It is widely acknowledged that racial or ethnic discrimination can negatively affect a person’s health. But how can a scientist measure this impact? The treatment that a person encounters due to the color of their skin, their language, or their country of origin is likely a chronic stimulus, encountered over their entire life rather than during a discrete period of time. How that person perceives or reacts to discrimination may also vary widely from individual to individual – some may shrug it off or internalize the damage, some may grow angry and lash out. Wrapping one’s statistical arms around such a huge variable is nearly impossible.
One way around this problem is to locate a finite period of elevated discrimination against a particular group, and measure the impact of that event upon health. Diane Lauderdale, professor of epidemiology in the Department of Health Studies, found just such an event in the terror attacks of September 11, 2001, and the brief but intense harassment of Arab-Americans that followed. In her talk for the MacLean Center of Clinical Medical Ethics seminar series in January, Lauderdale detailed how she studied a link between post-9/11 discrimination and birth outcomes for a paper in the journal Demography (pdf).
While it might be hard to pin down the discrimination experienced by people of Arabic origin over the course of their lives in the United States, their life in the months after the attacks was undoubtedly more stressful. According to the American-Arab Anti-Discrimination Committee, more than 700 violent incidents were directed toward persons who were perceived to be Arab in the nine weeks after 9/11. But a massive media and government pushback likely limited the duration of this out-of-control hatred, with the director of the ADC commenting in December 2001 that “My impression is that we are rapidly returning to what one would unfortunately call a normal amount of hate crimes.”
Lauderdale chose to focus on pregnant women, who are particularly sensitive to stress. High levels of corticotropin-releasing hormone – a peptide increased by stress – can induce early labor, producing babies that are premature and/or underweight. Lauderdale hypothesized that pregnant Arab-American mothers might have given birth to more low birth weight babies in the six months following 9/11 than they had during the same months in the previous year.
One problem: while Lauderdale had access to the birth certificates of more than 1.5 million children born in California from 2000-2002, the certificates categorized race only by black, white, American Indian, Asian, and Other – no Arabic. Fortunately, previous work by Lauderdale and colleagues had developed an algorithm for predicting a person’s Arab origin using their first and last names. While the algorithm was admittedly imperfect, it was able to create enough of an enriched sample to conduct the comparison, Lauderdale said.
Her analysis found that the births of most groups (white mothers, black mothers, foreign born mothers) were unaffected by the events of 9/11, with virtually no difference in the risk of having a low birth weight baby between the two years. But for the 15,000 Californian women with Arabic names analyzed, there was a small but significant spike in low birth weight babies from October 2001 to March 2002. Children born to those mothers were 34 percent more likely to be underweight than babies born to Arabic mothers from October 2000 to March 2001.
To break the Arabic group down even further, Lauderdale looked at whether each child was given a traditionally Arabic name, potentially a sign of stronger ethnic identity. While the number of Arab names given to newborns did not change before and after 9/11, babies with “traditional” names were more than twice as likely to be born underweight in 01-02 compared to the same months in 00-01. Babies with more “American” names, on the other hand, were almost unperturbed, with only a 16% higher chance of being underweight at birth.
“That was really the surprising finding,” Lauderdale said. “The elevation in risk for the Arab-Americans was really entirely explained by the risk among the women who gave their babies ethnically distinctive given names. That’s consistent with the hypothesis that these are women who identify more strongly as Arab-Americans and may also over-represent women who dress so as to be identifiable or have other features that identify them as more distinctively Arab.”
One potential confound of the study may be factors unique to California – an attempt to replicate Lauderdale’s findings in Michigan, the state with the highest proportion of Arab-Americans, failed to find an effect of 9/11 on birth outcome. Lauderdale proposed that discrimination against Arab-Americans may be felt more acutely in the more-dispersed Arabic population of California than Michigan, where there are many predominantly Arab-American communities near Detroit – a reminder that even this discrete example of discrimination is hard to wall off from complicating factors. But by isolating the size of one specific health effect likely caused by the stress associated with a short period of elevated ethnic discrimination, Lauderdale’s research suggests the dramatic damage that could result from a lifetime of racism and discrimination.
Lauderdale, D. (2006). Birth Outcomes for Arabic-Named Women in California Before and After September 11 Demography, 43 (1), 185-201 DOI: 10.1353/dem.2006.0008
[Each academic year, the MacLean Center for Clinical Medical Ethics organizes a series of lunchtime seminars by physicians, biologists, economists, social scientists and other experts covering the biggest questions in health care and ethics. This year’s theme is “Health Disparities: Local, National, Global,” and the series was put together with the Urban Health Initiative, the Global Health Initiative, and Finding Answers. ScienceLife will carry regular coverage of this unique series, and video of the lectures will be posted when available.]