Public Health Grand Rounds: Data shows disparities not always attributable to race

There are those who are quick to assert that African Americans are more likely to be overweight; that Hispanics are more likely to struggle with mental health issues.  And when the data, on the surface, appears to match the stereotype, confirmation bias can make it an easy story to perpetuate.

AnnDozier

Ann Dozier

But information collected in Monroe County over the past 17 years shows that researchers need to closely examine myriad contributing factors before declaring that race or ethnicity leads to an increased risk for certain behaviors or health conditions.

“We talk about racial disparities, but maybe there’s a lot more going on underneath that we need to be attentive to, rather than just labeling a difference due to race,” said Ann Dozier, RN, PhD, Associate Professor in Public Health Sciences.

Starting in 1997 and continuing in 2000, 2006, and 2012, the Monroe County Health Department began surveying a random sample of adults in Monroe County, and a deep dig into the results shows that over time a handful of factors are more strongly correlated than race or ethnicity when it comes to certain behaviors and health conditions.

Dozier, who spoke about the findings at a March seminar in Helen Wood Hall Auditorium, compared data from five of the categories: smoking, perceived health, mental health, obesity and overweight, and insurance coverage.

In addition to race or ethnicity, the data were adjusted for age, gender, education level, and whether the person was a city or suburban resident.  Once controlling for these other conditions, race and ethnicity became almost a complete non-factor.

Smoking

  • Raw data showed that, at various points from 1997 and 2012, both African Americans and Hispanics were significantly more likely to be smokers than whites.
  • However, after adjusting for the four factors listed above, Black race and Hispanic ethnicity were not shown to be indicator for whether a person was a smoker in the 2006 and 2012 studies.  (In 1997 and 2000, adjusted data showed that African Americans were actually less likely to be smokers than whites.)
  • Instead, significant indicators were age, education, and place of residence.  Those below the age of 65, those with less education, and city-dwellers were all significantly more likely to be smokers.

Perceived health

  • Again, raw data showed that both African Americans and Hispanics, at various times between 1997 and 2012, were more likely to self-report their health as “fair” or “poor” than whites were.
  • But after adjusting for the four factors, no significant difference between self-reported perceived health was shown between Whites and African Americans in 2006 and 2012.   African Americans did report worse perceived health at significant levels in 1997 and 2000.
  • Age, education, and place of residence were again seen as significant indicators.  Those above the age of 65, those with less education, and city-dwellers were all significantly more likely to report a lower score for their perceived health.

Mental health

  • Participants were asked “Have you experienced stress, depression, or problems with emotion on 14 or more of the last 30 days?”  Raw data showed that African Americans and Hispanics, at various times between 1997 and 2012, were more likely to answer “yes.”
  • After adjusting for the four factors, Black race and Hispanic ethnicity were rarely found to be significant indicators for poorer mental health.  In 1997, Hispanics were still significantly more likely to report mental health stressors than whites, while in 2012, blacks were significantly less likely to report than whites.
  • Age and education were significant indicators for poorer mental health.  Those below the age of 65 and those with less education were more likely to report mental health stressors.

Body Mass Index

  • BMI was broken into two groups: overweight and obese.  Both raw data and adjusted data from the overweight group showed that minority groups were no more likely to be overweight than whites.
  • For obesity, adjusted data showed African Americans were more likely to be obese than whites in 1997, 2000, and 2006.  By 2012, adjusted data showed no significant correlation.
  • Significant indicators were gender for the overweight group and education for the obese group.  Men were more likely to be overweight, and those with less education were more likely to be obese.

Insurance Coverage

  • Once again, raw data showed that both African Americans and Hispanics, at various times between 1997 and 2012, were less likely to have health insurance than whites.
  • However, adjusted data showed no connection between Black race or Hispanic ethnicity and insurance coverage, except for at one point: in 2000, African Americans were less likely to have coverage.
  • Significant indicators were gender, age, and place of residence.  Men, those below the age of 65, and city-dwellers were less likely to have health insurance.

Conclusions

Dozier said that the analyses did have some weaknesses.  For one, all the data points are self-reported.

“So think about weight.  Are people being honest about how heavy they are?  That could actually be understated,” she said.

Researchers also weren’t able to adjust for other factors that might potentially contribute to these categories, such as marital status or income.

But conclusions can still be drawn from these findings.  Less education is a factor in key health behaviors.  Age is a factor in smoking, mental health reporting, and health insurance coverage.  Men are more likely to be overweight and less likely to have insurance.  And city-dwellers are more likely to smoke, and less likely to have insurance.

For each of these categories, race and ethnicity isn’t the end of the story.

“Disparities may have a component that may be attributable to race or ethnicity,” said Dozier.  “But we should be cautious about overstating what that relationship is, and we should be looking at them in the context of potential confounding factors.”

 

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s