Prevalence of Diagnosed Diabetes Among U.S. Adults Aged ≥18 Years With Disabilities, 2021–2022

According to U.S. national data, one in four adults has some type of disability (1), placing them at increased risk of chronic diseases. In fact, for people with disabilities higher rates are consistently reported of diabetes (2), obesity (3), and cardiovascular disease (2), along with risk factors for chronic conditions, such as physical inactivity (3). Although some differences in rates of chronic diseases may be related to the underlying condition leading to the disability, others may be related to complications of the condition. For diabetes specifically, some disabilities have a bidirectional relationship with diabetes. For example, mobility impairment has been associated with incident diabetes (4) and adults with diabetes are at increased risk of mobility disability (5). Other disabilities such as vision loss (6), hearing loss (6), impaired cognition (7), and difficulty with self-care (8) may be indirectly linked to onset of diabetes. And conversely, through various mechanisms, adults with diabetes are at increased risk of vision loss (9), hearing loss (10), cognitive decline (11), and difficulty with self-care (5). Communication disabilities, defined as “difficulties in understanding or being understood by others when speaking in one’s usual language” (12), can result from other disabilities such as hearing loss (13) or diabetes-related complications such as stroke.

Moreover, adults with disabilities are disproportionately represented in underserved sociodemographic groups (e.g., non-Hispanic [NH] Black individuals, Hispanic individuals, individuals with less than a high school education, and individuals living below the federal poverty level [FPL]) (14–16) who are also at increased risk for diabetes. Adults with diabetes and disability incur increased health care costs, have increased health service use (including hospitalizations), and experience more barriers to health care access (17), so surveillance of diabetes prevalence among adults with disabilities has important clinical and public health implications.

In 2006, the United Nations Convention on the Rights of Persons with Disabilities mandated that countries that ratified the convention provide internationally comparable data on people with disabilities, using the Washington Group measures (18). The Washington Group Short Set on Functioning (WG-SS) indicator has been extensively tested and validated across the world to ensure its accuracy and universality (19). Although studies have shown the prevalence of various types of diabetes-related disability for years (20), such studies have largely included reliance on measures of disability with assessment of the different domains of disability individually (2,5–7) or in various geographic areas (6), or included use of the American Community Survey (ACS) measures (2,21). To our knowledge, prevalence of diabetes has not been comprehensively reported from studies among adults with multiple or distinct domains of disability with use of the WG-SS at the U.S. national level. The objective of this study is to use the WG-SS disability questions to describe and compare the prevalence of diagnosed diabetes among adults with and without disabilities, with two different measures of disability including a graded functioning severity measure, overall and by sociodemographic characteristics.

We conducted analyses with public use data from the 2021–2022 National Health Interview Survey (NHIS), an annual cross-sectional household survey that is representative of the noninstitutionalized civilian U.S. population. NHIS has a complex sample design, and data are collected on sociodemographics and health topics through in-person interviews. NHIS sample sizes for adults aged ≥18 years were 29,482 (in 2021) and 27,651 (in 2022). The Sample Adult (i.e., respondent) response rate in 2021 was 50.9% and in 2022 was 47.7% (22). To account for small samples of adults with diabetes and disabilities, we combined data for 2021 and 2022. Weights were divided by 2 to account for pooling. A total of 62 respondents were excluded for missing data on any or all of the disability or diabetes variables, resulting in an analytic sample of 57,071 respondents aged ≥18 years. This analysis of NHIS deidentified public use data was considered not to be human subject research and did not require review by Centers for Disease Control and Prevention Institutional Review Board.

Diagnosed diabetes was identified with the survey question, “Other than during pregnancy, has a doctor or other health professional ever told you that you had diabetes?” Disability survey questions include six functional domains: vision, hearing, mobility, cognition, self-care, and communication:

• 1.Do you have difficulty seeing, even when wearing glasses?

• 2.Do you have difficulty hearing, even when using a hearing aid?

• 3.Do you have difficulty walking or climbing stairs?

• 4.Do you have difficulty remembering or concentrating?

• 5.Do you have difficulty with self-care, such as washing all over or dressing?

• 6.Using your usual language, do you have difficulty communicating, understanding, or being understood?

The WG-SS has four response categories reflecting a continuum of functional difficulty: none, some, a lot, and cannot do at all. Information regarding the methodology and validation of the disability domains has previously been published (18). Though the WG-SS does not assign text labels to the four numeric categories, to ease reporting for this study, we assigned labels as follows indicating category placement on the continuum of severity: Sample Adults who indicated “some” difficulty were classified as having “milder” disability, Sample Adults who indicated “a lot” of difficulty or “cannot do at all” were classified as having “moderate to more severe” disability, and Sample Adults who indicated “none” to all six WG-SS questions were classified as having “no” disability. We report estimates for 1) no disability, 2) milder disability, 3) moderate to more severe disability, and 4) any disability (21). We chose to report “any” disability to identify those with any level of difficulty, whether it was milder, moderate, or more severe. In reporting “any disability” for any of the six functional domains, if a person experienced moderate to more severe disability for one domain but milder disability for another, they were assigned to moderate to more severe, so those who reported only “some” difficulty to any of the domains were assigned to milder disability.

We examined the national crude prevalence and age-standardized (to the 2000 U.S. standard population) prevalence of self-reported diagnosed diabetes (hereafter diabetes) by self-reported sociodemographic characteristics including age, sex, race and ethnicity, level of education, type of insurance, and poverty-to-income ratio (PIR), which is the ratio of the family’s income to the appropriate federal poverty threshold (FPL) (22). We also examined self-reported BMI, measured as weight in kilograms divided by the square of height in meters, by self-reported disabilities. We used BMI classifications of underweight (<18.5 kg/m²), normal weight (18.5–24.9 kg/m²), overweight (25.0–29.9 kg/m²), and obesity (≥30 kg/m²) for U.S. adults aged ≥18 years. We calculated the crude prevalence and age-standardized prevalence of diabetes among adults with each of the six functional disabilities. We also assessed the average duration of diabetes (time since self-reported date of diagnosis) by level of severity of any disability. For conservative comparisons, we compared CIs for overlap (23) to assess differences by sex, race and ethnicity, level of education, insurance type, and PIR, adjusting for age. Data were suppressed when appropriate using the National Center for Health Statistics data presentation standards for proportions (24).

During 2021–2022, the unadjusted prevalence of any disability among U.S. adults aged ≥18 years was 44.0% (95% CI 43.3–44.7); 34.9% (95% CI 34.3–35.6) reported milder disability, and 9.0% (95% CI 8.7–9.4) reported moderate to more severe disability. The most frequently reported age-standardized milder disability was cognition (17.3%, 95% CI 16.8–17.8), followed by vision (15.6%, 95% CI 15.1–16.0), hearing (12.4%, 95% CI 12.0–12.8), mobility (11.6%, 95% CI 11.3–12.0), communicating (4.4%, 95% CI 4.2–4.7), and self-care (2.9%, 95% CI 2.7–3.1) (Fig. 1A). The most frequently reported age-standardized moderate to more severe disability was mobility (4.5%, 95% CI 4.3–4.7), followed by cognition (2.6%, 95% CI 2.4–2.8), hearing (1.4%, 95% CI 1.3–1.5), vision (1.4%, 95% CI 1.3–1.5), self-care (1.0%, 95% CI 0.9–1.1), and communicating (0.8%, 95% CI 0.7–0.9).

Overall (all estimates presented are age standardized unless otherwise noted), 11.2% (95% CI 10.8–11.6) of U.S. adults (crude: 14.9%, 95% CI 14.4–15.4) with any functional disability had diabetes during 2021–2022 (Table 1). Diabetes prevalence increased by level of disability: from 9.5% (95% CI 9.1–9.9) among those with milder disability to 18.3% (95% CI 16.9–19.9) among those with moderate to more severe disability (Table 1 and Fig. 1B). Prevalence of diabetes increased linearly with the number of disability types (Fig. 2). Prevalence of diabetes among adults with no disabilities was 5.8% (95% CI 5.5–6.1), one disability 8.4% (7.9–8.9), two disabilities 11.8% (10.9–12.8), three disabilities 15.7% (14.2–17.3), and four or more disabilities 20.4% (18.1–23.0).

Among U.S. adults with milder disability, diabetes prevalence was highest among adults with self-care disability (20.0%, 95% CI 17.7–22.5), followed by mobility (14.7%, 95% CI 13.6–15.9), communicating (12.6%, 95% CI 11.2–14.1), cognition (12.2%, 95% CI 11.5–12.9), vision (11.8%, 95% CI 11.1–12.5), and hearing (9.9%, 95% CI 9.2–10.6) (Fig. 1B). Among adults with moderate to more severe disability, diabetes prevalence was highest among adults with mobility disability (25.0%, 95% CI 21.9–28.5), followed by self-care (21.8%, 95% CI 17.7–26.5), hearing (18.4%, 95% CI 14.5–23.0), vision (17.6%, 95% CI 14.5–21.2), cognition (16.2%, 95% CI 14.0–18.7), and communicating (14.5%, 95% CI 10.5–19.7). Prevalence of diabetes among adults with moderate to more severe disability was similar to the prevalence among those with milder disability for adults with communication and self-care disabilities. Prevalence of diabetes was higher among those with moderate to more severe disability compared with those with milder disability (Fig. 1B) for vision, hearing, mobility, and cognition.

Among adults with any disability, diabetes prevalence (age standardized) increased with age-group (from 18–44 to 65–74 years), was higher for men than women, was higher for NH Black and Hispanic adults than NH White adults, was higher among those with any public health insurance or Medicare compared with those with private insurance or without insurance, was inversely associated with level of education, and was inversely related to PIR for those <100% and at 100%–299% of FPL compared with those at 300%–499% and ≥500% of the FPL (Table 1). Within each level of disability, patterns of diabetes prevalence were similar within subgroups (characteristics). In comparisons with adults with any milder disability, diabetes prevalence was higher among those with any moderate to more severe disability. Diabetes prevalence was lower among those with no disability than among adults with any milder disability.

Among adults with a vision, hearing, mobility, or cognition disability, diabetes prevalence (age standardized) increased by age-group from 18–44 to 65–74 years, and among adults with self-care or communication disability, diabetes prevalence increased by age-group from 18–44 to 45–64 years (Table 2). Among adults with mobility disability, diabetes prevalence was higher for men than women. Among adults with a vision, hearing, mobility, or cognition disability, diabetes prevalence was higher for NH Black and Hispanic adults than for NH White adults; among adults with self-care disability, diabetes prevalence was higher for Hispanic adults than for NH White adults. Among adults with a vision, hearing, mobility, cognition, or communication disability, diabetes prevalence was higher for those publicly insured than for those privately insured. Among adults with a vision, hearing, or cognition disability, diabetes prevalence was inversely associated with level of education; for adults with mobility disability, diabetes prevalence was higher for those with a high school (HS) education and less than a HS education than for those with more than a HS education; and among adults with self-care or communication disabilities, diabetes prevalence was higher among those with less than a HS education than among those with more than a HS education. Diabetes prevalence was inversely related to PIR among adults with vision, cognitive, and self-care disability (for those <100% and at 100%–299% compared with those at 300%–499% and ≥500% of the FPL).

The mean duration of diabetes among those with no disability was 14.2 years (95% CI 13.1–15.2), with milder disability 17.1 years (95% CI 16.2–18.0), and with moderate to more severe disability 20.9 years (95% CI 19.6–22.2).

Overall, 44% (crude) of U.S. adults had milder, moderate, or more severe disability during 2021–2022, and among them, 14.9% reported that they had ever been diagnosed with diabetes. We observed a dose response of increasing diabetes prevalence with increasing number of disability types. Diabetes prevalence was higher among adults with moderate to more severe disability than among those with milder disability for the four most frequently reported disabilities (cognition, vision, hearing, and mobility). However, for the least frequently reported (≤5% of U.S. adults) disabilities (communication and self-care), there was no difference in prevalence of diabetes between those with milder and those with moderate to more severe disability. Consistent with the prevalence of diabetes in the general population (25), prevalence of diabetes among those with disabilities generally increased with age, was higher among men than women, was higher among NH Black and Hispanic than NH White adults, and was inversely associated with level of education and PIR.

In this study, 9.5% of adults with milder disability and 18.3% with moderate to more severe disability reported a diabetes diagnosis, which may be related to the progression to severe disability experienced with longer duration of diabetes (26). Although this study was cross-sectional, we observed that longer duration of diabetes was associated with increased severity of disability. Notably, there was no difference in diabetes prevalence between those with milder disability and those with moderate to more severe disability among adults with the two least frequently reported disabilities. This may be in part explained by the severity of the disability, even if it was reported to be “milder.” In other words, the least frequently reported disabilities may be “severe” or rare in this study because the survey only includes community-dwelling individuals (i.e., self-care [27] and communication [28] disabilities are more prevalent among residents of long-term care facilities). Another explanation is that those with more severe disability may be less likely to respond to the survey. In another national study of community-dwelling adults with examination of the association of diabetes and disability, investigators found that even after they controlled for A1C, diabetes duration, and cardiovascular disease, adults with diabetes had the highest odds of self-care disability (measured according to activities of daily living including dressing and bathing), compared with lower-extremity mobility, instrumental activities of daily living, general physical activities, and leisure and social activities (29). This strong association may indicate why the prevalence of diabetes was similar regardless of whether this disability was milder or moderate to more severe. Difficulty with dressing and bathing can result from decreased muscle strength, lower muscle quality, and accelerated loss of muscle mass, especially in the lower extremities, which have all been reported in adults with diabetes (30).

Also notably, the WG-SS is designed to identify milder to more severe disability, whereas the ACS disability measure has a binary response of “yes” or “no” to six similar questions about serious difficulty (31). The Behavioral Risk Factor Surveillance System (BRFSS) uses the ACS indicator and in 2022 reported that 28.7% of U.S. adults have challenges in hearing, vision, mobility, cognition, self-care, or independent living (32). The BRFSS reported a somewhat lower proportion of adults with disabilities compared with the 44% found in the NHIS. In studies where investigators compared the ACS with the WG-SS, they found that individuals with poor or fair health as well as low socioeconomic status, defined according to education, income, and employment, chose to respond “some difficulty” (milder disability) in the WG-SS but “no disability” in the ACS, as the ACS only asked about “serious difficulty” (31). Since those with milder disability had lower prevalence of diabetes, that could explain the findings from the ACS in 2022 that 17.0% of adults with disabilities had diabetes, compared with 11.2% in 2021–2022 as identified by any level of disability in the WG-SS.

Both diabetes (33) and self-care disability (defined according to three or more activities of daily living) (27), independently, have been found to be strong predictors of nursing home admission. With the shrinking number of U.S. nursing home beds, an ongoing problem for decades and recently accelerated, the prevalence of adults with diabetes and self-care disability may increase in the community-dwelling population. Monitoring prevalence of diabetes among those with self-care disability will be important for policy-makers, health care providers, and researchers.

We observed a dose response of increasing diabetes prevalence with increasing number of disability types. In another study among older adults, investigators found that higher number of chronic diseases, as well as diabetes alone, was associated with higher number of functional disabilities (34). That increased prevalence of diabetes is associated with higher number of disabilities is likely multifactorial. Longer duration of diabetes, especially if not managed well, could contribute to multiple disability types (29). Some physiological changes that occur with aging, such as loss of muscle mass and strength (35), may contribute, although many chronic diseases can contribute to the development of impairments independently. Additionally, factors associated with diabetes and disability may be more prevalent and result in multiple disabilities (34). Some examples include obesity, insulin resistance, diabetes-related complications (e.g., cardiovascular heart disease, lower-extremity arterial disease, stroke), and other comorbidities that may or may not be caused by diabetes (e.g., depression and arthritis) (36).

Diabetes prevalence was highest among adults with moderate to more severe mobility disability; conversely, in our previous study mobility was the most frequently reported disability among adults with diabetes (B.H.B et al., Centers for Disease Control and Prevention, unpublished data). This may in part be due to pathophysiological factors such as decreased muscle strength and impaired blood flow or to the increased risk of decline in lower-extremity physical functioning among adults with diabetes (37). Moreover, obesity is a well-known risk factor for both mobility disability and diabetes, confirmed by higher diabetes prevalence for those with overweight or obesity than for those with normal weight among adults with mobility disability. Populations with mobility impairment could be targeted for effective strategies for slowing progression, such as physical activity, which has been found to slow down the impairment process in populations at high risk (38). The American Diabetes Association strongly recommends physical activity and exercise for individuals with diabetes, and that it be tailored to meet the specific needs of each individual (39).

Maintaining the health and wellness of individuals with disabilities is an important public health concern, yet those with disabilities face health disparities and barriers to disease prevention programs and health promotion. Studies have shown that adults with disabilities receive significantly fewer preventive services and have poorer health status than individuals without disabilities (17). Yet, adults with disabilities, especially groups with high prevalence of diabetes (who have difficulty with activities of daily living [e.g., dressing or bathing] or communicating or with other moderate to more severe disabilities), may benefit from diabetes prevention programs. Moreover, adults with disabilities who are at risk of diabetes and could benefit from prevention programs may differ from other adults at risk of diabetes, as they have unique health needs, and may benefit from educational materials or health promotion programs universally designed to work for everyone, including people with disabilities. In addition, adults with diabetes and disability need diabetes self-management support and education to delay or prevent diabetes-related complications. One way to accomplish this would be to ensure that the existing, evidence-based diabetes programs (and diabetes self-management support and education programs) meet the needs of individuals with various types of disabilities, through evaluation and adaptation.

There were some notable limitations in our study. First, the study design was cross-sectional, so we could not determine the directionality between disability and diabetes. Second, due to the small sample of adults with type 1 diabetes, we could not report differences in disability prevalence by diabetes type (1 and 2) separately (24). Third, in using the moderate to more severe definition of disability, some subgroup estimates had to be suppressed due to small cell size for four of the six individual measures of disability. However, pooling data from 2 years improved precision and the level of suppression within subgroup estimates. Fourth, NHIS only includes noninstitutionalized individuals, which could lead to underestimation of disability prevalence. In addition, as adults with disabilities may have been less likely to respond, incidence of disability may have been underestimated overall. Also, the survey may be subject to recall bias; however, self-report of diabetes diagnosis in surveys has been considered valid and reliable, with substantial agreement (κ 0.71–0.80) for diabetes (40). Finally, we used the method of examining overlapping CIs to assess comparisons, which is not ideal for formal significance testing but provides conservative comparisons (23) that are reasonable given the number of disabilities, levels of disabilities, and subgroups assessed.

Notwithstanding these limitations, our study suggests that adults with multiple disability types, self-care and communication disabilities, or moderate to more severe disabilities may benefit from diabetes prevention programs because they are at increased risk of diabetes (3) and experience inequities in access to and receipt of health care (15,17).

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. B.H.B. was primarily responsible for writing the manuscript and for the final content of the manuscript. B.H.B., J.D.O., J.B.S., I.H., and K.M.B. contributed to the design of the study. All authors read, reviewed, and approved the final manuscript. B.H.B. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Handling Editors. The journal editors responsible for overseeing the review of the manuscript were Steven E. Kahn and Alka M. Kanaya.