The Impact of the COVID-19 Pandemic on Ethnic Minority Groups With Diabetes

A significant body of evidence indicates that many of the most vulnerable in our society have faced the greatest burden of both direct and indirect consequences of the coronavirus disease 2019 (COVID-19) pandemic, including the elderly, people living in deprived areas, people with learning difficulties, the homeless, and migrants. In particular, ethnic minority populations in many European countries and North America are at greater risk of infection and severe outcomes due to COVID-19 (1–3), although recent data on later waves and variants suggest some of these disparities have reduced for some minority groups (4).

There is now strong evidence that having a long-term condition, such as diabetes, increases risk of severe COVID-19 outcomes. Due to the greater prevalence of comorbidities in some ethnic minority populations (5), this may be one factor explaining outcome disparities between ethnic groups (Fig. 1). However, despite the large volume of literature and many reviews focusing on the association between ethnicity and COVID-19, there is, to our knowledge, far less attention paid to examining differences in COVID-19 outcomes by ethnic groups in patients with diabetes.

Therefore, this Perspective summarizes the literature on prepandemic ethnic disparities in diabetes care as well as the wider context of disrupted diabetes care during the pandemic. It also provides an overview of the limited literature on ethnic disparities in COVID-19 in people with diabetes. Finally, this article concludes with a summary of potential explanatory mechanisms for ethnic disparities in COVID-19, interventions implemented to improve diabetes outcomes during the pandemic, structural determinants that require addressing, and recommendations for future research.

Prior to the COVID-19 pandemic, numerous studies had reported major ethnic and socioeconomic disparities in diabetes care, particularly for intermediate outcomes and diabetes complications. For patients with type 1 diabetes, inequalities in access to technology and insulin pumps also exist. The combination of these ethnic disparities in diabetes care and outcomes has been a significant contributor to the predictable and undesirable inequity in COVID-19 outcomes for people with diabetes.

Internationally, ethnic and socioeconomic disparities in intermediate diabetes outcomes have been repeatedly demonstrated. For example, a 2013 U.S. study found that the Hispanic population was less likely to receive five diabetes quality-of-care recommendations, including twice-yearly HbA_1c tests and yearly foot exam, dilated eye exam, cholesterol test, and flu vaccination (6). Black patients were also less likely to receive HbA_1c tests, eye exams, and flu vaccination, and Asian patients were less likely to receive HbA_1c tests and annual foot and eye examinations. These differences were eliminated after adjusting for insurance status and education.

Similarly, in a nationally representative study of 164 English primary care practices, Black patients and people from the lowest socioeconomic status (SES) groups had higher HbA_1c than White patients and people from the highest SES groups (7). In a systematic review of 24 studies, Black patients were ∼30% less likely to have HbA_1c and blood pressure control but were ∼20–50% more likely to receive eye or foot examinations than White patients (8). A high school degree or higher was associated with a greater likelihood of HbA_1c control and eye examination. Other social determinants of health (i.e., income, access to care, and food insecurity) were also related, albeit more weakly, to these intermediate outcomes.

The results of disparities in diabetes quality of care and intermediate outcomes have translated into well-characterized ethnic and socioeconomic disparities in eye, kidney, and cardiovascular disease. Black, Latino, and Asian patients all have higher rates of diabetic eye disease than non-Hispanic White patients (8). Additionally, Black, Latino, Asian, Filipino, and mixed-ethnicity patients all have been found to have higher rates of end-stage renal disease than non-Hispanic White patients within the Kaiser Permanente Northern California integrated health care delivery system (9). Furthermore, in a 2020 study of people with type 2 diabetes, South Asian patients had a higher cardiovascular disease risk than people of White ethnicity (10).

Studies in China, Sweden, and the U.S. have identified an association between low SES and increased diabetes complication rates. For example, in a Chinese cohort of over 25,000 patients with type 2 diabetes, the least educated population groups had the highest risk for cardiovascular disease, cerebrovascular disease, and retinopathy, and the lowest household income population had the highest rate of retinopathy and neuropathy (11). In a study of over 24,000 Swedish patients with type 1 diabetes, similarly, the risk of death was three times higher among patients in the two lower income quintiles compared with the highest income quintile, and the risk for cardiovascular disease and stroke was twofold (12). Higher education was also associated with a lower risk of stroke. In the U.S., risk of death was ∼1.5 times greater among those with the lowest education level and without measures of financial wealth, such as home ownership (13). Importantly, accounting for U.S. socioeconomic differences by ethnicity narrows the excess mortality attributed to diabetes for White and Hispanic populations but not for non-Hispanic Black populations (14).

Data from England suggest there was between a 74% and 88% reduction in care processes during the pandemic compared with prepandemic levels (15). Other work has also suggested reduced consultation rates and hospital admissions, 30% reduction in diabetes diagnoses, and 70% reduction in HbA_1c testing during the pandemic (16). Older patients from deprived areas experienced the largest reductions in health checks, but high-quality data on the differences by ethnic groups are lacking.

The few studies that have examined the impact of the pandemic on diabetes care have suggested that there have been no ethnic disparities in diabetes care. For example, in the U.K., a study of 618,161 people with type 2 diabetes from 1,744 general practices identified similar lower rates of performing health checks and prescribing diabetes medications by ethnicity (15). Interestingly, a study of a large academic U.S. health system reported a higher rate of eye exams among Black, Asian, and Hispanic patients with diabetes compared with White patients and no difference in HbA_1c and nephropathy testing (17). Of note, a major shift to telemedicine occurred during the pandemic, and one study of 1,292 patients did not identify ethnic disparities in telemedicine use for subspecialty diabetes care but did report lower rates of telemedicine use among people with a non-English primary language and older adults (18).

Diabetes is a risk factor for severe COVID-19, including intensive care unit (ICU) admission and death (19). However, despite multiple publications reporting the association of ethnicity with COVID-19 outcomes in ethnic minority populations, very few studies have examined differences in COVID-19 outcomes by both ethnicity and diabetes (Table 1). The largest is a National Population Database study, which included 264,390 people with type 1 diabetes and 2,874,720 people with type 2 diabetes in England (20). Overall, in individuals with type 1 diabetes, South Asian, Black, mixed, and other ethnic groups had significantly higher mortality risk from COVID-19 versus White populations (20). For people with type 2 diabetes, risk of in-hospital mortality was greater for Asian, Black, and mixed ethnic groups but not for the other ethnic groups (20). The other ethnic groups included Vietnamese, Japanese, Filipino, Malaysian, and other ethnicities.

Another U.K. study of over 19,000 COVID-19 high-dependency-unit and ICU admissions found that of 3,524 patients with diabetes, 58.5% were White, 21.9% were Asian, 11.6% were Black, and 8% were mixed or other ethnicity. There was no evidence of a difference in COVID-19 mortality associated with type 2 diabetes in subgroups defined by ethnicity (21). Other U.K. data include a small, single-center study of 39 patients with end-stage renal disease secondary to diabetic kidney disease, which reported a higher prevalence of patients of African-Caribbean ethnicity hospitalized with COVID-19 (60%) versus White or other ethnic groups (22).

In a multisite, cross-sectional, observational study of 113 people in the U.S. with type 1 diabetes (n = 58 hospitalized), people of ethnic minority background (Black, Hispanic, and other) with confirmed COVID-19 were, on average, 3.63 times more likely to be hospitalized than non-Hispanic White patients (23). The same research group also subsequently reported a cohort of 180 patients with type 1 diabetes and laboratory-confirmed COVID-19, and they found that Black patients were significantly more likely to present with diabetic ketoacidosis (DKA) than non-Hispanic White patients (24). Additionally, in a multisite prospective study of 137 service locations and 313 patients with type 1 and type 2 diabetes, compared with the White group, the Black group was significantly associated with hospitalization and worsening illness severity within 114 days of a positive COVID-19 test (25). However, a retrospective study of 6,104 people with type 2 diabetes from an ethnically diverse southern U.S. sample found 239 (39%) tested positive for COVID-19, but there were no significant differences in mortality between Black and White patients (26).

In a retrospective study from the University of Michigan, among those with diabetes, in the whole cohort Black patients were significantly more likely to be tested for COVID-19 and have positive test results than White patients; however, hospitalization, ICU admission, and mortality were not different for White and Black COVID-19 patients with type 2 diabetes (5). Another multisite study of 4,413 COVID-19 patients with type 2 diabetes admitted to New York hospitals reported that most patients were non-White (67.3%), with 27.1% Black, 17.2% Hispanic, and 24.9% multiethnic or other ethnicity (27). There was no significant association between ethnicity and mortality (27). Similarly, a nationwide retrospective cohort study in the U.K. (n = 19,256) reported no difference in mortality associated with type 2 diabetes in subgroups defined by ethnicity (21).

Overall, results from the few studies of COVID-19 in ethnic minority populations with diabetes are mixed, but there is some suggestion that the rates of hospitalization and mortality are higher than those of White populations. Although a number of larger population-level studies have been performed, these were all conducted in the U.K. or U.S., and there is a need for evidence from other countries, particularly lower- and middle-income countries. In addition, further work is needed to examine disparities for more granular ethnic groups (e.g., breaking down South Asian into Indian, Pakistani, etc.), although efforts may be hampered by poor ethnicity coding in routine health care data in many countries.

Ethnic disparities in rate, severity, and mortality from COVID-19 infection were evident early in the pandemic. Understanding the impact of ethnicity on COVID-19–related outcomes was initially hampered by a lack of data reporting patient ethnicity. However, with a greater appreciation, as the pandemic evolved, of associations of ethnicity with poorer COVID-19 outcomes, studies are now shedding light on the potential factors predisposing ethnic minorities to poorer clinical COVID-19 outcomes. The exact reasons for the higher incidence and severity of COVID-19–related outcomes in minority ethnic groups are complex and involve social, economic, cultural, and lifestyle factors and pathophysiological differences (28). These parameters then modulate, within ethnic populations, exposure to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), vulnerability to the disease, social environment, control measures, and biological differences (Fig. 1).

Greater exposure to the virus raises the likelihood individuals will contract COVID-19. During the pandemic, workers with essential occupations had increased frequency of contact with other individuals, for example, workers in health care settings, those who work indoors in crowded spaces, or workers making frequent contact with multiple people, such as those in transport and retail. Indeed, ethnic minority groups were overrepresented in a nationwide investigation of occupations classified as essential during the pandemic (29). This included medical staff, such as registered nurses, medical assistants, and bus drivers. A study of frontline health care workers (n = 99,795) versus the general community (n = 2,035,395) found higher risk of a positive COVID-19 test result for frontline health care workers (hazard ratio [HR] 3.40) after adjusting for multiple parameters, including their elevated likelihood of receiving a COVID-19 test (30). Secondary post hoc analyses found that frontline minority workers are at increased risk of a positive COVID-19 test compared with White coworkers. In New York City (NYC), frontline workers were overrepresented in Black (29.4%) and Hispanic (35.5%) populations (31). Therefore, one possible reason predisposing ethnic minorities to poorer outcome is greater exposure to the virus through overrepresentation of these groups in essential frontline jobs.

Another potential contributing factor to adverse COVID-19 outcomes, which was considered relatively early in the pandemic, is the high prevalence of comorbid conditions in ethnic minority populations (32), which is associated with increased hospitalization rate, ICU admission, and mortality. A study of UK Biobank data with 5,623 COVID-19 cases found that BMI was associated with a positive test in all patients, but that higher BMI (>30 kg/m²) was a stronger determinant of a positive test in Black and minority ethnic patients than in White patients (33). Evaluation of NYC data found a relationship in the two boroughs with the highest COVID-19 mortality rate and the highest obesity rate, which was most prevalent among Black and Hispanic residents (34). Notably, however, the study did not fully adjust for all potential risk factors. Another NYC study found that poorer neighborhoods had greater hypertension risk among Black residents (relative risk [RR] 3.4) and diabetes risk among Hispanic residents (RR 5.5) compared with residents of poorer White neighborhoods (31). Furthermore, poorer predominantly Black and Hispanic neighborhoods had more ICU admissions than poorer predominantly White neighborhoods. In addition, a nationwide U.S. county-level investigation found, after multiple adjustments, that counties with higher percentages of Black residents had a greater prevalence of comorbid conditions and COVID-19 diagnoses (RR 1.24) and deaths (RR 1.18) (35).

Beyond potential health determinants, such as comorbidities, on COVID-19 outcomes for Black and Hispanic patients, several socioeconomic factors have been investigated. Urban setting, household crowding, income, lack of transportation, health insurance availability, and health care access could conceivably impact COVID-19 outcomes (29–31,35,36), although there is some discordance across studies (37). These determinants may also be differentially present by ethnicity, although few COVID-19 studies have been stratified by ethnicity.

Analysis of NYC electronic health records (n = 23,300) found a correlation between social disadvantage (compared by quintiles) and higher likelihood of Black or Hispanic ethnicity, of chronic comorbid conditions (e.g., diabetes, obesity, and hypertension), and higher hospitalization risk from COVID-19 (38). County-level analysis also found, after multiple adjustments, correlations between counties with greater burden of COVID-19 cases and the proportion of Black and Hispanic residents and counties with greater burden of COVID-19 deaths and the proportion of Black and Native American residents (36). A greater proportion of multiunit households and households lacking a vehicle correlated positively with cases and deaths, respectively. Another study also found a relationship between public transportation use (i.e., household lacking vehicles) (39) and mean household size (40) with COVID-19 cases, which are likely contributors in disadvantaged communities. In a nationwide U.S. county-level study, lack of health insurance was a greater risk factor for COVID-19 diagnosis in counties with a percentage of Black residents above the national average (35).

In addition to housing, transportation, income, and insurance status, opioid and other substance use disorders are contributors to adverse COVID-19 outcomes in the U.S. Nationwide, Black individuals with opioid use disorders (odds ratio [OR] 4.16) and substance use disorders (OR 2.17) were at higher COVID-19 risk than White individuals (41), although the analysis did not adjust for comorbidities.

Control measures are used to limit viral spread in the community through awareness campaigns. However, ethnic minorities may experience language barriers, lower access to public health messaging, poor uptake or access to screening, and structural discrimination, leading to suboptimal control measures. County-level analysis by a U.S. study found that individuals with limited English proficiency were more likely to suffer COVID-19 deaths after multiple adjustments (36). In a nationwide study, ethnic minority status and language subindex correlated positively with COVID-19 incidence and mortality rates after adjusting for population density, urbanicity, and COVID-19 testing rate (42). Regarding uptake of public health messaging, a cross-sectional U.S. survey of 1,435 adults found that non-Hispanic White (30.7%) and Asian (25.0%) participants were likelier to correctly answer all 14 questions concerning COVID-19 versus Hispanic (19.7%) and non-Hispanic Black (15.8%) participants (43). Thus, it is possible that preventative control measures through public health campaigns are failing in ethnic minority communications.

An additional possible predisposing factor is medical mistrust, which can result from prior experience of unethical medical practices. Though understandable, this mistrust can represent a barrier to seeking medical care in the present. In a California cohort of Black individuals with HIV (n = 101), most (97%) held at least one general COVID-19 mistrust belief and over 50% held COVID-19 vaccine or treatment hesitancy beliefs (44). In a large U.K. study of 32,361 adults in the University College London COVID-19 Social Study, distrust of vaccines was higher among people of ethnic minority backgrounds than among White people (45). Therefore, overall, in addition to greater exposure and susceptibility, people from ethnic minorities may also be predisposed to more serious COVID-19 outcomes due to poorer control measures.

Finally, it is important to note that although systematic reviews and meta-analyses associate ethnicity with poorer COVID-19 outcomes, some studies report no disparities or even protective effects (46–49). Nevertheless, in studies that do identify risk in Black, Hispanic, or other minority ethnic groups, this risk persists even after adjusting for comorbidities and socioeconomic factors (2,50–53). This suggests there are biological differences underlying differential responses of ethnicity to COVID-19. One possibility, given its pivotal role in promoting viral entry into host cells, is differential ACE2 expression by ethnicity (54) as well as of other molecules related to SARS-CoV-2 pathophysiology, e.g., immune function. Another possible aspect is single nucleotide polymorphisms in genes involved in SARS-CoV-2 infection, such as TMPRSS2, in addition to ACE2 (55,56). Caution in interpretation is required, however, because these potential mechanisms require further work. It is also possible that plausible biological mechanisms will be politicized to minimize criticisms of structural racism, which underpins social inequalities driving long-standing health inequalities in ethnic minority populations.

Overall, emerging evidence implicates multiple potential determinants of ethnic disparity in COVID-19 outcomes. Moving forward, a better understanding will be facilitated by stratifying study data sets by ethnicity (the most granular ethnic group data should be used when available). Additionally, most studies to date have been correlative. Although it will be challenging, future investigations should establish causality in order to implement evidence-based policies to alleviate disparities.

As we now plan for recovery, improved surveillance, and risk factor management, it will be imperative that primary and specialist care services urgently address the disproportionate impact the pandemic has had on ethnic minority groups. Recommended interventions for diabetes patients during the pandemic, which has disproportionately impacted ethnic minority populations, focused on individual-level strategies and barriers to diabetes care (57–60) but excluded the larger structural barriers, which placed ethnic minorities with diabetes at particular risk for poor COVID-19 outcomes from the outset. These barriers include structural inequities in adequate housing, food, education, employment opportunities, and neighborhood resources, among others, which are important determinants of health for both diabetes and COVID-19 independently but particularly for individuals in high-risk populations within both groups (e.g., ethnic minorities).

Prioritizing affordable housing, a living wage, economic investments in ethnic minority communities, and policy changes to mitigate ethnic residential segregation are some long-term strategies to address the underlying housing and related economic challenges faced by ethnic minorities, particularly low-income minorities. Implementing these changes would strengthen the resilience of this population in weathering pandemics and natural disasters as well as in managing chronic diseases, such as diabetes, which require stable housing, refrigeration (for many types of insulin storage), and safe spaces for physical activity as part of optimal self-management.

Structural interventions, like equal access to high-quality education and employment opportunities, implementing a living wage, and job retraining for those impacted by the changing economy or who are reentering the workforce, can affect the short-term and long-term health trajectories of ethnic minority individuals with diabetes. Regular employment usually provides access to health insurance (for U.S. residents), and the use of supporting federal policies to provide paid sick leave and quarantine leave for COVID-19 is another way, along with pandemic mitigation, to ensure individual assistance reaches patients.

The pandemic also impacted food insecurity, which saw a large rise during this period. Before the pandemic, over 35 million Americans faced food insecurity, which rose to over 50 million, disproportionately affecting families and children (61). Food insecurity also disproportionately impacts ethnic minorities, low-income individuals, and other socially marginalized populations. Previous studies showed that for patients with diabetes, food insecurity is associated with worse diabetes self-management, up to twice the odds of poor diabetes control, and higher hospitalization rates (62). Supporting economic policies that reduce poverty and food policies that reduce food insecurity, such as increased eligibility for the Supplemental Nutrition Assistance Program in the U.S., will help improve diabetes outcomes.

As we move into the recovery phase, postpandemic research should focus on developing and evaluating interventions to address these disparities in the short and long term for all ethnic minority populations with long-term conditions such as diabetes. This should involve both individual-level and wider systems-level approaches. In addition, a number of other priorities require further attention. Given limited published reports, there is a need to utilize large population cohorts to further examine disparities in COVID-19 outcomes in ethnic minority populations with diabetes. It will also be important to further examine the impact of the pandemic on diabetes care for ethnic minorities, as limited work has been done in this area. Other outstanding issues include whether there are therapeutic differences in patients from ethnic minority populations with diabetes who were admitted with COVID-19 or who are positive for COVID-19 compared with White populations and if there are differential effects of therapies in ethnic minority groups with diabetes. Are new models of care, such as telemedicine, widening health disparities for ethnic minority populations with diabetes?

In summary, the structural inequities highlighted by the coronavirus pandemic, which have driven many of the disparities in COVID-19 morbidity and mortality, have exacerbated existing health disparities in diabetes among ethnic minority populations. While there have been many helpful guidelines and approaches for diabetes self-management as well as outpatient and inpatient care during the pandemic, few have recommended addressing structural drivers. Only by taking a long-term, holistic view of health and health care will we, and particularly our most vulnerable populations, be better able to weather future pandemics.

Funding. K.K. is supported by the National Institute for Health Research (NIHR) Applied Research Collaboration East Midlands and the NIHR Leicester Biomedical Research Centre. E.L.F. is supported by the National Institutes of Health (NIH) National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) (U01DK119083), the NeuroNetwork for Emerging Therapies, the NeuroNetwork Therapeutic Discovery Fund, the A. Alfred Taubman Medical Research Institute, and the Sinai Medical Staff Foundation. N.L. is supported by NIH NIDDK (2P30DK092949), National Institute on Aging (R01 AG063391), and National Institute on Minority Health and Health Disparities (R01 MD013420). A.R. is supported by the NIHR Applied Research Collaboration East Midlands. M.P. is supported by NIH NIDDK (2P30DK092949, R01DK124597, R01DK127961, and R21DK121262), National Institute on Minority Health and Health Disparities (R01 MD013420), National Heart, Lung, and Blood Institute (1OT2HL156812), and the Merck Foundation. The views expressed are those of the authors and not necessarily those of the NIHR, NHS, or the Department of Health and Social Care.

Duality of Interest. K.K. has acted as a consultant or speaker or received grants for investigator-initiated studies for AstraZeneca, Novartis, Novo Nordisk, Sanofi, Lilly, Merck Sharp & Dohme, Boehringer Ingelheim, Bayer, Berlin-Chemie AG/Menarini Group, Janssen, and Napp Pharmaceuticals. K.K. is director of the University of Leicester Centre for Black Minority Ethnic Health, trustee of the South Asian Health Foundation, chair of the Ethnicity Subgroup of the U.K. Scientific Advisory Group for Emergencies (SAGE), and member of SAGE. M.P. has acted as a consultant or speaker for Bayer Pharmaceuticals, CME Outfitters, and PRIME in the past 3 years. M.P. is currently or has been a consultant or advisory board member for the Agency for Healthcare Research and Quality, American Diabetes Association, the American College of Physicians, the Centers for Disease Control and Prevention, the Centers for Medicare and Medicaid Services, the Illinois Department of Health, the Society of General Internal Medicine, and Physicians for Human Rights in the past 3 years. No other potential conflicts of interest relevant to this article were reported.