Original Research

Hypothyroidism and Diabetes Mellitus in an American Indian Population

Author and Disclosure Information



BACKGROUND: The clinical relationship between diabetes and hypothyroidism is becoming more widely recognized. Although American Indians are disproportionately afflicted with diabetes, the occurrence of hypothyroidism within this group has not been previously reported.

METHODS: We present data from a retrospective chart review of health clinic data from a rural isolated northeastern tribe. A total of 156 cases of diabetes and 25 cases of hypothyroidism were identified among 892 eligible individuals living in the service area.

RESULTS: Both conditions exhibited strong sex differences. The prevalences of diabetes (21%) and hypothyroidism (5%) among women were higher than those observed among men (13% and 0.2%, respectively). The overall prevalence of hypothyroidism among women with diabetes (8.8%) varied by age ranging from 5% among women younger than 60 years to 21% among women aged 60 years and older.

CONCLUSIONS: Our findings support the need for further investigation of the association between diabetes and hypothyroidism in American Indian populations with high prevalence rates of diabetes. This association may be of particular interest to family physicians and other clinicians caring for American Indian populations.

Thyroid disease and diabetes mellitus account for considerable morbidity in the United States. The annual incidence of hypothyroidism in adults is 0.08% to 0.2%. Hypothyroidism is more common among women than men, and incidence increases with age. An estimated 6% of the US population has diabetes, and more than 600,000 new cases are diagnosed each year. Both diseases are more prevalent in women. Thyroid disease has been reported to occur with such frequency within populations with diabetes that some authorities recommend routine screening for those patients.1-3 Hypothyroidism has been observed to be quite prevalent in patients with diabetes.2-6 Although American Indians are known to have high prevalence rates of diabetes, no published reports have examined the occurrence of hypothyroidism among American Indians with diabetes.


A medical record review was undertaken for all patients attending the health clinic of a northeastern American Indian tribe. This tribal community is located in a rural area that is approximately 60 miles from major medical services. The Tribal Health Center is operated as a local comprehensive health care delivery system under contract with the Indian Health Service (IHS). The IHS, an agency of the Department of Health and Human Services, provides funding and oversight to federally recognized tribes for a broad-spectrum program of preventive, curative, rehabilitative, and environmental services.7 This system integrates health services provided directly through IHS facilities, purchased by IHS through contractual arrangements with providers in the private sector, and delivered through programs operated by tribes. The main objective of the health center is to provide the highest level of health care possible for eligible recipients in the service delivery area. The center is staffed by full-time physicians, physician assistants, nurse practitioners, pharmacists, counselors, aides, and clerical and administrative personnel.

Although patient medical records are maintained in paper form, key elements of the patient visit, including diagnosis, are entered into a central IHS patient care database. We used this clinic database to identify diagnostic groups for our study. A medical record review was undertaken for the 892 individuals (415 men, 477 women) eligible for clinic services during calendar year 1998. These individuals represent the denominator and are the number of eligible American Indians living within the geographic service area of the clinic. Of the 477 women, 431 were aged younger than 60 years. We directed our attention to patient records listed with a diagnosis of either diabetes mellitus or hypothyroidism. Odds ratios (ORs) were computed, and 95% confidence intervals (CIs) were calculated as described by Rothman.8


An electronic search of clinic records for 1998 identified a total of 156 cases of diabetes (17.5% prevalence) and 25 cases of clinical hypothyroidism (2.8%) occurring among the 892 eligible patients ([Table 1]). Among men (n=415) there were 54 cases of diabetes (13%) and 1 case of hypothyroidism (0.2%). The 1 case of hypothyroidism occurred in a man with diabetes. For women (n=477) prevalence rates of diabetes and hypothyroidism were higher, as expected, with a total of 102 cases of diabetes (21.4%) and 24 cases of hypothyroidism (5%).

Prevalence rates for hypothyroidism differed markedly by age and diabetes status. The prevalence of hypothyroidism among women with diabetes was 8.8% overall, ranging from 5% among women younger than 60 years to 21% among women 60 years and older ([Table 2]). The OR for hypothyroidism occurring among women with diabetes relative to those who did not have diabetes was 2.32 (95% CI, 0.9-5.8). Also, the ORs of these 2 conditions among women appeared to be modified by age, with the OR being 1.4 (95% CI, 0.4-4.8) among women younger than 60 years and 2.63 (95% CI, 0.4-22.6) among those aged 60 years and older. However, these data should be viewed cautiously, because the CIs include 1.0, and because of the small number of women aged 60 years or older.


Recommended Reading