Diabetes among Hispanics in Oklahoma: Assessing Disparities to Guide Basic and Applied Research

Abstract

Background –

Diabetes, a leading cause of morbidity and mortality in the United States, disproportionally affects minority populations. In 2015, Hispanics, the largest minority in the country, had the third highest rate of diabetes prevalence and the third highest age-adjusted rate of diabetes-related mortality. Substantial progress in understanding diabetes disparities nationally and in many areas of the country has been made. However, little is known about diabetes and related mortality among Hispanics in Oklahoma, which is known as a Hispanic “new Settlement” state due to the relatively recent and substantial growth of this population.

Methods –

We used Oklahoma Behavioral and Risk Factor Surveillance Survey data (2011–2016) to calculate population estimates of diabetes prevalence and selected sociodemographic characteristics for Hispanic and Non-Hispanic adults in the state. We used Oklahoma Death Registry data to estimate diabetes-related mortality rates for Hispanic and Non-Hispanic adults for the same five-year period. We examined differences in diabetes prevalence and diabetes-related mortality across selected sociodemographic characteristics.

Results –

Hispanics are the largest minority group in Oklahoma. Spanish is the most common non-English language spoken in the state. Hispanics are younger, poorer, less educated and experience less access to health care compared to other populations in Oklahoma. While Hispanics had the fifth highest reported diabetes prevalence rate during the five-year period examined, they had the third highest diabetes-related mortality rate in the state.

Discussion and Conclusions –

There is a need for community engagement and basic and applied research to help identify and reduce diabetes disparities in the growing Hispanic population in Oklahoma.

BACKGROUND

In 2015, the estimated prevalence of both diagnosed and undiagnosed diabetes in the United States was 9.4%, and diabetes was the seventh leading cause of death.¹ Diabetes imposes significant financial and human costs on individuals, families, and communities. In the U.S. in 2017, diagnosed diabetes was associated with an estimated $237 billion in direct medical costs and $90 billion in reduced productivity, in addition to human costs of reduced quality of life and nonpaid caregiving.² Diabetes disparities disproportionally affect minority groups in terms of prevalence and disease burden.^1,3–10 Between 2013 and 2015, the overall age-adjusted prevalence of diagnosed diabetes nationally was 15.1% among American Indians/Alaska Natives, 12.7% among non-Hispanic Blacks, 12.1% among Hispanics, and 8.0% among Asians, compared to 7.4% among non-Hispanic Whites.¹ As the largest U.S. minority group with 15% of the total U.S. population,¹¹ Hispanics had the third highest age-adjusted rates of diabetic mortality in the nation from 1999–2017.¹² The Hispanic Community Health Survey/Study of Latinos (HCHS/SOL),¹³ a national prospective cohort study from 2008 through 2011 that estimated diabetes prevalence rankings among several Hispanic nationalities, found that Mexicans had the highest prevalence (18.3%), followed by Dominicans and Puerto Ricans (18.0%), Central Americans (17.7%), Cubans (13.4%), and South Americans (10.2%).¹³

During the past decades, Oklahoma has seen a dramatic growth in its Hispanic population, and thus is known as a Hispanic “new settlement” state.¹⁴ From 2000 to 2010, Oklahoma’s Hispanic population increased by 85.2%, substantially higher than the growth of the state’s non-Hispanic population which increased only 4.5% during the same period. While Hispanics were the largest minority group (10.9%) in Oklahoma in 2018,¹⁵ relatively little is known about this population in terms of demographics and diabetes. In this study, we used Oklahoma Behavioral and Risk Factor Surveillance Survey (BRFSS) data for a five-year period (2011–2016) to examine select sociodemographic characteristics of Oklahoma’s Hispanic population and calculate population estimates of diabetes prevalence for Hispanic and Non-Hispanic adults in the state. We used Oklahoma Death Registry data to estimate diabetes-related mortality rates for Hispanic and Non-Hispanic adults for the same five-year period. We examined differences in diabetes prevalence and diabetes-related mortality across selected sociodemographic characteristics. We discuss findings in relationship to Hispanic diabetes prevalence and mortality nationally and make recommendations for future research and interventions to improve our understanding of diabetes and health outcomes among this important and growing segment of Oklahoma’s population.

METHODS

Determining Diabetes Prevalence among Hispanics in Oklahoma

Data Sources.

We used sociodemographic and diabetes prevalence data for Oklahoma from the Centers for Disease Control and Prevention (CDC) Behavioral Risk Factor Surveillance System (BRFSS), an annual nation-wide telephone survey that collects cross-sectional Information about health behaviors, chronic conditions, and use of preventive and healthcare services among U.S. adults. We used Oklahoma BRFSS data for a 6-year period (2011–2016) in order to avoid small subgroup sizes with stratification by multiple sociodemographic characteristics.

We also used the U.S. Census Bureau’s American Community Survey (ACS),¹⁶ an annual nation-wide survey that collects Information about social, economic, housing, and demographic characteristics, as well as data from the Migration Policy Institute’s 2016 data Compilation¹⁷ and Data USA.¹⁸

Variables and Data Analysis.

Using 2011–2016 combined BRFSS data for Oklahoma,¹⁹ we obtained population estimates for demographics. We classified each population subgroup according to sex (male or female), age (18–44, 45–64, and 65+ years of age), education (Less than high school, High school graduated, Some College, and College graduated), and income (<$15,000, $15,000-$24,999, $25,000-$49,999, $50,000-$74,999, and $75,000+). We also included Information about health care coverage (answering YES to the question “Do you have any kind of health care coverage, including health Insurance, prepaid plans such as HMOs, or government plans such as Medicare, or Indian Health Service?”), whether survey respondents were unable to visit a doctor due to cost (answering YES to the question “Was there a time in the past 12 months when you needed to see a doctor but could not because of cost?”), and whether respondents had a personal doctor or health care provider (answering YES to the question “Do you have one person you think of as your personal doctor or health care provider?”). Population estimates and corresponding 95% confidence intervals were created.

Additional demographic data included Information on population growth, place of birth, and language. Hispanic population growth in Oklahoma was calculated from data provided by the State of Oklahoma.²⁰ Migration Policy Institute¹⁷ data, tabulated from the U.S. Census Bureau’s ACS and the Decennial Census, was used to assess place of birth for Oklahoma Hispanics. We obtained data on non-English speakers in Oklahoma from the Data USA website,¹⁸ which compiles its data from the ACS 5-year and ACS 1-year estimates. We determined population change percentage for Hispanics from 2000–2010 and Hispanic population growth from 1970–2010.²⁰ We used population estimates for foreign-born Hispanics in Oklahoma from the Migration Policy Institute’s 2016 data Compilation.¹⁷ Specifically, we utilized population estimates and percentages of foreign-born Hispanics in Oklahoma, including data on region/country born (i.e., South America, Central America, Mexico, and the Caribbean). We also examined the most common non-English language data on the website Data USA,¹⁸ which compiles U.S. government data into a single source, by selecting the state of Oklahoma and its “diversity” section.

We used the 2011–2016 combined BRFSS data for Oklahoma to calculate age-adjusted prevalence of diabetes by sociodemographic characteristics.²¹ We identified diabetes diagnosis by a positive response to the question “Has a doctor, nurse or other health Professional ever told you that you have diabetes?”²¹ We classified those diagnosed with diabetes according to age, sex, education, income, race/ethnicity, not visiting a doctor due to cost, having a personal doctor or health care provider, and insurance status. We compared population estimates for those diagnosed with diabetes among all residente, Hispanic residente, and non-Hispanic residents of Oklahoma. We created summaries of diabetes prevalence after stratifying by sociodemographic characteristics and made formal statistical comparisons between Hispanic and non-Hispanic populations for the age-adjusted prevalence of diabetes. Point estimates and 95% confidence intervals were calculated. The SAS SURVEYLOGISTIC procedure was used to compare the odds of diabetes between Hispanic and Non-Hispanic populations using survey weights that were age adjusted.²²

Determining Mortality Rates among Hispanics in Oklahoma

Data Sources.

We obtained age-adjusted diabetes-specific mortality estimates from the Oklahoma Statistics on Health Available for Everyone (OK2SHARE) web-based query system that is supported by the Oklahoma State Department of Health.¹⁹

Data Analysis.

We calculated diabetes-specific mortality using ICD-10 mortality codes during the period 2011–2016¹⁸ to ensure a sufficient number of diabetes-related deaths within sociodemographic subgroups.

Age-adjusted mortality rates per 100,000 population were estimated. Mortality rates are presented by age, sex and race/ethnicity. Stratification by additional sociodemographic factors was not possible using the OK2SHARE data query system. Age adjustment was made relative to the 2000 U.S. Standard Population.

Since all data used in this study are publicly available and de-identified, no institutional review board review or approval were required.

RESULTS

In 2017, the total population of Oklahoma was approximately 3.9 million, with 5.7% or 223,883 being foreign-born. Of these, Latin America was the place of birth for 56.6% or 126,609 people. The majority of foreign-born Latin Americans in Oklahoma in 2017 were from Central America (51.8%), more specifically Mexico (44.2%). Only 3.5% of foreign-born Hispanics in Oklahoma were from South America, and 1.2% were from the Caribbean.¹⁷ In 2015, the most common non-English language spoken in Oklahoma was Spanish, with 7.16% of the overall population being native Spanish or Spanish Creole speakers.¹⁸ The sociodemographic characteristics of Hispanics in Oklahoma and select social determinants of health (education, income, and access to health care) are presented in Table 1.

Table 1:

Sociodemographic Characteristics of Hispanic and Non-Hispanic Adults, Oklahoma Behavioral Risk Factor Surveillance Survey 2011–2016

Demographics		Overall		Hispanic		Non-Hispanic
		%	95% CI	%	95% CI	%	95% CI
Sex	Male	49.0	(48.4, 49.6)	50.7	(48.0, 53.3)	48.9	(48.2, 49.5)
	Female	51.0	(50.4, 51.6)	49.3	(46.7, 52.0)	51.1	(50.5, 51.8)
Age	18–44 years	47.4	(46.8, 48.0)	75.1	(73.1, 77.0)	45.0	(44.4, 45.7)
	45–64 years	33.1	(32.6, 33.6)	21.0	(19.2, 22.9)	34.1	(33.6, 34.7)
	65+ years	19.5	(19.1, 19.8)	3.9	(3.2, 4.6)	20.8	(20.4, 21.2)
Education	Less than high school	14.7	(14.1, 15.2)	42.4	(39.8, 45.1)	12.2	(11.7, 12.8)
	High school graduated	32.0	(31.4, 32.6)	30.1	(27.7, 32.6)	32.2	(31.6, 32.8)
	Some college	32.3	(31.7, 32.9)	19.5	(17.4, 21.6)	33.4	(32.8, 34.1)
	College graduated	21.0	(20.5, 21.4)	7.9	(6.9, 9.0)	22.1	(21.6, 22.6)
Could not visit a doctor due to cost	Yes	16.8	(16.3, 17.3)	30.2	(27.7, 32.6)	15.7	(15.1, 16.2)
Have a personal doctor or health care provider	Yes	75.1	(74.5, 75.7)	50.5	(47.8, 53.1)	77.3	(76.7, 77.9)
Insurance status	Any coverage	83.2	(82.6, 83.7)	51.5	(48.8, 54.1)	85.9	(85.4, 86.4)
Income	<$15,000	13.0	(12.6, 13.5)	18.3	(16.2, 20.5)	12.6	(12.1, 13.1)
	$15,000-$24,999	20.1	(19.5, 20.6)	33.9	(31.1, 36.6)	18.9	(18.3, 19.4)
	$25,000-$49,999	27.9	(27.3, 28.5)	29.6	(27.0, 32.3)	27.8	(27.2, 28.4)
	$50,000-$74,999	15.0	(14.6, 15.5)	8.7	(7.2, 10.3)	15.6	(15.1, 16.1)
	$75,000+	23.9	(23.4, 24.5)	9.4	(7.8, 11.1)	25.2	(24.6, 25.8)

Diabetes Prevalence among Hispanic and Non-Hispanic Adults

Diabetes Prevalence by Race/Ethnicity.

Hispanic adults in Oklahoma had a lower prevalence of diabetes (9.4%) than non-Hispanic American Indians (18.7%), non-Hispanic African Americans (14.9%), non-Hispanic Whites (12.2%), and non-Hispanic Multiracial adults (11.3%).²¹ The odds of diabetes, after adjustment for age, were significantly higher among non-Hispanics compared to Hispanics (odds ratio 1.41, 95% CI: 1.20 to 1.67, p<0.0001).

Diabetes Prevalence by Sex.

Hispanic males and females had similar diabetes rates (9.4% and 9.3%), which mirrors the sex-specific diabetes rates among the state’s overall population (12.4% and 12.5%).²¹

Diabetes Prevalence by Age.

Diabetes rates were higher among Hispanics who were 65+ years of age (32.7%), compared to Hispanics 45–64 years of age (21.5%), and non-Hispanics 65+ years of age (23%).²¹

Diabetes Prevalence by Education.

Hispanics who were high school graduates had the lowest diabetes rate (6.6%), followed by those who had some College education (9.0%).²¹ Hispanics with less than high school had the highest rate of diabetes (11.4%), followed by Hispanics who graduated from College (10.2%).²¹ Similarly, adults with less than a high school education had the highest rate of diabetes in the non-Hispanic population (18.8 %). ²¹ In contrast, the lowest diabetes rates in the non-Hispanic population were among College graduates (10.3%).²¹

Diabetes Prevalence by Income.

Among Hispanics, those in the lowest income category (<$15,000/year) had the highest diabetes rates (12.9%).²¹ This rate was lower, however, than non-Hispanics in the same income group (18.1%).²¹ Diabetes rates (12.5%) among Hispanics who made between $50,000 and $74,999 and who made more than $75,000 (11.1%) were similar to those in the lowest income group (12.9%).²¹ This pattern differs from that found for non-Hispanics among whom rates of diabetes declined for each increase in income group. The lowest rates of diabetes for Hispanics were found among those making $15,000 to $24,000 (8.4%) and $25,000 to $49,000 (9.0%), both of which were lower than non-Hispanics in the same income groups (17.1% and 12.9%, respectively).²¹

Diabetes Prevalence by Health Care Access Factors.

Age-adjusted diabetes prevalence rates for Hispanics were lower for three distinct access to health care variables, i.e. insurance status, doctor visit cost, and personal health care provider. The diabetes rates were lower for Hispanics who reported holding any health care coverage (11.4%), not visiting a doctor due to cost (10.5%) and having a personal doctor or health care provider (13.7%) compared to non-Hispanics (13.5%, 13.3%, and 15.0%, respectively). ²¹

Diabetes Mortality Rates among Hispanic and Non-Hispanic Adults

The age-adjusted diabetes-specific mortality rate for Hispanic adults in Oklahoma from 2011 to 2016 was 33.4 per 100,000, which was the third highest after American Indians (56.5 per 100,000) and African Americans (43.7 per 1000,000).²¹ Hispanics aged 65 years or older had the highest age-adjusted diabetes-specific mortality rate (187.4 per 100,000), which exceeded the rates for non-Hispanics in the same age category (159.7 per 100,000).²¹ Hispanics aged 44 years or younger had the lowest age-adjusted diabetes-specific mortality rate (0.9 per 100,000) followed by those 45 to 64 years old (28.3 per 100,000).²¹ Hispanic males had a higher age-adjusted diabetes-specific mortality rate (38.1 per 100,000) than Hispanic females (29.1 per 100,000).²¹ This pattern was similar in the non-Hispanic population. Mortality rates for the variables of education, health care access (insurance status, doctor visit cost, and personal health care provider) and income are not available in OK2SHARE.

DISCUSSION

The Hispanic population in Oklahoma has grown rapidly. In 2018, Hispanics comprised 10.9% of the state’s population making them the largest minority group.¹⁵ Oklahoma’s Hispanic population is diverse in terms of country of origin, with Mexico having the largest representation. Spanish is the most common non-English language spoken in Oklahoma.¹⁸ Overall, Hispanics were younger, less educated, poorer, and experienced less access to health care compared to non-Hispanic populations (Table 1). These demographic characteristics of Hispanics in Oklahoma are reflections of national demographics.¹⁵

The age-adjusted prevalence of diabetes for Hispanics in Oklahoma (9.4% based on 2011–2016 combined BRFSS data) was the fifth highest rate in the state. This finding differs from national diabetes statistics according to which Hispanics ranked third in diabetes prevalence (12.1%).² It is likely that diabetes among Hispanics is underdiagnosed or underreported in Oklahoma. A low level of health care access may contribute to undiagnosed diabetes. Even with access to health care, language and cultural barriers²³ may contribute to undiagnosed diabetes. Previous studies have shown that Spanish-speaking or Spanish-preferring Hispanics had less access to health care and preventive care.^24,25 As a “new settlement state,”¹⁴ Oklahoma lacks the infrastructure to integrate new Hispanic immigrants, with associated language and cultural barriers. The current political climate of anti-immigration policies is another underlying factor affecting Oklahoma¹⁴ where, between 2010 and 2014, 125,989 Oklahomans lived with at least one undocumented family member.²⁶ Hispanic families and individuals – regard-less of Immigration status – live with fear and trepidation which negatively affects their health and willingness to seek health care,^27–29 and to respond to state-wide surveys. Considering these barriers, undiagnosed diabetes among Hispanics in Oklahoma is a serious concern.

Examining diabetes prevalence among Hispanics, we found potential, although limited, influence of social determinants of health (SDOH).³⁰ We included the SDOH of Education, Economic Stability, and Health and Health Care (Table 2). We found that less education and poverty were correlated with higher diabetes rates, and that Hispanics with less than high school education had the highest rate of diabetes. Unexpectedly, the second highest rate was among College graduates.

Table 2:

Age-adjusted Diabetes Prevalence by Sociodemographic Characteristics for Overall, Hispanic and Non-Hispanic Adult Populations, Oklahoma Behavioral Risk Factor Surveillance Survey 2011–2016

Demographics		Overall		Hispanic		Non-Hispanic
		%	95% CI	%	95% CI	%	95% CI
Sex	Male	12.4	(11.9, 13.0)	9.4	(7.3, 11.5)	12.7	(12.1, 13.2)
	Female	12.5	(12.1, 13.0)	9.3	(7.6, 11.0)	12.8	(12.3, 13.3)
Age	18–44 years	4.6	(4.2, 5.0)	4.4	(3.2, 5.5)	4.6	(4.2, 5.1)
	45–64 years	16.8	(16.1, 17.5)	21.5	(17.7, 25.4)	16.5	(15.8, 17.2)
	65+ years	23.1	(22.3, 23.9)	32.7	(24.5, 40.9)	23.0	(22.2, 23.8)
Education	Less than high school	17.2	(15.9, 18.5)	11.4	(9.0, 13.7)	18.8	(17.3, 20.4)
	High school graduated	12.6	(12.0, 13.2)	6.6	(4.7, 8.4)	13.1	(12.4, 13.7)
	Some college	11.6	(11.0, 12.2)	9.0	(5.8, 12.1)	11.7	(11.1, 12.3)
	College graduated	10.4	(9.8, 11.0)	10.2	(6.3, 14.0)	10.3	(9.7, 10.9)
Could not visit a doctor due to cost	Yes	12.9	(12.0, 13.9)	10.5	(7.9, 13.1)	13.3	(12.3, 14.4)
Have a personal doctor or health care provider	Yes	14.9	(14.5, 15.4)	13.7	(11.6, 15.9)	15.0	(14.5, 15.4)
Insurance status	Any health care coverage	13.5	(13.1, 13.9)	11.4	(9.5, 13.4)	13.5	(13.1, 13.9)
Income	<$15,000	17.6	(16.3, 18.8)	12.9	(9.1, 16.6)	18.1	(16.7, 19.4)
	$15,000-$24,999	15.9	(15.0, 16.9)	8.4	(6.0, 10.7)	17.1	(16.0, 18.2)
	$25,000-$49,999	12.6	(11.9, 13.3)	9.0	(6.2, 11.8)	12.9	(12.1, 13.6)
	$50,000-$74,999	11.4	(10.5, 12.4)	12.5	(7.0, 18.0)	11.3	(10.4, 12.2)
	$75,000+	8.8	(8.1, 9.4)	11.1	(5.4, 16.7)	8.7	(8.1, 9.4)
Race/Ethnicity	Non-Hispanic White	12.2	(11.8, 12.6)			12.2	(11.8, 12.6)
	Non-Hispanic Black or African American	14.9	(13.3, 16.4)			14.9	(13.3, 16.4)
	Non-Hispanic American Indian	18.7	(16.6, 20.7)			18.7	(16.6, 20.7)
	Non-Hispanic Asian	8.9	(6.1, 11.7)			8.9	(6.1, 11.7)
	Non-Hispanic Other race	6.4	(2.9, 9.9)			6.4	(2.9, 9.9)
	Non-Hispanic Multiracial	11.3	(5.0, 17.7)			11.3	(5.0, 17.7)
	Hispanic	9.4	(8.0, 10.7)	9.4	(8.0, 10.7)

The correlation of income with diabetes prevalence followed a similar pattern. The lowest income was correlated with the highest diabetes rate among Hispanics, but the two highest incomes were correlated with the second and third highest rates. The causes for these unexpected findings are not clear and require further study. However, causes might be related to acculturation, whereby greater acculturation is associated with higher diabetes prevalence.³¹,³² While our analysis by health care access showed diabetes prevalence rates for Hispanics to be lower than those for non-Hispanics, this finding might reflect overall diabetes rankings in Oklahoma, with Hispanics having the fifth highest prevalence. Considering that only 51.5% of Hispanics in Oklahoma reported health care coverage, Hispanic diabetes rates may be underreported. Lack of insurance is known to result in poorer health outcomes for reasons that include delayed care seeking and not receiving needed care.³ Overall, the potential impact of SDOH on health outcomes among Oklahoma Hispanics is supported by a growing scientific literature that highlights the disproportionate impact of SDOH on minority populations and its contributing role to poorer diabetes outcomes.^3,27,33 Additional research needs to examine how specific SDOH factors influence diabetes among Hispanics in Oklahoma.

While we found that Hispanics in Oklahoma had a lower diabetes prevalence (9.4%) compared to non-Hispanic Whites (12.2%), age-adjusted diabetes-specific mortality was higher among Hispanics (33.4/100,000) compared to non-Hispanic Whites (21.2/100.000), and among Hispanics 65+ of age (187.4/100.000) compared to non-Hispanic Whites in the same age group (159.7/100.000) (Table 3). These state-level mortality rates among Hispanics are similar to those found at the national level, with Hispanic Americans being 1.5 times more likely to die from diabetes compared to non-Hispanic Whites.⁹ Additional research is needed to better understand the drivers of diabetes mortality disparities in Oklahoma.

Table 3:

Age-adjusted Diabetes-specific Mortality Rates by Sociodemographic Characteristics for Overall, Hispanic and Non-Hispanic Adults Population, Oklahoma Vital Statistics 2011–2016

Demographics		Overall Age-adjusted	Hispanic Age-adjusted	Non-Hispanic Age-adjusted
		Diabetes-specific	Diabetes-specific	Diabetes-specific
		Death Rate (per 100,000 population)	Death Rate (per 100,000 population)	Death Rate (per 100,000 population)
Sex	Male	35.6	38.1	35.6
	Female	25.4	29.1	25.4
Age	≤ 44 years	2.3	0.9	2.5
	45–64 years	36.1	28.3	36.5
	65+ years	160.5	187.4	159.7
Race/Ethnicity	Non-Hispanic White	21.2		21.2
	Non-Hispanic Black or African American	43.7		43.7
	Non-Hispanic American Indian	56.5		56.5
	Non-Hispanic Other race	19.7		19.7
	Hispanic	33.4	33.4

NOTE: rates presented after stratification by age are age-specific instead of age-adjusted estimates.

Citation: Oklahoma State Department of Health (OSDH), Center for Health Statistics, Health Care Information, Vital Statistics 2011 to 2016, on Oklahoma Statistics on Health Available for Everyone (OK2SHARE). Accessed at http://www.health.ok.gov/ok2share on 010CT2Q18:19:32:53.

Limitations included the small sample sizes and the data sources used for this analysis. Due to limitations of small sample sizes, we combined multiple years (2011–2016) of BRFSS data for Oklahoma to obtain population estimates for sociodemographic variables and calculate age-adjusted diabetes prevalence and mortality. While the BRFSS is the only state-level source of Information about diabetes prevalence, it relies on self-report, which likely under-estimated prevalence rates,³⁴ particularly among groups with limited health care access such as Hispanics. While the BRFSS does not provide estimates of undiagnosed diabetes,³⁴ it currently is the only source of data on state-level diabetes prevalence rates.³⁴ Since local statistics for Hispanics in Oklahoma remain difficult to access,²³ the state-level BRFSS data are a good source for an initial understanding of diabetes among Hispanics in Oklahoma.

Nevertheless, our findings point to the need of community-engaged basic and Intervention research. Engaging with community partners and other stakeholders who serve and/or are members of Hispanic communities will enhance the development of culturally appropriate data collection strategies and relevant multilevel interventions to help reduce diabetes disparities among diverse Hispanic communities in Oklahoma. Such approaches are valuable for and successful in addressing health disparities.^3,8,9,35 Community-engaged, culturally appropriate data collection will likely yield more representative, rich, relevant and actionable data. Partnerships with community-based Hispanic organizations^23,36 and community members such as community health workers (CHWs),^8,37 are some strategies to collect more meaningful and representative data that leverage important knowledge of local Hispanic communities. Such partnerships can also be beneficial in establishing and strengthening trusting relationships and translating data into locally relevant public health interventions aimed at reducing diabetes disparities. While targeted interventions with patients, providers, or health care systems are effective,⁸ multi-level and multi-disciplinary interventions beyond clinical settings are an important element of a multi-pronged strategy to reduce persistent diabetes disparities and their underlying SDOH.^33,38 Interventions that go beyond increasing knowledge and healthy lifestyles are needed. Shifts in economic, social, health care and policy environments that make healthy choices easy choices, and reduce drivers of inequity, are fundamental to population level improvements in diabetes.²⁷ Utilizing culturally appropriate and strength-based strategies such as the Integration of CHWs into teams to reduce chronic disease in low-income and hard-to-reach populations will improve community and clinical environments, and related outcomes.^3–42 Finally, recent expansion of Medicaid in Oklahoma might positively impact diagnosis and clinical treatment options of diabetes among Hispanics and other disadvantaged groups.

CONCLUSIONS

Assessing demographic factors and diabetes among Hispanics in Oklahoma, we found potential underreporting of diabetes, disparities in prevalence and mortality rates, and a potential impact of SDOH on diabetes prevalence and mortality. Considering the high human and monetary costs of diabetes, we propose community-engaged, culturally appropriate research that can inform the development of locally relevant multilevel interventions. Better understanding and reducing diabetes disparities among Hispanics in Oklahoma is fundamental to improvements in health equity and population health.