07 (−0.32 to 0.18) | 0.60 |
|
| ||||
| Quintile 3: 1.59 | −0.85 (−1.11 to −0.59) | <0.001 | −0.81 (−1.02 to −0.60) | <0.001 |
|
| ||||
| Quintile 4: 6.49 | −1.00 (−1.10 to −0.91) | <0.001 | −0.92 (−1.02 to −0.83) | <0.001 |
|
| ||||
| Quintile 5: 23.2 | −1.86 (−2.31 to −1.41) | <0.001 | −1.76 (−2.14 to −1.38) | <0.001 |
|
| ||||
| Increase in no. of alcohol drinks/day§ | 0.39 (0.23 to 0.55) | <0.001 | 0.41 (0.23 to 0.59) | <0.001 |
|
| ||||
| Total daily hr of sleep¶
| ||||
| <6 | Reference | Reference | ||
|
| ||||
| 6 to 7 | −0.36 (−0.43 to −0.29) | <0.001 | −0.32 (−0.38 to −0.25) | <0.001 |
|
| ||||
| >7 to 8 | −0.35 (−0.48 to −0.22) | <0.001 | −0.30 (−0.42 to −0.18) | <0.001 |
|
| ||||
| >8 | −0.09 (−1.11 to 0.92) | 0.86 | −0.09 (−0.96 to 0.78) | 0.84 |
|
| ||||
| Missing data | −0.49 (−0.93 to −0.06) | 0.03 | −0.41 (−0.72 to −0.10) | 0.01 |
|
| ||||
| Increase in hr/day watching television|| | 0.32 (0.13 to 0.51) | 0.001 | 0.31 (0.20 to 0.42) | <0.001 |
|
| ||||
| Change in smoking status
| ||||
| Never smoked, no change | Reference | Reference | ||
|
| ||||
| Former smoker, no change | 0.19 (0.13 to 0.25) | <0.001 | 0.14 (0.08 to 0.20) | <0.001 |
|
| ||||
| Current smoker, no change | −0.62 (−0.97 to −0.26) | <0.001 | −0.70 (−1.10 to −0.31) | <0.001 |
|
| ||||
| Current smoker, changed to former smoker | 4.99 (3.89 to 6.09) | <0.001 | 5.17 (4.06 to 6.29) | <0.001 |
|
| ||||
| Former smoker, changed to current smoker | −2.47 (−3.82 to −1.12) | <0.001 | −2.81 (−4.24 to −1.38) | <0.001 |
|
| ||||
| Never smoked, changed to current smoker | 0.67 (−1.48 to 2.82) | 0.02 | 0.32 (−2.31 to 2.95) | 0.81 |
Data are based on a 20-year follow-up (1986–2006) in the Nurses’ Health Study (NHS), a 12-year follow-up (1991–2003) in the Nurses’ Health Study II (NHS II), and a 20-year follow-up (1986–2006) in the Health Professionals Follow-up Study (HPFS). Findings within each study and restricted to persons who never smoked were generally similar to these pooled results (see the tables in the Supplementary Appendix). To convert pounds to kilograms, divide by 0.45.
Values were adjusted for age, baseline body-mass index at the beginning of each 4-year period, and all the variables shown in the table simultaneously.
Changes in dietary habits associated with weight change were derived by totaling the ordinal values for the quintiles of change for each dietary habit in ascending order (1 to 5) or descending order (5 to 1) for habits that were inversely or positively associated with weight gain, respectively. Scores ranged from 17 to 85.
Heterogeneity was evident across beverage types and changes (see the Supplementary Appendix for details).
Data are based on absolute levels, owing to limited data on serial assessments of sleep for each cohort to assess change in each 4-year period.
Differences in weight change are for increased television watching; the inverse difference would be associated with decreased watching. Values are based on data from the NHS II and the HPFS; limited serial assessments precluded an analysis of changes in television watching in the NHS.