Socioeconomic Disparities In Obesity And Diabetes

So far, we have discussed overall trends. At every point in time, obesity and diabetes rates are higher in some populations than in others. Populations with higher rates generally include groups with less education, lower household income, blacks, and Native Americans, but there are some noticeable variations and interactions depending on whether we distinguish between moderate and severe obesity. For example, obesity as defined by a BMI of over 30 has been higher among men than women (although that gap has narrowed noticeably). However, severe obesity (BMI > 40) has been much higher among women for some time, and women also make up the vast majority (close to 85%) of bariatric surgery patients (15).

Table 1 shows descriptive prevalence statistics for obesity, severe obesity, and diagnosed diabetes for men and women and by socioeconomic characteristics. Individuals without health insurance are, on average, 10 yr younger than those with health insurance (partly an effect of Medicare), yet have somewhat higher obesity rates and much higher severe obesity rates. Or, put in another way, severely obese patients who have the highest need for treatment are less likely to have health insurance. Largely because of the age difference between the insured and the uninsured populations, diagnosed diabetes is less common among the uninsured. But once age and sociodemographic variables (but not obesity) are adjusted for, the rate of diagnosed diabetes is slightly higher in the uninsured than in the insured population. However, it is less than what would be expected given the differences in obesity rates, suggesting that there is a higher rate of undetected diabetes among the uninsured. There are several reasons for this. People without insurance are far less likely to have a usual source of care—nearly 50% of the nonelderly uninsured, compared with about 10% of the insured, reported no usual source of health care—and a usual source of care and a sustained patient-physician relationship are strongly associated with use of preventive care (16,17). Low-socioeconomic status (SES) patients are less likely to receive regular screening for common health risk factors. Using the third NHANES study (1988-1994) (8,9), Qi Zhang at the University of Chicago estimated that undiagnosed diabetes was twice as common among adults without a high school diploma than among those with higher levels of education (Zhang, personal communication).

Table 1

Socioeconomic Status and Prevalence of Obesity, Severe Obesity, and Diabetes

Table 1

Socioeconomic Status and Prevalence of Obesity, Severe Obesity, and Diabetes

Women

Men

Women

Men

Women

Men

All adults

22.4

23.4

3.4

1.8

7.7

8.4

By insurance

Uninsured (mean age = 37)

25.6

23.0

4.4

2.6

6.2

5.4

Insured (mean age = 47)

21.9

23.5

3.3

1.8

8.0

9.0

By annual household income

<$15,000

31.0

24.9

6.4

3.4

14.4

14.3

>$50,000

16.9

22.9

2.2

1.5

3.5

6.0

By education

Less than high school

30.8

27.0

5.4

2.4

15.5

11.6

Completed college

15.3

18.7

2.1

1.3

4.2

7.0

By race/ethnicity

Non-Hispanic white

19.9

22.9

2.9

1.8

7.1

8.0

Non-Hispanic black

36.8

28.6

7.5

2.9

11.5

11.1

Asian

6.6

7.7

0.2

0.1

3.6

7.4

Native American

32.2

29.7

5.8

2.5

13.2

12.4

Hispanic

26.0

25.2

3.3

2.2

7.9

8.0

Source: Authors' calculation based on BRFSS 2004

Source: Authors' calculation based on BRFSS 2004

For women, a strong income gradient exists in obesity prevalence: those with annual household income less than $15,000 had an obesity rate that was almost twice as high as that among women with household income above $50,000. There is no such effect for men—obesity rates are fairly similar throughout the income range, a fact that many authors have noted. However, the same is not true for severe obesity, where there is an income gradient for both men and women. As we saw in Fig. 3, diagnosed diabetes is more prevalent among the severely obese than the moderately obese. The prevalence of diabetes therefore displays an income gradient for men as well as for women. This gradient exists even after adjusting for age (retirees have lower incomes) and other sociodemographics.

The causal direction of these associations is unclear. Severe obesity can cause disabilities that prevent people from working, and obesity therefore becomes the cause of lower income. At least for women, obesity also has been implicated as the cause of marrying lower-income spouses, lower educational achievements, and lower wage rates. Alternatively, low income may be a cause of obesity, as hypothesized by several authors and discussed in the next section. It is also possible that other factors cause both low income and obesity. One of the most prominent factors is severe mental illness, especially psychotic disorders. Some of the newer antipsychotics are strongly associated with weight gain and hyperglycemia (18,19) and mental illness obviously limits earning capacity. However, we tried to calculate to what extent this group drives the income gradient and realized that mental illness is not the reason for the association between lower income and severe obesity or diabetes.

Table 2 BMI and Employment Status

BMI

% homemaker, women

% Working for pay, women

% Working for pay, men

<25

13.6

58.5

59.7

25-30

12.1

59.3

63.6

30-35

12.0

55.7

62.8

35-40

11.9

54.9

59.9

40+

10.7

46.7

52.0

Source: BRFSS 2002, work for pay excludes self-employment

Source: BRFSS 2002, work for pay excludes self-employment

For both men and women, education is a socioeconomic factor that is strongly associated with obesity and diabetes. Among women, the rates of both obesity and severe obesity are more than twice as high in the group with less than a high school education than in the group of college graduates. There are noticeable race differences, with Asians having much lower rates of obesity and diabetes, and blacks and Native Americans having the highest rates. These differences remain after adjusting for age, income, and education, except that after this adjustment, Asians are no longer at lower risks for diabetes. Table 1 also shows the high rates of severe obesity among black and Native American women.

For women, several other adverse socioeconomic outcomes are associated with obesity (and although causality runs both ways, it is thought that the main effect is from obesity to adverse social outcomes). The effects appear to be larger for white women than for black or Hispanic women. Obese women are less likely to be married, tend to have lower family income, and are more likely to live in poverty. Table 2 shows additional details for employment status. The heavier a woman is, the less likely she is to work for pay and also less likely to be a homemaker. Women in higher weight categories are increasingly likely to be unemployed or unable to work. For men, there is no similar gradient: the BMI 35-40 group has the same employment rate as the BMI less than 25 group. Only in the severe obesity range—i.e., with a BMI over 40—do we see an adverse employment effect for men. For that group, employment rates are about 10 percentage points lower than that for moderate obese men. The primary reason reported by severely obese men is "unable to work," which is 10 times more common among severely obese men than among moderately obese men.

So far, we have discussed existing disparities in recent years, but there are concerns that current obesity epidemic will increase gaps. Figures 4-6 show that average weight gain has been fairly similar across all sociodemographic groups. Even the more advantaged groups, such as those with higher education and the non-Hispanic white, have seen large increases in BMI.

However, the weight distribution is more skewed to the right among those with less education, African Americans, and women. If we focus on the heaviest individuals, there actually may be an increasing gap by race or education and a decreasing gap by gender. Compare Fig. 7, which shows the 80th percentile, with Fig. 6, which shows the mean. Even though men have a higher mean BMI and this gap has not narrowed much, at the 80th percentile, the story is quite different. Based on self-reported height and weight,

Fig. 4. Average body mass index by education. Source: Authors' calculation based on the BRFSS. Reprinted with permission from ref. 54.
Fig. 5. Average body mass index by race. Source: Authors' calculation based on the BRFSS.
Fig. 6. Average body mass index by gender. Source: Authors' calculation based on the BRFSS.
Fig. 7. Eightieth percentile body mass index by gender. Source: Authors' calculation based on the BRFSS.

women have been catching up very rapidly (Fig. 7). In fact, when measuring weight objectively, a BMI of over 30 has already become more common among women than men, a dramatic reversal of the situation 20 yr ago (6). Furthermore, as we saw in Table 1, severe obesity rates are already much higher among women than among men (the lines would cross if we plotted the 90th percentile).

My Life My Diet

My Life My Diet

I lost over 60 pounds and 4+ inches off my waist without pills, strenuous exercise, or any of the things that the diet experts tell you to do...and I did it in less than 4 months! If you have the desire, and can read through my e-book , then this is for you! I could have easily made it a lot more difficult, with stacks of information that people will never read, but why?

Get My Free Ebook


Post a comment