Differences in health behaviors and health outcomes among non-Hispanic Whites and Arab Americans in a population-based survey in California

Arab Americans are an understudied minority population in the United States (US) for whom health research is slowly and steadily increasing [1]. It is estimated that there are approximately 3.7 million Arab Americans in the United States [2]. California has the largest number of Arab Americans while Michigan has the highest proportion of Arab Americans in the population [2]. The majority of studies examining Arab American health in the US do so in Michigan in the context of an ethnic enclave (Dearborn, Michigan) due to ease of identification and recruitment [1]. There is reason to believe, however, that Arab Americans living in ethnic enclaves differ from those living outside of them [3]. Very few population-based studies [4‐7] allow the examination of the potential bias that may occur when categorizing Arab Americans as non-Hispanic Whites in health studies outside of Michigan due to difficulties in identifying Arab Americans.

Arab Americans lack an ethnic/racial identifier due to Office of Management and Budget standards created in 1997 [8]. Persons who classify themselves as Arab American are told to self-identify as White, and in some cases, like in the case of the Census, are recoded from Other to White based on an indication of origin from the Middle East and North Africa [9]. Investigators interested in Arab American health are therefore unable to isolate Arab populations despite evidence that Arab Americans have different health behaviors, different levels of healthcare access, and experience more cultural stigma that affects health when compared to the general non-Hispanic White population [1, 10]. Additionally, the social pressures this group faces, including an increased risk of hate crimes [11] and other instances of racism, influence health in this population.

We aimed to better understand how the health of Arab Americans differs from non-Hispanic Whites in California by: 1) estimating the prevalence of various health outcomes and behaviors among Arab Americans and 2) examining the association between race/ethnicity and these health outcomes and behaviors while adjusting for confounders.

A total of 1359 Arab Americans and 192,868 non-Hispanic Whites were identified from the CHIS between 2003 and 2016. More male Arab Americans were identified than non-Hispanic Whites (54.9 vs. 49.0%, p < 0.05), and the age distribution of the two groups differed with more Arab Americans aged 30–49 (42.2 vs. 33.8% among non-Hispanic Whites, p < 0.001) and more non-Hispanic Whites aged 50 or over (49.9 vs. 25.5% among Arab Americans, p < 0.001). More Arab Americans were born outside the United States (65.1 vs. 8.7%, p < 0.001) and fewer Arab Americans were citizens when compared to non-Hispanic Whites (84.9 vs. 90.6%, p < 0.05). Table 1 compares demographic, socioeconomic, and health related factors between Arab Americans and non-Hispanic Whites.

The goals of this study were to provide estimates for the prevalence of health behaviors and outcomes among Arab Americans in California and to examine differences with non-Hispanic Whites. Overall, Arab Americans reported more positive health behaviors and health outcomes than non-Hispanic Whites in our CHIS sample. Arab Americans had decreased odds of reporting alcohol use (including binge drinking), ever contemplating suicide, and hypertension than non-Hispanic Whites. The only outcome Arab Americans had increased odds of self-reporting was diabetes.

Previous studies on alcohol use and misuse among Arab Americans in national and Michigan-based surveys have also found that Arab Americans are less likely to report alcohol use than non-Hispanic Whites [17]. Binge drinking in the past month was found to be lower among Arab Americans identified through the National Survey on Drug Use (10%) than in the Michigan Behavioral Risk Factor Survey (17%) [17]. Arab Americans immigrate to the United States from countries with low alcohol consumption rates due to the religious prohibition and social discouragement of drinking in many origin countries [18, 19]. Some theories exist as to why Arab Americans may begin using alcohol including social change and acculturation [20], trauma from origin countries [21], and the larger composition of Christian Arab Americans than Muslim Arab Americans [2], but alcohol consumption remains low in this population.

We found that Arab Americans are less likely to have ever contemplated suicide than non-Hispanic Whites. Only one other study, based in Michigan, has examined suicide among Arab Americans. That study found that Arab Americans were less likely to die of suicide than non-Arab Whites in Michigan between 1990 and 2007 [22]. Studies among ethnic minority groups suggest that Arab Americans may be at lower risk for suicide and suicidal ideation due to socially-oriented norms like communalism and strong family bonds [23]. Positive ethnic group identity and affective expression may also be important mechanisms through which the risk of suicide is reduced for these populations [23].

We found higher diabetes (8.6 vs. 7.9%), lower hypertension (16.0 vs. 29.7%), and lower heart disease prevalence (3.9 vs. 8.2%) among Arab Americans than in non-Hispanic Whites. Arab Americans had increased odds of self-reporting diabetes and lower odds of self-reporting hypertension than non-Hispanic Whites in our sample. Diabetes incidence and prevalence have been shown to be higher among immigrant populations in the US [24]. Additionally, diabetes prevalence in the Middle East is high [25] and Arab immigrants to other countries have been found to have a high burden of diabetes [26]. The pattern found in this paper, higher levels of diabetes but not hypertension or obesity, is in congruence with evidence from the National Health Interview Survey in which Middle Eastern immigrants had higher rates of diabetes but not hypertension or obesity than European immigrants to the US [27]. Some reasons why Arab Americans may be at increased risk for diabetes include changes to diet associated with acculturation [28] and the potential for lower access and under-utilization of health care for this minority population [29].

Disease prevalence in our sample differs from that examined in national health surveys and Michigan based surveys. The National Health Interview Survey found a lower diabetes prevalence among Arab Americans (4.8%, identified through place of birth) and non-Hispanic Whites (6.9%) than in our sample [15]. Hypertension prevalence was also higher in our sample than in the National Health Interview Survey for both Arab Americans (13.4%) and non-Hispanic Whites (24.5%) [15], while heart disease prevalence in our sample was lower than that reported in the Michigan Behavioral Risk Factor Survey for Arab Americans (8.8%) and non-Hispanic Whites (8.1%) [4]. The differences in chronic disease prevalence between Michigan and California is of particular concern given the reliance on Arab American data coming out of Michigan. If the health behaviors and health outcomes differ for Arab Americans in contexts outside of Michigan, this could have large ramifications for the overall knowledge of Arab American health risks nationally. Some potential reasons for the differences in chronic disease prevalence between the samples may include the methods of recruiting and isolating Arab Americans, differences in community context and acculturation, and differences in length of stay in the US of the Arab Americans sampled. In our study, individuals did not self-identify as Arab American but were isolated through the use of country of birth and language spoken at home. Studies in Michigan have traditionally allowed for self-identification of Arab American identity. These differences may result in different samples with different characteristics, including sociodemographic and immigration level variables. Additionally, differences in socioeconomic status and opportunity between California and Michigan may be driving differences in results. Although data is limited, there is some evidence to suggest that Arab Americans in California are wealthier than those living in Michigan [30] despite the poor socioeconomic indicators reported by Arab Americans in our sample. Additionally, the ethnic community context in which people live likely plays a large factor in the available nutritional and social activities that may influence these outcomes. The concentration of Arab Americans in Michigan is higher than that in California and may therefore drive ethnic community context [31]. Finally, some work has shown that length of residence in the United States is directly correlated with obesity and chronic disease risk [32] and length of stay in the US may differ from state to state.

There are some limitations to our analysis. First, the use of random digit dialing for recruitment may not be yielding a representative sample of Arab Americans. The Arab American Institute estimates approximately 817,455 Arab Americans living in California (a state with an adult population of 34.4 million). We estimate therefore that approximately 2% of the total adult Californian population is Arab American. Since the CHIS enrollment is estimated at 20,000 Californians every year, we might expect that 400 Arab Americans should be identified each year through the survey. Our sample shows 80–150 Arab Americans are being recruited annually suggesting under-recruitment of this population or an inability to properly identify Arab Americans using the questions on birth place and language. Second, all outcomes are self-reported by the participant and are not verified with medical records potentially leading to reporting bias due to stigma and shame associated with particular health behaviors and outcomes [17]. Third, due to the lack of recruitment of a large number of Arab Americans, we were underpowered to detect differences in particular health outcomes and health behaviors. Finally, the group of Arab Americans isolated from CHIS is heterogeneous and this heterogeneity may mask disease risk patterns for this population. Specifically, the inclusion of Arabic speaking individuals from North Africa, who may self-identify as Black on standard race surveys, may in fact have different social experiences and exposures that may put them at differential risk for various health outcomes.

Despite these limitations, our study reports on one of the largest rigorously conducted population-based samples of Arab Americans from California and provides one of the first looks at differences between Arab Americans and non-Hispanic Whites in California on a number of health behaviors and outcomes. We found Arab Americans participated in less risky health behaviors and had better health outcomes, except with regard to diabetes. Future work in California should allow for self-identification of Arab American ethnicity in order to properly understand the health needs of this minority population. Understanding the differences in demographics, socioeconomic status, health behaviors, and health outcomes between Arab Americans and non-Hispanic Whites in California will help public health officials and clinicians better target health messaging and health promotion to Arab Americans, an important minority subgroup.