Skip to main content
. 2022 May 6;2022:7652408. doi: 10.1155/2022/7652408

Table 3.

Mean differences in 10-year weight gain across age categories in US women and men, after adjusting for the covariates.

Outcome Age categories F P
36–39 years 40–49 years 50–59 years 60–69 years 70–79 years
Women only (n = 7108)
 10-year weight gain (kg) 9.0a ± 0.7 7.7a ± 0.5 5.9b ± 0.6 3.3c ± 0.6 1.8d ± 0.5 30.2 <0.0001
 10-year weight gain (%) 14.5a ± 0.9 12.8a ± 0.7 10.1b ± 0.8 6.5c ± 0.8 3.7d ± 0.6 46.4 <0.0001
Men only (n = 6694)
 10-year weight gain (kg) 6.5a ± 0.7 5.0b ± 0.4 2.7c ± 0.4 0.9d ± 0.5 −1.1e ± 0.5 33.7 <0.0001
 10-year weight gain (%) 8.5a ± 0.7 6.8b ± 0.5 3.9c ± 0.4 1.7d ± 0.5 −0.6e ± 0.5 45.2 <0.0001
Combined (n = 13,802)
 10-year weight gain (kg) 7.8a ± 0.5 6.4b ± 0.3 4.3c ± 0.4 2.1d ± 0.4 0.3e ± 0.4 52.6 <0.0001
 10-year weight gain (%) 11.5a ± 0.6 9.8b ± 0.4 7.0c ± 0.4 4.1d ± 0.5 1.5e ± 0.4 76.2 <0.0001

10-year weight gain is the average amount of weight gained over the previous 10 years, expressed in kg or as a percentage of initial body weight. Means on the same row with different superscript letters (a,b,c,d,e) are significantly different (P < 0.05). SE represents the standard error of the mean. Means have been adjusted for differences in race for the women only and men only analyses, and sex and race when the analysis used the entire sample. The mean difference between 10-year weight gain (kg) for those in the 36–39 and 40–49 age categories, using the combined sample, was significant at the P=0.0854 level. Across the five categories of age, the sample size percentages (%) and sample size numbers (n), when combined, were 36–39 years (10.3%, n = 1347), 40–49 years (26.8%, n = 3415), 50–59 years (28.3%, n = 3425), 60–69 years (22.2%, n = 3578), and 70–79 years (12.4%, n = 2037). Because NHANES sample weights were applied to each individual, differences in the sample size of each category should be interpreted based on the percentages (%), not the raw number (n). The sample size percentages (%) can be generalized to the US adult population, but the sample numbers (n) cannot.