Our results show that SAD was lower in women than in men in the overall population, as well as in the majority of subgroups defined by age, race/ethnicity, birth place, BMI, and household income. For example, in each BMI category (normal weight, overweight, and obesity), women tended to have an approximate 1.3 cm lower mean SAD than men, and the sex difference in SAD did not change over BMI categories. In addition, women had a 0.7 cm, 1.3 cm, 1.1 cm, 0.6 cm, and 1.2 cm lower mean SAD than men in the groups aged < 30, 30 to < 40, 40 to < 50, 50 to < 60, and ≥ 60 years, respectively. SAD is a manifest measure of visceral adipose tissues. The sex difference in SAD may be attributed to the observation that women have more abdominal subcutaneous adipose tissue but less visceral adipose tissue compared to men.22,23
WC is a measure of abdominal obesity, and measures both subcutaneous and visceral adipose tissues.3 By examining results from diverse studies of WC after matching the age ranges,24−27 we found that WC was lower in women than in men and the change patterns in WC was similar to the patterns in SAD in this study for women and men. WC is widely used to define abdominal obesity with differential reference ranges/cut points for women and men (88 cm for women and 102 cm for men). Similar differential patterns in SAD across sex groups from our study provide evidence for consideration of sex-specific cut points of SAD when assessing the obesity-related health risk by use of SAD.
Although women have a lower mean WC than men, data from two prospective cohort studies of the Health Professionals Follow-up Study and the Nurses’ Health Study showed that WC predicted the adjusted relative risk of coronary heart disease in both women and men, and the correlation between WC and heart disease was even stronger in women than in men.28 A different study followed up half a million men and women aged 40–69 years in the United Kingdom and showed that women with bigger waists and waist-to-hip ratios faced a greater excess risk of heart attack than men who had a similar ‘apple shape’.29 SAD was thought to have stronger associations with cardio-metabolic disorders than WC.5,30 In our study, women tended to have a lower SAD than men, similarly to WC, in the overall population, as well as the subgroups determined by age, race, education, etc. However, it remains unclear whether there is a sex difference in the associations of SAD with cardio-metabolic risk, and further study is needed to examine if the association is stronger in women than in men as WC showed in the previous reports. The strong health risk associations of SAD and WC would indicate that more intensive screening for the risk of cardiovascular disease might help prevent the onset of disease in individuals with an apple shape, especially in women.
SAD increased with age, and the trends in men and women were similar with respect to age. The results can be partly explained by the reports on visceral fat change over age in the previous studies.23,31 The process of aging was associated with substantial fat redistribution among depots.32 Redistribution of fat from subcutaneous to visceral depots was observed from late middle age until the ninth decade of life. Our results also show that the mean SAD increase was faster among younger adults aged < 50 years compared to older ones aged > 50 years for both women and men (Fig. 3). The change in SAD over age in our study is similar to WC,23 implying that body fat (both subcutaneous and visceral fat) is mostly accumulated in the first half of the life even though it continues until an older age. Both SAD5 − 7 and WC33 have been associated with cardiometabolic risk which is the leading cause of death in the US and worldwide. The prevention of excess fat in the early life could play a critical role in preventing or delaying obesity-related cardiovascular risk for both women and men.
Although mean BMI and WC among women and men increase in trends for the last two decades, the trends have leveled off in recent years after 2010.34 In this study, increasing trends in SAD were not significant for both women and men in 2011–2016, indicating no overall SAD increase over time. This is consistent with trends in BMI and WC, implying that the increases in anthropometric measures of obesity reach a plateau. Numerous obesity prevention and education programs may play an important role in reducing the increasing trend.
In this study, socio-economic correlates of SAD were similar in women and men. Older age, higher BMI, born in the US, and lower household income were all associated with higher SAD in both women and men. However, race/ethnicity groups were differently associated with SAD. Compared to non-Hispanic whites, being Hispanic was associated with lower SAD in men but not in women; being non-Hispanic black was associated with higher SAD in women but not in men; being Asian was associated with lower SAD in both women and men. A study of subjects, including 66 African American, 72 Hispanic, and 47 white men and women, showed that middle-aged and older African-American men and women had lower visceral fat than Hispanic and white peers. The reports are controversial to our results about racial/ethnic difference in SAD that measures levels of visceral fat. We conducted an analysis using a nationally representative sample of 14,054 individuals, including 3,134 non-Hispanic blacks, 3,359 Hispanics, and 5,425 non-Hispanic whites, and therefore, our data had more power to reflect the difference in visceral fat among racial/ethnic groups.
NHNAES is a national survey of the US civilian non-institutionalized population using a complex stratified multi-stage sampling design. By incorporating into data analysis the features of NHANES design including sampling weights, selection probabilities, and geographic clustering, our results can reasonably be generalized to the entire US non-institutionalized population and the socioeconomic subpopulations of men and women. Another strength of our study is the large sample size that ensures the analysis power for robust and unbiased estimates when comparing the difference in SAD between women and men across socio-economic subgroups. There are several limitations in this study. NHANES is a cross-sectional survey, and the associations of socio-economic factors with SAD could not be interpreted as causal effects. SAD was measured by a two-arm sliding-beam caliper with the possibility of certain levels of measurement errors in SAD. However, trained examiners recorded 4 repeated SAD readings that could minimize the measurement errors. Certain other factors not considered in this study could confound the results, as we focused on sex difference in SAD and its correlates.