Noninsulin-dependent diabetes mellitus, only six decades ago a rare phenomenon among the Navajo Indians, has become increasingly common in this population. In one of the earliest reports, Salsbury (1937) found only one case of diabetes mellitus (DM)⁴ in more than 6000 Navajo hospitalizations at Sage Memorial Hospital in Ganado, Arizona. Three years later, Joslin (1940), in a statewide survey of physicians, reported only four cases of Navajo with symptoms of DM. In the early 1960s, however, investigators began noting that the number of patients with DM admitted to clinics and hospitals on the Navajo reservation was increasing (Saiki and Rimoin 1968). A decade later, 11% of Navajo in the Fort Defiance area aged 35 y and older had DM (Bennett et al. 1972); in the early 1980s, 12% of adult Navajo in the Many Farms-Rough Rock area of the reservation had the disorder (Hall et al. 1992). These rates are about 10 times higher than the prevalence in the Many Farms-Rough Rock area in 1955-61 (Hall et al. 1992). Most recently, Sugarman et al. (1992) reported that 17% of persons aged 20-74 y living in the Teec Nos Pos chapter of the Shiprock Service Unit in northeastern Arizona had DM.

As in other populations (Donahue and Orchard 1992), Navajo with DM are at high risk of atherosclerotic macrovascular complications. In a 1990 study, about one fourth of Navajo aged 20 or over with DM who attended an IHS clinic had cardiac disease (Hoy et al. 1995). This was more than five times the prevalence of heart disease in their age-matched counterparts without DM. Furthermore, in that study almost 60% of persons with DM had cholesterol levels above 200 mg/dL (1 mg/dL = 0.05551 mmol/L). In a similar study conducted by the same investigators (Hoy et al. 1994), 53.8% of Navajo women with DM and 64.6% of Navajo men with DM had hypertension. This was more than six times the prevalence of hypertension in their age-matched counterparts without DM.

Navajo with DM are also at high risk of microvascular complications. In 1979-1980, 17 of 59 (29%) adult Navajo with DM were found by dilated ophthalmoscopy to have diabetic retinopathy (Rate et al. 1983). Duration of DM was a strong predictor because retinopathy was present in about half of those with DM for 10 y (Rate et al. 1983, Sugarman 1990). In 1990, the risk of microalbuminuria was found to be four times greater among Navajo subjects with DM than in their age- and blood pressure-matched counterparts without DM, and more than half of Navajo with DM had microalbuminuria (Hoy et al. 1994).

In this report, we present the following new information: DM estimates that are both population-based and representative of all Navajo living on or near the Navajo reservation, the rate of undiagnosed DM by sex and age, blood glucose control among a population-based sample of persons with DM, medical conditions associated with DM and diabetes care as reported by Navajo with diagnosed DM.

MATERIALS AND METHODS

IHS field staff initially approached selected households to explain the study and to invite household members to be interviewed, provide fasting blood samples, participate in OGTT and provide anthropometric measurements. If household members agreed, morning sessions to be held at the subjects' homes were scheduled, usually within 1 wk of the initial visit. Subjects were contacted by field staff the evening before the study session to remind them to fast overnight. Sessions began with determining eligibility for OGTT. Pregnant women, persons who reported a medical history of DM, and those who were acutely ill during the session were not eligible for OGTT. For eligible respondents, field staff measured glucose tolerance, plasma lipid levels, blood pressure, height and body weight, and administered a standardized study questionnaire to assess diet, physical activity, tobacco use, family history of various diseases and medical history.

Venous plasma samples were collected in grey-top vacutainer tubes containing sodium fluoride, centrifuged to separate plasma within 45 min of collection, transferred to red-top tubes, chilled in ice and delivered daily to the reference laboratory (Corning Clinical Laboratories, El Paso, TX). The plasma was analyzed for glucose using the Technicon RAXT (Technicon Instrument Corporation, Tarrytown, NY).

The WHO criteria require only fasting and 2-h values to determine diabetes status. The NDDG estimates require fasting and 1- and 2-h values. Classification criteria for the two methods are presented in Table 1.

Table 1. Unweighted number of persons aged 20 y and older detected with diabetes and impaired glucose tolerance using two classification methods, Navajo Health and Nutrition Survey, 1991-19921 [View Table]

After persons were seated for at least 10 min, three blood pressure measurements were taken with the use of a random-digit sphygmomanometer, and the mean of the last two values was used in the analyses. Hypertension was defined as a systolic blood pressure 140 mm Hg or a diastolic pressure 90 mm Hg or use of hypertension medication (National Heart, Lung, and Blood Institute 1993).

Height and weight were measured with respondents wearing light clothing without shoes. Overweight was defined using the sex-specific 85th-percentile values of body mass index (BMI) (kg/m²) from the second National Health and Nutrition Examination Survey (NHANES II) for 20- to 29-y-olds (Najjar and Rowland 1987). A BMI 27.8 for men and 27.3 for women indicated overweight.

All plasma lipid measurements were performed in a single laboratory (Corning Clinical Laboratories) with the fasting specimen. Specimens were transported to the laboratory on a daily basis, and levels of total cholesterol and triglycerides were measured with a Technicon SMAC-III; a Technicon RAXT (Technicon Instrument Corporation) was used for determinations of HDL cholesterol. We used ADA clinical practice recommendations (1996) to define high concentrations of triglycerides (200 mg/dL) (1 mg/dL = 0.01129 mmol/L) and LDL cholesterol (130 mg/dL), and low concentrations of HDL cholesterol (<35 mg/dL) (1 mg/dL = 0.02586 mmol/L).

Subjects were considered to have engaged in physical activity if they responded positively to the following question: During the past month, did you participate in any physical activities or exercises such as running, basketball, softball, aerobics, walking, hunting, getting wood, swimming or other activity for exercise? We assessed dietary fat intake by analyzing reports of all food consumed during the previous day (Ballew et al. 1997).

RESULTS

Table 2. Prevalence of diabetes mellitus (DM) and impaired glucose tolerance (IGT) in persons aged 20 y and older using WHO and National Diabetes Data Group (NDDG) criteria, by sex and age, Navajo Health and Nutrition Survey, 1991-19921 [View Table]

Table 3. Potential risk factors for diabetes mellitus (DM) among persons aged 20 y and older, by diabetes status and sex, Navajo Health and Nutrition Survey, 1991-1992 [View Table]

Other factors related to DM complications. Medians and age-standardized geometric means are presented for blood pressure levels and lipid concentrations by diabetes status (Table 4). Most notable, more than half of all persons with DM or IGT had fasting triglyceride values 160 mg/dL. The median triglyceride value for persons without DM was only 124 mg/dL, and persons with DM were three times more likely than persons without DM to have triglyceride values > 200 mg/dL. Persons with DM were 2.6 times more likely to have hypertension than were persons without DM. Furthermore, they were three times more likely to have low HDL values than were their counterparts without DM. Persons with DM were as likely to smoke cigarettes or chew tobacco as were persons without DM.

Table 4. Factors associated with the development of heart disease and other complications of diabetes mellitus (DM), by diabetes status and sex, Navajo Health and Nutrition Survey, 1991-1992 [View Table]

Table 5. Type of therapy and diabetes care among persons aged 20 years and older with a medical history of diabetes, by sex, Navajo Health and Nutrition Survey, 1991-1992 [View Table]

DISCUSSION

Demographic trends, better detection and nutritional changes may all have contributed to the increasing prevalence of DM among the Navajo. Because about two thirds of newly diagnosed cases of DM occur after age 45 (Kenny et al. 1995), the aging of the Navajo people is likely to have been an important factor (life expectancy at birth rose from 71.1 y in 1980 to 73.7 y in 1987) (Office of Program Planning, IHS 1993, Sugarman et al. 1990). In 1979, the IHS established a diabetes program charged with the detection and control of DM among the Navajo and other American Indians (Mayfield et al. 1994). This program may have had a measurable effect on reported DM prevalence. For example, in the Many Farms-Rough Rock area of the Navajo Nation in 1988, only about 25% of persons with DM determined by oral glucose tolerance testing (OGTT) were found to be previously undiagnosed (Hall et al. 1991), a proportion substantially below the rate of about 50% in the U.S. general population (Harris et al. 1987). Some of the increase in DM among the Navajo may be due to the same nutritional factors (increased caloric and dietary fat consumption, combined with a sedentary lifestyle) that have affected the U.S. general population. Nationally, the prevalence of DM has risen substantially  it more than doubled between 1935 and 1960 and tripled from 1960 to 1992 (Kenny et al. 1995). Still, some researchers have suggested that these lifestyle changes cannot fully account for the Navajo's higher DM rates and have postulated an underlying genetic mechanism (Hall et al. 1992). However, Navajo-specific genes linked to DM have yet to be identified, although a gene linked to DM has been isolated in a Southwestern Mexican-American population (Hanis et al. 1996).

Gender discrepancies exist among the Navajo in the prevalence of diabetes and the awareness of the condition. Women were 27% more likely to have DM and 18% more likely to have IGT than were men. About one third of all Navajo with diabetes in the study were not aware of having this disorder at the time of the survey. This proportion is similar to that reported previously in the Many Farms-Rough Rock area of the Navajo reservation in 1988 (Hall et al. 1991). Men were much more likely to be unaware of their diabetes than were women (51 vs. 23%).

Little has been published regarding blood glucose control among a population-based sample of Navajo with DM (O'Connor et al. 1990). We found that blood glucose control (using fasting plasma glucose which is not as informative as glycosylated hemoglobin) was poor among almost 40% of the persons with a known history of DM. Although tight glycemic control (155 mg/dL) has been shown to delay the onset of retinal, renal and neurologic abnormalities in persons with insulin-dependent diabetes mellitus (Diabetes Control and Complications Trial Research Group 1993), many diabetes experts (American Diabetes Association 1996) believe that it may also delay these complications in persons with noninsulin-dependent diabetes. Insulin was rarely used by either men or women in our study, and almost a third of women used no medication to control their diabetes. Furthermore, <10% of Navajo had ever monitored their own blood glucose.

Other than good glycemic control (Diabetes Control and Complications Trial Research Group 1993), components of good diabetes care endorsed by the ADA (1996) include regular exercise, consumption of <10% of energy from saturated fat, weight loss or maintenance, achievement of optimal serum lipid and blood pressure levels, foot inspections to find ulcers or sores, blood pressure checks at every doctor's visit, smoking cessation and annual dilated ophthalmoscopic eye examinations. In our study, persons with DM were approximately three times more likely to be overweight than persons without DM. Women with a medical history of DM were especially likely to be overweight, and men with DM were likely to distribute body fat in the truncal region. Three fourths of all persons with DM consumed >10% of their energy from saturated fat and ~40% were sedentary. Approximately 20% of Navajo with a history of DM had never been advised to exercise. According to self-reports, >10% of persons with a medical history of DM had not received an eye examination of any kind or a blood pressure check during the past year.

The Navajo have already taken some important measures in addressing DM and its complications. As early as the late 1970s, programs were developed to target diabetes among the Navajo and other American Indians (Mayfield et al. 1994). In part, screening and awareness campaigns may explain the greater likelihood that a Navajo person with diabetes has been diagnosed than has someone with diabetes in the general U.S. population (Harris et al. 1987). Navajo with a history of DM seem to have responded to educational messages about the importance of tobacco cessation and checking one's feet for sores and ulcerations.

Unlike previous studies of diabetes in the Navajo, this study used oral glucose tolerance testing and reservation-wide population-based sampling to present more comprehensive information on this subject. Yet, the study has some limitations. General sampling and design flaws are discussed in White et al. (1997). Because the Navajo tend to live in extended families, samples such as ours that use the family as the unit of analysis may include many cases of diabetes in one sampling unit, because diabetes clusters in families; this may result in large standard errors when the cluster sampling is accounted for (such as those reported in this study). Furthermore, the sample of men with diabetes may have been biased, because it is possible that many men were away from the home when the family unit was approached, leaving an unrepresentative group for the study. The 1990 census of the population of the Navajo Nation indicates that the male-to-female ratio among persons aged 21-64 y is 0.90 (personal communication with the Navajo Nation), but in NHANS, the male-to-female ratio is 0.56 for the same age group. If the men who were missed were more likely to be workers, it is quite possible that they were healthier and less likely to have diabetes. This would contribute to inflated estimates of diabetes among men. Much of the information on diabetes care and complications relied on self-reports. Furthermore, only one question was used to assess physical activity, and the duration and intensity of activity could not be determined. Finally, having small samples of persons with diabetes sometimes resulted in imprecise statistical estimates; clearly, larger samples of persons with diabetes should be included in future studies.

In conclusion, diabetes has become an important public health problem for the Navajo. The results of this study provide important baseline data for monitoring the future trend of diabetes and for evaluating the effectiveness of interventions to prevent and control diabetes in this population. To reduce substantially the future burden of diabetes, it will be important to continue current efforts such as diabetes screening, monitoring of diabetes care through medical audits, and patient education. New initiatives may also be required, such as aggressive campaigns to improve blood glucose control that include dietary management, assessment of the feasibility of treatment with insulin or hypoglycemic oral agents and the promotion of regular monitoring of blood glucose by the patient either at home or in the clinic. Development of structured programs to facilitate reducing overall calories and dietary fat (e.g., medical nutrition therapy) (Franz et al. 1995) and increasing physical activity may be important for both clinical control and prevention of DM among the Navajo. Population-wide strategies aimed at weight maintenance and control should also be implemented and evaluated to determine their effectiveness in preventing or delaying glucose intolerance. Such efforts are already beginning (personal communication, Diabetes Primary Prevention Trial study committee) in the Navajo Nation and other selected sites throughout the U.S.

ACKNOWLEDGMENTS