Sleep Problems associated with Race/Ethnicity, Income, Education, Marital Status and Employment Status

In our recent paper, entitled "Who Gets the Best Sleep? Ethnic and Socioeconomic Factors Related to Sleep Complaints" published in the journal Sleep Medicine, we performed a set of analyses on some data that was gathered by the CDC in 2006 on over 150,000 people all across the USA.

These people were asked, "In the past 2 weeks, how many nights did you have trouble falling asleep, staying asleep, or sleeping too much?" The wording of this question is important, because it is not asking for any particular sleep disorder (like insomnia or sleep apnea), but rather a general sleep complaint. Also, it addresses issues that can exist in any sleep disorder, and involves "sleeping too much" which opens the possibility that it is not just measuring insufficient sleep. So this is a very broad question. When we looked at the data, we quickly realized that most people reported values at the very low end or very high end of the scale. Because of this, we re-coded the data so that those whose values were less than 6 were rated as not having a complaint and those with values 6 or higher were rated as having a complaint. This translates to an average of 3 nights per week across 2 weeks.

Next, we realized that this is a very large and diverse group of people. We wanted to know if there was any group of people that was more or less likely to report problems with their sleep. So we chose a number of variables, including Race/Ethnicity (White, Black/African-American, Hispanic/Latino, and Asian/Other), Income (<$10k, $10-15k, $15-20k, $20-25k, $25-35k, $35-50k, $50-75k, and $75k+), Education (Did not finish high school, High school graduate, Some College, and College graduate), Marital Status (Married, Never Married, In a relationship, Divorced, Widowed, and Separated), and Employment Status (Employed, Self-employed, Retired, Student, Homemaker, Unemployed <1 year, Unemployed >1 year, and Unable to work). We analyzed the data separately for men and women.

To perform our analyses, we used a technique called logistic regression, which computes the odds ration (OR) that a certain outcome will occur. The "outcome" in question was having sleep problems. So we created an equation (actually 2 equations -- one for each gender) that included age (to control for age), as well as the variables mentioned above. The equation computes the OR simultaneously for all of the levels of all of the variables, so they all control for each other. So, for example, when we report the OR for being Married, it is controlling for gender (separate analysis), age, race/ethnicity, income, education, and employment all at the same time. So we know the effects are not due to any of those other factors, which makes them much more reliable. Also, for this type of analysis, when you have multiple categories (like in Race/Ethnicity), one category is set to be the "reference" and all others are compared to it.

What we found was surprising.

For Race/Ethnicity (reference = White): For women, if you were Black/African-American or Hispanic/Latina, you were less likely to complain (26% less likely than White in both cases), and if you were Asian/Other, you were much less likely to complain (58% less likely than White), but if you were Multiracial, you were more likely to complain (67% more likely). For men, there were no differences among groups (none was different from White) except for men who were Asian/Other, who were 57% less likely to complain of sleep problems than White men.

For Education (reference = College Graduate): College graduates slept best. For both men and women, the more education you had, the less likely you were to report sleep problems. In men, those with some college were 22% more likely to complain, those who graduated high school were 27% more likely to complain, and those that did not complete high school were 35% more likely to complain (all of these were significantly different from college graduates). For women, those with some college were 30% more likely to complain, those who graduated high school were 31% more likely to complain, and those that did not complete high school were 64% more likely to complain (again, all of these were significantly different from college graduates).

For Marital Status (reference = Married): Married people slept best. Those that slept worse were those that were never married (62% more likely for women and 74% more likely for men) or in a relationship but not married (39% more likely for women and 102% more likely for men). Men who were divorced or separated were more likely to report problems (52% more likely for divorced and 43% more likely for separated) and this increase was greater than in women (divorced women 24% more likely and separated women 15% more likely). Widowed men and women did not sleep better than married, after controlling for all of the other variables.

For Employment (reference = Employed): Employed people slept best. If you were self-employed, you were 21% (women) or 25% (men) more likely to report sleep problems. If you were retired and male, you were 22% more likely to report problems (no difference in women). If you were a student, there was such a wide range of responses that it was impossible to tell if they were different (some were less likely to report problems and some were almost twice as likely). If you were unemployed <1 year, you were 97% more likely to report problems if you were a woman and 191% more likely if you were a man. If you were unemployed >1 year, you were 122% more likely if you were a woman and 197% more likely if you were a man. If you were unable to work, you were 319% more likely to report problems as a woman and 454% more likely as a man. Although these patterns were similar for men and women, there was a large difference for homemakers. Women homemakers were 18% more likely to report problems (similar to self-employed), but male homemakers were 244% more likely to report sleep problems (similar to unemployed).


For Income (reference $75k+):  For women, all categories reported significantly more sleep problems than those making the most money. Generally, the less money, the more complaint (15% more likely for $50-75k, 32% for $35-50k, 39% for $25-35k, 58% for $20-25k, 75% for $15-20k, 84% for $10-15k, and 52% for <$10k). For men, the pattern was a little different. Although the least amount of complaint was still in the highest-earning group, only those making <$25k were significantly different with those making $20-25k being 45% more likely to complain, $15-20k 53% more likely, $10-15k 88% more likely and <$10k 47% more likely.

We know that there is a very strong relationship between sleep and health. Overall, these results show that social and economic factors can play an important role in that relationship.

For more, see the complete paper at http://www.michaelgrandner.com/pages/research-publications.html.