Correlates of Glycemic Status Among American Indian Adults With Type 2 Diabetes in Rural Southeastern Oklahoma

Abstract

Background:

Elevated glycosylated hemoglobin (HbA1c) levels have been associated with an increased risk for type 2 diabetes (T2D) complications.

Objective:

The aim of this study was to examine the electronic health record (EHR) correlates of HbA1c levels among adults with T2D using Tribal health services.

Method:

EHR data from one Tribal health center were used for this study. Variables included driving distance to clinics, socioeconomic status, use of primary care and diabetes educators, age, sex, glycemic status, and HbA1c. Descriptive, bivariate and linear regression analyses were performed.

Results:

Nearly 40% of the 5,205 patients in the sample had HbA1c levels < 7% in 2021. HbA1c in 2020 was a strong predictor of HbA1c in 2021. Older age was weakly associated with lower HbA1c. A lack of primary care visits or diabetes educator visits was associated with lower HbA1c, likely due to patients with lower HbA1c levels had less need for T2D-related primary care or diabetes educator visits.

Discussion:

This study provided insight into factors associated with HbA1c levels that deserve further exploration, but it is important to note that factors included in social determinant of health frameworks specifically developed for Indigenous communities were not available in the EHR data set. Therefore, future studies are needed that examine Indigenous social determinants of health specific to the local community.

Keywords: American Indian adults, rural health, type 2 diabetes

High glycosylated hemoglobin (HbA1c) levels elevate the risk of type 2 diabetes (T2D)-related complications (Sun et al., 2022). Glycemic status is multifaceted and complex (Ali et al., 2012; Zakaria et al., 2023). In order to develop targeted interventions to improve T2D outcomes, more information is needed on the contributory factors associated with HbA1c levels that may provide protective factors or increase the risk of diabetes-related complications.

Research has examined factors associated with HbA1c levels, such as access to primary care and socioeconomic status (SES). For some rural residents with T2D, access to care may be particularly challenging (Dugani et al., 2021). Issues rural residents may face include limited access to health care providers, lack of transportation options, and longer travel times (Maganty et al., 2023; Zgibor et al., 2011). Improved geographic access to care has been associated with improved preventive care and patient outcomes in some cases (Jewett et al., 2018; Kelly et al., 2016). While not specific to diabetes, a systematic review concluded that longer travel for patients to health care was associated with poorer health outcomes in 77% of studies (Kelly et al., 2016). Conversely, results from a study of older American Indian adults did not find travel time to be a factor in T2D outcomes (Nicklett et al., 2017).

T2D cases are often treated in the primary care setting, so it is vital to assess driving distance to primary care as well as the number of T2D-related primary care and diabetes educator visits (Kushner et al., 2022). Diabetes educators, important members of the T2D primary care team, may improve the overall patient experience and patient comprehension of T2D (Grohmann et al., 2017).

In addition to accessibility, demographic characteristics such as SES, age, and sex may also influence glycemic status. Low SES has been associated with an increase in T2D complications (Tatulashvili et al., 2020). There is also evidence that HbA1c levels, as well as the proportion of adults with T2D, increase with age (Centers for Disease Control and Prevention [CDC], 2026; Dubowitz et al., 2014). Although differences in glycemic status between men and women have been inconsistent, sociocultural differences by sex could affect glycemic status (Mondesir et al., 2016). These demographic and social determinants intersect with broader structural factors that shape diabetes burden. In addition, American Indian communities have faced systemic issues such as discrimination and historical trauma along with higher rates of chronic diseases, including T2D (Findling et al., 2019; Gillson et al., 2024; U.S. Department of Health and Human Services, Office of Minority Health, 2026).

This study contributes to a deeper understanding of factors associated with HbA1c levels within a rural Tribal health system, which provides health care to eligible patients at no direct cost to the patient. Thus, as described in our conceptual framework, our study examines the various electronic health record (EHR) correlates of HbA1c levels among adults with T2D using Tribal health services (Supplemental Digital Content [SDC] 1). Guided by the existing literature and available variables in the EHR, our conceptual framework categorizes each variable as a glucose management indicator, nonmodifiable correlate, or a modifiable correlate that may influence HbA1c levels. The variables we obtained from the EHR include driving distance to the nearest Choctaw Nation Health Services Authority’s (CNHSA) clinic, SES, age, sex, and frequency of diabetes educator and T2D-related primary care visits.

Methods

Study Design

We examined the association between EHR-derived factors with HbA1c levels in adults with T2D using CNHSA 2020–2021 EHR data, part of a larger 2017–2021 data set. Specifically, the factors we examined included driving distance to the nearest CNHSA clinic, SES, T2D-related primary care visits, diabetes educator visits, age, and sex. The Choctaw Nation of Oklahoma (CNO) and the University of Florida Institutional Review Boards (IRBs) approved this study. The senior author has a longstanding research relationship with the CNO.

Eligibility Criteria

All patients included in the 2017–2021 CNHSA EHR who met the following criteria were included in the study: ≥ 18 years of age and a documented diagnosis of T2D using the International Classification of Diseases, 10th Revision (ICD–10) codes (World Health Organization, 2019). Our study focused on 2021 data, so patients who also met the following criteria were included: CNHSA visit data in 2021, a zip code in 2021 within 200 miles of the nearest CNHSA clinic, and at least one HbA1c value in 2021. Patients were excluded if they did not have a documented HbA1c level in 2020, since 2020 HbA1c levels were used to predict 2021 HbA1c levels. Patients diagnosed with end-stage renal disease (ESRD) documented with an N18.6 ICD–10 code were excluded due to potential issues with the reliability of HbA1c in patients with ESRD (Shrishrimal et al., 2009; Wouk, 2021). Patients with diagnosis codes for other specified diabetes, pure hyperglyceridemia, and diabetic chronic kidney disease were also excluded from the study. Our study flow diagram is displayed in SDC 1.

Setting

CNHSA comprises a 44-bed hospital as well as outlying clinics (Choctaw Nation of Oklahoma, n.d.-c). There are over 225,000 CNO Tribal citizens, the third largest Tribal Nation in the United States, and the CNO reservation covers nearly 11,000 square miles in Southern Oklahoma (Choctaw Nation of Oklahoma, n.d.-b). CNHSA is a Tribally run health care system, and patients are members of a federally recognized tribe (Choctaw Nation of Oklahoma, n.d.-a). However, a percentage of patients (1.5% of our total sample) are not enrolled in a federally recognized Tribe. Instead, they are family members of Tribal citizens or CNHSA employees and can receive some CNHSA services.

Data Retrieval

A limited data set was extracted by CNHSA staff members after CNO IRB approval was received, and a data use agreement was completed. The limited data set included diagnosis codes, sociodemographic data, laboratory HbA1c values, and T2D-related health visit information.

Variables and Measures

HbA1c Levels

A laboratory measure and a marker for glucose levels over the preceding 60 to 90 days, HbA1c, was used to assess glycemic status (Sherwani et al., 2016). We used mean HbA1c levels if a patient had > 1 HbA1c level listed in the EHR in 2020 and again in 2021 for this study. HbA1c levels in 2020 were used to predict HbA1c levels in 2021. HbA1c levels of < 7% are often considered as a HbA1c goal for patients with T2D, so HbA1c levels were dichotomized (< 7% and ≥ 7%) to examine the sociodemographic characteristics in our study (ElSayed et al., 2023).

Driving Distance to the Nearest CNHSA Clinic

Patient addresses were unavailable in the limited data set. Instead, the first zip codes listed in the EHR in 2021 for each patient were used to determine the driving distance to the nearest CNHSA clinic. Geocoded geographic centroids—the geographic centers of zip codes—were used to approximate each patient’s home address. The addresses of CNHSA clinics were available. Geocoded approximate patient locations and exact locations for CNHSA clinics locations were used, and we then calculated the shortest driving distance (miles) to the nearest CNHSA clinic for each patient. We excluded patients with a driving distance of more than 200 miles from their home address to the nearest CNHSA clinic. Members of our study team determined ≤ 200 miles as a reasonable driving distance from the nearest CNHSA clinic.

Frequency of Visits With the Diabetes Educator

Visit information listed within the EHR in 2021 was used to derive the number of visits with a diabetes educator.

Frequency of T2D-Related Primary Care Visits

The number of T2D-related primary care visits in 2021 was counted for each patient using EHR diagnosis codes and visit information. Visits with primary care providers and endocrinologists were included in our count as both provide T2D-related primary care within CNHSA. We also included visits with residents within the Family Medicine Residency Program at CNHSA. We were unable to distinguish between visit types, so it’s possible that a small portion of visits were T2D-related emergency department visits rather than T2D-primary care visits. Few telehealth visits were conducted within CNHSA during the COVID-19 pandemic.

Patient Demographic Characteristics

The first 2021 value in the EHR data set was used for patient characteristics, including sex, age, and race. Since self-reported SES measures such as income, educational attainment, and occupation were unavailable, Medicaid status was used as a proxy for SES. Medicaid status is often used to estimate SES in research using EHR data (Casey et al., 2018). However, it is essential to note that Medicaid status likely does not capture the complexities of all SES domains (Casey et al., 2018). Oklahoma voters approved Medicaid expansion on June 30, 2020, for households with incomes up to 138% of the Federal Poverty Level, with coverage beginning on July 1, 2021 (Oklahoma Health Care Authority, 2021). Thus, we used the final Medicaid status in 2021 reported in the EHR for each patient in our study. SES was defined as low if the patient had Medicaid listed in the EHR; for all other patients, it was defined as not low.

Statistical Analysis

Descriptive statistics were summarized for all data, including sociodemographic variables, HbA1c levels, number of T2D-related primary care and diabetes educator visits, Medicaid status, and the distance to the nearest CNHSA clinic. Pearson’s r and Spearman’s rho correlations were examined. Chi-square tests of independence and independent t-tests were used to compare characteristics of patients with HbA1c levels < 7% and ≥ 7%. Effect sizes, Cohen’s d, Cramer’s V, and Phi were calculated to examine the strength of the association between variables. A linear regression model was used to examine the association between HbA1c levels in 2021 and other factors in 2021, adjusting for HbA1c levels in 2020. SAS (Version 9.4) was used for all statistical analyses, and p-values < .05 were considered significant.

The driving distance to the nearest CNHSA clinic was calculated using SAS. We used the geographic center of the patient’s zip code (Weiss et al., 2021) and the geocoded location of the CNHSA clinics. Google Maps was used to determine the driving distance from the patient’s zip code to the nearest CNHSA clinic (Weiss et al., 2021).

Results

There were 10,506 patients with T2D within the CNHSA 2017–2021 EHR data set. Our study focused on 2021 data, and after applying our exclusion criteria, 8,244 patients with T2D had EHR data in 2021. A total of 5,205 patients met inclusion criteria for this study (Figure 1). Of the 5,205 patients included in our study, 3,712 patients were prescribed non-insulin glucose-lowering medication for at least 90 days in 2021.

Figure 1.

Study flow diagram.

Note. CNHSA = Choctaw Nation Health Services Authority; T2D = type 2 diabetes.

The sample characteristics, overall and by HbA1c levels, are presented in Table 1. The ages of patients ranged from 21 to 90 years, with a mean age of 59.3 (SD = 13.2) years. The sample was 54% female, and 98.5% of patients were American Indian adults; the remaining 1.5% were members of other race categories. There were 15.9% of patients who received Medicaid, which was considered low SES for our study. The mean HbA1c of patients in 2020 was 7.8 (SD = 1.8), ranging from 4.8 to 17.2. In 2021, the mean HbA1c was 7.8 (SD = 1.8), ranging from 4.5 to 16.9. The mean driving distance to the nearest CNHSA clinic was 22.9 (SD = 31.2) miles in 2021. The median driving distance to the nearest CNHSA clinic in 2021 was 12.2 miles, with an interquartile range (IQR) of 3.9–26.7 miles. In 2021, the mean number of T2D-related primary care visits per patient was 2.8 (SD = 2.0), ranging from 0 to 17 visits, and the median was 2 (IQR = 1–4). The mean number of diabetes educator visits per patient was 0.6 (SD = 1.0), ranging from 0 to 18 visits, and the median was 0 (IQR = 0–1).

Table 1.

Sample sociodemographic characteristics (overall and by 2021 HbA1c level)

2021 HbA1c Level

Characteristic
Overall
<7%
≥7%

p

Effect size

Number of patients n

5205
2046
3159

Age (y) mean (SD)
59.3 (13.2)
60.4 (13.0)
58.5 (13.2)
<.001a

d=.15

Patient sex n (%)

<.001b

Φ=.05

 Female
2816 (54.1)
1173 (41.7)
1643 (58.3)

 Male
2389 (45.9)
873 (36.5)
1516 (63.5)

Race n (%)c, d

.99e

Φ=.0002

 American Indian
5124 (98.5)
2014 (39.3)
3110 (60.7)

 Other
79 (1.5)
31 (39.2)
48 (60.8)

SES n (%)e

.002b

Φ=.04

 Low SES
825 (15.9)
284 (34.4)
541 (65.6)

 Not low SES
4380 (84.1)
1762 (40.2)
2618 (59.8)

Distance to nearest clinic n (%)

.82b

V=.01

 <5 miles
2185 (42.0)
843 (38.6)
1342 (61.4)

 5–<25 miles
1623 (31.2)
643 (39.6)
980 (60.4)

 25–<50 miles
781 (15.0)
314 (40.2)
467 (59.8)

 50–200 miles
616 (11.8)
246 (39.9)
370 (60.1)

PC visits per year n (%)

<.001b

V=.20

 0
464 (8.9)
276 (59.5)
188 (40.5)

 1–2
2142 (41.2)
979 (45.7)
1163 (54.3)

 3–5
2087 (40.1)
673 (32.2)
1414 (67.8)

 >5
512 (9.8)
118 (23.0)
394 (77.0)

DE visits n (%)

<.001b

Φ=.14

 Yes
2364 (45.4)
756 (32.0)
1608 (68.0)

 No
2841 (54.6)
1290 (45.4)
1551 (54.6)

Correlates of HbA1c Levels

In 2021, 39.3% of patients had a mean HbA1c < 7%, while 60.7% had a mean HbA1c ≥ 7%. There was a significant difference between HbA1c of < 7% and ≥ 7% on age, sex, SES, number of T2D related primary care visits, and number of diabetes educator visits; the effect sizes were small for age, primary care visits, and diabetes educator visits and negligible for others (Table 1).

Bivariate analysis (Table 2) using Spearman’s rho showed a weak, positive correlation between number of primary care visits and age (r = .21, p < .001), as well as between number of primary care visits and number of diabetes educator visits (r = .21, p < .001). Primary care visits were positively, but weakly, correlated with HbA1c levels in 2020 (r = .23, p < .001) and in 2021 (r = .20, p < .001). Diabetes educator visits were also positively, but weakly, correlated with HbA1c levels in 2020 and 2021 (r = .17, p < .001; r = .17, p < .001). Using Pearson’s r, there was a weak, negative correlation between both age and mean HbA1c in 2020 (r = −.17, p < .001) and age and mean HbA1c in 2021 (r = −.15, p < .001). Mean HbA1c in 2020 was strongly, positively correlated with mean HbA1c levels in 2021 (r = .72, p < .001).

Table 2.

Pearson’s r Correlations (below the diagonal) and Spearman Correlations (above the diagonal)

Age
Distance to nearest clinic†

PC visits
DE visits
2020 mean HbA1c
2021 mean HbA1c

Age
1
.02
(.22)
.21
(<.001)
.02
(.29)
−.14
(<.001)
−.12
(<.001)

Distance to nearest clinic†

.02
(.16)
1
−.05
(<.001)
−.07
(<.001)
−.01
(.60)
−.03
(0.07)

PC visits
.22
(<.001)
−.06
(<.001)
1
.21
(<.001)
.23
(<.001)
.20
(<.001)

DE visits
.0008
(.95)
−.08
(<.001)
.20
(<.001)
1
.17
(<.001)
.17
(<.001)

2020 mean HbA1c
−.17
(<.001)
−.01
(.39)
.15
(<.001)
.17
(<.001)
1
.76
(<.001)

2021 mean HbA1c
−.15
(<.001)
−.04
(.01)
.12
(<.001)
.16
(<.001)
.72
(<.001)
1

The mean distance to the nearest CNHSA clinic from patients’ home zip codes, by the number of primary care visits, is shown in Table 3. Patients with > 5 T2D-related primary care visits in 2021 had lower mean and median distances to the nearest clinic than patients with ≤ 5 such visits. In 2021, 464 patients (8.9% of the sample) had no T2D-related primary care visits, with a mean distance of 23.2 miles (SD = 31.0) and a median distance of 12.5 miles to the nearest CNHSA clinic. Whereas 512 patients (9.8% of the sample) had > 5 T2D-related primary care visits, with a mean distance of 16.1 (SD = 21.4 miles and a median distance of 8.6 miles to the nearest CNHSA clinic).

Table 3.

Distance to the nearest clinic by number of primary care visits

Primary care visits

n

Mean (SD)

Median (IQR)
10th −90th Percentile
Minimum-Maximum

0 Visits
464
23.2 (31.0)
12.5 (3.9–27.0)
1.4–67.0
.9–173

1–2 Visits
2142
25.6 (35.1)
13.0 (3.9–29.0)
1.6–79.8
.9–197

3–5 Visits
2087
21.7 (28.5)
13.0 (3.9–26.2)
1.6–55.7
.9–196

>5 Visits
512
16.1 (21.4)
8.6 (3.3–20.3)
1.4–39.6
.9–164

A linear regression model was used to examine the HbA1c in 2020 and the correlates of HbA1c levels in 2021 (Table 4). While adjusting for all other factors listed in Table 4, HbA1c levels in 2020 were strongly associated with HbA1c levels in 2021 (b = .686, p < .001). Older age was weakly associated with lower HbA1c levels in 2021 (b = −.004, p = .002). No primary care visits compared to > 5 were associated with lower HbA1c levels in 2021 (b = −.191, p = .02). No diabetes educator visits compared to ≥ 1 diabetes educator visits were associated with lower HbA1c levels in 2021 (b = −.084, p = .02).

Table 4.

Multiple regression associations between 2021 mean HbA1c levels with 2020 mean HbA1c levels and 2021 correlates

Health status indicator & correlates
B
SE B

t

p

2020 mean HbA1c
.686
.010
70.11
<.001

Age
−.004
.001
−3.05
.002

Female sex
−.029
.034
−0.84
.40

Male sex†

Not low SES
−.072
.048
−1.50
.14

Low SES†

Nearest clinic <5 miles
.112
.057
1.98
.048

Nearest clinic 5–<25 miles
.056
.059
.96
.34

Nearest clinic 25–<50 miles
.044
.067
.66
.51

Nearest clinic ≥50 miles†

No PC visits
−.191
.082
−2.34
.02

1–2 PC visits
−.004
.063
−.06
.95

3–5 PC visits
−.028
.061
−.46
.65

>5 PC visits†

No DE visits
−.084
.035
−2.41
.02

DE visits†

Discussion

We used EHR data to examine correlates of glycemic status among patients from a single rural Tribal health system. Nearly 40% of patients in our sample had a mean HbA1c < 7% in 2021, compared with roughly 50% of U.S. adults with diabetes who had an HbA1c < 7% according to the 2020 National Diabetes Statistics Report (CDC, 2026). However, individualized approaches to glycemic goals are significant, so HbA1c target levels for some patients in our sample were likely > 7% (ElSayed et al., 2023). Consistent with prior research, past blood glucose levels were a strong predictor of future HbA1c levels. Interestingly, older age was weakly associated with lower HbA1c levels. Furthermore, our study found that patients with no T2D-related primary care visits had lower HbA1c levels than those with > 5 visits. Additionally, no diabetes educator visits were associated with lower HbA1c levels than ≥ 1 diabetes educator visits. This study contributes importantly to improving our understanding of glycemic status and correlates of HbA1c levels that may be unique to American Indian adults receiving CNHSA services. CNHSA provides comprehensive health services to Tribal citizens, including several outlying clinics, which may differ from those in other Tribal communities. It is crucial to note that there are 574 unique federally recognized nations in the U.S. (Bureau of Indian Affairs, n.d.), with variations in health care access, geographic locations, languages, and culture.

After adjusting for other covariates, older age was weakly associated with lower HbA1c levels. However, this association had a trivial effect size, with a difference in mean age of less than 2 years between patients who achieved and did not achieve an HbA1c level < 7%. This finding is consistent with previous research, which found that younger age (< 58 years) and higher baseline HbA1c were both associated with poor glycemic outcomes (Garvey et al., 2024). Furthermore, a separate study found that a younger age at T2D diagnosis was associated with a higher risk of all-cause mortality (Zhang et al., 2024). These results underscore the importance of glycemic control, particularly in patients diagnosed at a younger age with T2D.

We also observed a small effect size between age and primary care visits in our bivariate analysis. A study with a diverse sample but an unknown number of American Indian participants found that younger age at T2D diagnosis was associated with reduced likelihood of in-person primary care visits and lower likelihood of achieving glucose control at 1-year post-diagnosis (Gopalan et al., 2020). This finding suggests a potential relationship between age, primary care use, and glycemic status. Further research is needed to investigate this association in greater detail, particularly within American Indian populations.

Longer distances to primary care were not associated with higher HbA1c levels in our study, as in some previous studies (Strauss et al., 2006; Zgibor et al., 2011). This finding may be due to strengths of the Tribal health system, such as the large number of outlying clinics. Also, Tribal Transit, which provides transportation to health care visits, is offered to Tribal members, which may partially explain why HbA1c levels were not associated with longer distances to primary care. Consistent with our findings, a previous study among older American Indian adults with T2D found that access-related barriers, including travel time, were not associated with higher HbA1c levels (Nicklett et al., 2017). Future studies should further explore the access-related facilitators identified by the community for patients with T2D who use Tribal health care services.

The lack of T2D-related primary care or diabetes educator visits was associated with lower HbA1c levels than those with more visits. A plausible explanation may be that individuals with lower HbA1c levels did not need a primary care or diabetes educator visit specifically to address T2D in 2021. Also, nearly 30% of our sample was not prescribed non-insulin glucose-lowering medication for at least 90 days in 2021, which may have influenced the number of visits. Patients are likely referred to a diabetes educator when experiencing complications or barriers to self-management of T2D (James, 2021), which may explain the higher percentage of patients with at least one diabetes educator visit who had HbA1c levels ≥ 7%. Further analysis is also needed to examine how HbA1c levels change over time with ongoing visits with a diabetes educator. Health literacy should also be included as a variable in these studies. Moreover, the small effect size observed between primary care and diabetes educator visits is likely due to the integration of diabetes educators into the primary care setting, which provides patients with same-day access to multiple services.

We found no association between sex and HbA1c levels while adjusting for other variables. Sex differences in HbA1c levels have been inconsistent in the literature (Kautzky-Willer et al., 2023; Mondesir et al., 2016; Patel et al., 2022), which may be due to differences in sociocultural factors, such as variation in types of social support between men and women, which may influence T2D self-management (Goins et al., 2022; Mondesir et al., 2016). Specific sociocultural factors that may be unique to sex or gender should be further explored in future studies.

The association between SES and HbA1c levels was not significant when adjusting for other factors. Medicaid expansion in Oklahoma was not implemented until July 1, 2021, so the number of individuals eligible is likely higher than those included in our sample. Nearly 16% of our sample was enrolled in Medicaid, which is slightly lower than the estimated 16.8% of adult Oklahomans who were enrolled in Medicaid in December 2021 (Oklahoma Health Care Service Authority, 2021; U.S. Census Bureau, n.d.). It is also critical to note that the number of American Indian adults enrolled in Medicaid is estimated to be lower than the number eligible, due, among other reasons, awareness of expanded coverage options in 2021 (Medicaid and CHIP Payment and Access Commission, 2021). SES is a multidimensional construct, so in future studies it will be important to examine additional socioeconomic factors that were unavailable in our EHR data set.

Limitations

Our study has several limitations, including the potential for selection bias. We excluded patients who were missing EHR data, such as at least one HbA1c level in 2020 and 2021, potentially resulting in a sample not fully representative of the population. Also, at the time of this study, Medicaid was newly expanded in Oklahoma, and eligible patients may not have been enrolled. There are also limitations in using Medicaid status as a proxy for a complex construct like SES. We used the geocoded geographic center of each patient’s zip code to measure driving distance, a less precise method because we did not have access to each patient’s home address. However, it should be noted that driving distance is only one aspect of access to health care. It has been estimated that nearly 6 million Americans experience delays in medical care due to transportation barriers (Wolfe et al., 2020). Also, body mass index (BMI) is correlated with HbA1c levels, but we did not include BMI or other measures of adiposity as a variable in this study (Boye et al., 2021). We also did not exclude patients with conditions or medications that may affect HbA1c levels (e.g., pregnancy, anemia; ElSayed et al., 2023). This was a retrospective study using data from 2020–2021 during the COVID-19 pandemic, which may have led to atypical visit patterns, limiting the generalizability of the findings. Using the EHR data set, we were also unable to include elements of an Indigenous framework of social determinants of health, which incorporates Indigenous knowledge and worldviews, as well as community-determined factors grounded in health and emphasizing relationships among constructs (Carroll et al., 2022; Oré et al., 2025).

Conclusion

Our findings provide insight into factors that deserve further exploration. Importantly, factors included in social determinants of health frameworks specifically developed for Indigenous communities were not available in the EHR data set. Therefore, future studies are needed that examine Indigenous determinants of health specific to the local community.

Supplementary Material

SDC

Acknowledgments

Research reported in this manuscript was supported by the National Institute of Nursing Research of the National Institutes of Health (Award No. 1R01NR020386). S.M.M. was supported by the National Institute of Diabetes, Digestive, and Kidney Disease (Grant No. 1P30DK092923). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

The authors acknowledge support for PhD student funding from the University of Florida College of Nursing.

The authors would like to thank Choctaw Nation of Oklahoma for their support of this research. We would also like to thank Dr. Michael Weaver, PhD, Professor Emeritus, University of Florida, for his guidance and analytic support for this research.

Footnotes

Institutional review board approval of the study protocol was obtained from the University of Florida and Choctaw Nation of Oklahoma.

De-identified patient electronic health record data, with the exception of zip codes, was used for this study.

The authors have no conflicts of interest to report.

References

1. Ali MK, McKeever Bullard K, Imperatore G, Barker L, & Gregg EW (2012). Characteristics associated with poor glycemic control among adults with self-reported diagnosed diabetes—National Health and Nutrition Examination Survey, United States, 2007–2010. MMWR Mobidity Mortality Weekly Report, 61, 32–37. https://www.cdc.gov/mmwr/preview/mmwrhtml/su6102a6.htm [Google Scholar]
2. Boye KS, Lage MJ, Shinde S, Thieu V, & Bae JP (2021). Trends in HbA1c and body mass index among individuals with type 2 diabetes: Evidence from a US database 2012–2019. Diabetes Therapy, 12, 2077–2087. 10.1007/s13300-021-01084-0
   [DOI] [PMC free article] [PubMed] [Google Scholar]
3. Bureau of Indian Affairs. (n.d.). About us. https://www.bia.gov/about-us
4. Carroll SR, Suina M, Jäger MB, Black J, Cornell S, Gonzales AA, Jorgensen M, Palmanteer-Holder NL, De La Rosa JS, & Teufel-Shone NI (2022). Reclaiming Indigenous health in the US: Moving beyond the social determinants of health. International Journal of Environmental Research and Public Health, 19, 7495. 10.3390/ijerph19127495
   [DOI] [PMC free article] [PubMed] [Google Scholar]
5. Casey JA, Pollak J, Glymour MM, Mayeda ER, Hirsch AG, & Schwartz BS (2018). Measures of SES for electronic health record-based research. American Journal of Preventive Medicine, 54, 430–439. 10.1016/j.amepre.2017.10.004
   [DOI] [PMC free article] [PubMed] [Google Scholar]
6. Centers for Disease Control and Prevention. (2026). National diabetes statistics report. https://www.cdc.gov/diabetes/php/data-research/index.html
7. Choctaw Nation of Oklahoma. (n.d.-a). About: Health. https://www.choctawnation.com/about/health/
8. Choctaw Nation of Oklahoma. (n.d.-b). About the Choctaw Nation. https://www.choctawnation.com/about/
9. Choctaw Nation of Oklahoma. (n.d.-c). Family medicine residency program. https://www.choctawnation.com/services/residency/
10. Dubowitz N, Xue W, Long Q, Ownby JG, Olson DE, Barb D, Rhee MK, Mohan AV, Watson-Williams PI, Jackson SL, Tomolo AM, Johnson TM II, & Phillips LS (2014). Aging is associated with increased HbA1c levels, independently of glucose levels and insulin resistance, and also with decreased HbA1c diagnostic specificity. Diabetic Medicine, 31, 927–935. 10.1111/dme.12459
    [DOI] [PubMed] [Google Scholar]
11. Dugani SB, Mielke MM, & Vella A (2021). Burden and management of type 2 diabetes in rural United States. Diabetes/Metabolism Research and Reviews, 37, e3410. 10.1002/dmrr.3410
    [DOI] [PMC free article] [PubMed] [Google Scholar]
12. ElSayed NA, Aleppo G, Aroda VR, Bannuru RR, Brown FM, Bruemmer D, Collins BS, Hilliard ME, Isaacs D, Johnson EL, Kahan S, Khunti K, Leon J, Lyons SK, Perry ML, Prahalad P, Pratley RE, Seley JJ, Stanton RC, … Gabbay RA (2023). 6. Glycemic targets: Standards of care in diabetes—2023. Diabetes Care, 46, S97–S110. 10.2337/dc23-S006
    [DOI] [PMC free article] [PubMed] [Google Scholar]
13. Findling MG, Casey LS, Fryberg SA, Hafner S, Blendon RJ, Benson JM, Sayde JM, & Miller C (2019). Discrimination in the United States: Experiences of native Americans. Health Services Research, 54, 1431–1441. 10.1111/1475-6773.13224
    [DOI] [PMC free article] [PubMed] [Google Scholar]
14. Garvey WT, Cohen RM, Butera NM, Kazemi EJ, Younes N, Rosin SP, Suratt CE, Ahmann A, Hollander PA, Krakoff J, Martin CL, Seaquist E, Steffes MW, Lachin JM, & GRADE Research Group (2024). Association of baseline factors with glycemic outcomes in GRADE: A comparative effectiveness randomized clinical trial. Diabetes Care, 47, 562–570. 10.2337/dc23-1782
    [DOI] [PMC free article] [PubMed] [Google Scholar]
15. Gillson SL, Hautala D, Sittner KJ, & Walls M (2024). Historical trauma and oppression: Associations with internalizing outcomes among American Indian adults with type 2 diabetes. Transcultural Psychiatry, 61, 372–384. 10.1177/13634615221079146
    [DOI] [PMC free article] [PubMed] [Google Scholar]
16. Goins RT, Grant MK, Conte KP, & Lefler L (2022). Social support and diabetes management among older American Indians. Frontiers in Public Health, 10, 780851. 10.3389/fpubh.2022.780851
    [DOI] [PMC free article] [PubMed] [Google Scholar]
17. Gopalan A, Mishra P, Alexeeff SE, Blatchins MA, Kim E, Man A, Karter AJ, & Grant RW (2020). Initial glycemic control and care among younger adults diagnosed with type 2 diabetes. Diabetes Care, 43, 975–981. 10.2337/dc19-1380
    [DOI] [PMC free article] [PubMed] [Google Scholar]
18. Grohmann B, Espin S, & Gucciardi E (2017). Patients’ experiences of diabetes education teams integrated into primary care. Canadian Family Physician Medecin de famille Canadien, 63, e128–e136.
    [PMC free article] [PubMed] [Google Scholar]
19. James TL (2021). Improving referrals to diabetes self-management education in medically underserved adults. Diabetes Spectrum, 34, 20–26. 10.2337/ds20-0001
    [DOI] [PMC free article] [PubMed] [Google Scholar]
20. Jewett PI, Gangnon RE, Elkin E, Hampton JM, Jacobs EA, Malecki K, LaGro J, Newcomb PA, & Trentham-Dietz A (2018). Geographic access to mammography facilities and frequency of mammography screening. Annals of Epidemiology, 28, 65–71. 10.1016/j.annepidem.2017.11.012
    [DOI] [PMC free article] [PubMed] [Google Scholar]
21. Kautzky-Willer A, Leutner M, & Harreiter J (2023). Sex differences in type 2 diabetes. Diabetologia, 66, 986–1002. 10.1007/s00125-023-05891-x
    [DOI] [PMC free article] [PubMed] [Google Scholar]
22. Kelly C, Hulme C, Farragher T, & Clarke G (2016). Are differences in travel time or distance to healthcare for adults in global north countries associated with an impact on health outcomes? A systematic review. BMJ Open, 6, e013059. 10.1136/bmjopen-2016-013059 [DOI] [Google Scholar]
23. Kushner PR, Cavender MA, & Mende CW (2022). Role of primary care clinicians in the management of patients with type 2 diabetes and cardiorenal diseases. Clinical Diabetes, 40, 401–412. 10.2337/cd21-0119
    [DOI] [PMC free article] [PubMed] [Google Scholar]
24. Maganty A, Byrnes ME, Hamm M, Wasilko R, Sabik LM, Davies BJ, & Jacobs BL (2023). Barriers to rural health care from the provider perspective. Rural and Remote Health, 23, 7769. 10.22605/RRH7769
    [DOI] [PubMed] [Google Scholar]
25. Medicaid and CHIP Payment and Access Commission. (2021). Medicaid’s role in health care for American Indians and Alaska Natives. https://www.macpac.gov/wp-content/uploads/2021/02/Medicaids-Role-in-Health-Care-for-American-Indians-and-Alaska-Natives.pdf
26. Mondesir FL, White K, Liese AD, & McLain AC (2016). Gender, illness-related diabetes social support, and glycemic control among middle-aged and older adults. Journals of Gerontology. Series B: Psychological Sciences and Social Sciences, 71, 1081–1088. 10.1093/geronb/gbv061
    [DOI] [PubMed] [Google Scholar]
27. Nicklett EJ, Omidpanah A, Whitener R, Howard BV, & Manson SM (2017). Access to care and diabetes management among older American Indians with type 2 diabetes. Journal of Aging and Health, 29, 206–221. 10.1177/0898264316635562
    [DOI] [PMC free article] [PubMed] [Google Scholar]
28. Oklahoma Health Care Authority. (2021). SoonerCare fast facts. https://oklahoma.gov/content/dam/ok/en/okhca/docs/research/data-and-reports/fast-facts/2021/december/TotalEnrollment12_21.pdf
29. Oré CE, Loerzel E, Marziale E, & Parker M (2025). Developing Indigenous-centered healing, health, and wellness frameworks to strengthen indigenous health systems, decolonize public health, and achieve health equity. American Journal of Public Health, 115, 726–731. 10.2105/AJPH.2024.307958
    [DOI] [PMC free article] [PubMed] [Google Scholar]
30. Patel N, Swami J, Pinkhasova D, Karslioglu French E, Hlasnik D, Delisi K, Donihi A, Siminerio L, Rubin DJ, Wang L, & Korytkowski MT (2022). Sex differences in glycemic measures, complications, discharge disposition, and postdischarge emergency room visits and readmission among non-critically ill, hospitalized patients with diabetes. BMJ Open Diabetes Research & Care, 10, e002722. 10.1136/bmjdrc-2021-002722 [DOI] [Google Scholar]
31. Sherwani SI, Khan HA, Ekhzaimy A, Masood A, & Sakharkar MK (2016). Significance of HbA1c test in diagnosis and prognosis of diabetic patients. Biomarker Insights, 11, 95–104. 10.4137/BMI.S38440
    [DOI] [PMC free article] [PubMed] [Google Scholar]
32. Shrishrimal K, Hart P, & Michota F (2009). Managing diabetes in hemodialysis patients: Observations and recommendations. Cleveland Clinic Journal of Medicine, 76, 649–655. 10.3949/ccjm.76a.09054
    [DOI] [PubMed] [Google Scholar]
33. Strauss K, MacLean C, Troy A, & Littenberg B (2006). Driving distance as a barrier to glycemic control in diabetes. Journal of General Internal Medicine, 21, 378–380. 10.1111/j.1525-1497.2006.00386.x
    [DOI] [PMC free article] [PubMed] [Google Scholar]
34. Sun S, Hisland L, Grenet G, Gueyffier F, Cornu C, Jaafari N, & Boussageon R (2022). Reappraisal of the efficacy of intensive glycaemic control on microvascular complications in patients with type 2 diabetes: A meta-analysis of randomised control-trials. Therapies, 77, 413–423. 10.1016/j.therap.2021.10.002 [DOI] [Google Scholar]
35. Tatulashvili S, Fagherazzi G, Dow C, Cohen R, Fosse S, & Bihan H (2020). Socioeconomic inequalities and type 2 diabetes complications: A systematic review. Diabetes & Metabolism, 46, 89–99. 10.1016/j.diabet.2019.11.001
    [DOI] [PubMed] [Google Scholar]
36. U.S. Census Bureau. (n.d.). QuickFacts Oklahoma. https://www.census.gov/quickfacts/fact/table/OK/PST045222
37. U.S. Department of Health and Human Services, Office of Minority Health. (2026). Diabetes and American Indians/Alaska natives. https://minorityhealth.hhs.gov/diabetes-and-american-indiansalaska-natives
38. Weiss AJ, Pickens G, & Roemer M (2021). Methods for calculating patient travel distance to hospital in HCUP data. HCUP Methods Series Report US Agency for Healthcare Research and Quality. https://www.hcup-us.ahrq.gov/reports/methods/methods.jsp [Google Scholar]
39. Wolfe MK, McDonald NC, & Holmes GM (2020). Transportation barriers to health care in the United States: Findings from the National Health Interview Survey, 1997–2017. American Journal of Public Health, 110, 815–822. 10.2105/AJPH.2020.305579
    [DOI] [PMC free article] [PubMed] [Google Scholar]
40. World Health Organization. (2019). International statistical classification of diseases and related health problems (10th rev.). https://icd.who.int/browse10/2019/en#/
41. Wouk N (2021). End-stage renal disease: Medical management. American Family Physician, 104, 493–499.
    [PubMed] [Google Scholar]
42. Zakaria NI, Tehranifar P, Laferrère B, & Albrecht SS (2023). Racial and ethnic disparities in glycemic control among insured US adults. JAMA Network Open, 6, e2336307. 10.1001/jamanetworkopen.2023.36307
    [DOI] [PMC free article] [PubMed] [Google Scholar]
43. Zgibor JC, Gieraltowski LB, Talbott EO, Fabio A, Sharma RK, & Hassan K (2011). The association between driving distance and glycemic control in rural areas. Journal of Diabetes Science and Technology, 5, 494–500. 10.1177/193229681100500304
    [DOI] [PMC free article] [PubMed] [Google Scholar]
44. Zhang HZ (2024). Age at type 2 diabetes diagnosis and the risk of mortality among US population. Scientific Reports, 14, 29155. 10.1038/s41598-024-80790-8
    [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

SDC