Diet has been implicated in the etiology of diabetes, cardiovascular disease and cancer in many populations including the Navajo and other American Indian tribes (National Research Council 1989a, Rhoades et al. 1987). Before European contact, the Navajo cultivated corn and other crops but also ate a wide variety of wild animal and plant foods (Kluckhohn and Leighton 1974). After the Spanish introduced sheep and goats in the 17th century, the Navajo adopted mutton as a dietary staple but continued to grow and gather many plant foods. By the 1930s, a trend toward increased reliance on mutton and purchased staples (sugar, flour, coffee, bacon) was observed, but the Navajo's nutritional status was reported to be adequate and the absence of obesity was noted (Carpenter and Steggerda 1939, Steggerda and Eckhardt 1941).

In 1955, a survey of 1246 individuals in two regions of the Reservation found that the Navajo diet was good and that there appeared to be no nutritional problems, with the possible exception of inadequate vitamin C intake (Darby et al. 1956, Sandstead et al. 1956, Stockell et al. 1956). In 1968, a large survey of a single community on the Navajo Reservation found a high risk of inadequacy of iron, protein and vitamin C intakes (University of Pittsburgh 1969). Three small surveys between 1969 and 1981 found evidence of widespread nutritional inadequacy among both children and pregnant or lactating women, including low intakes of vitamin A, folate, calcium and iron. These studies led to the initiation of several nutritional support programs on the Navajo Reservation (Butte et al. 1981, Van Duzen et al. 1969 and 1976).

There has never been a comprehensive survey of the nutritional status of a representative sample of the Navajo people living on or near the Reservation. This report presents the dietary findings of the 1991-1992 Navajo Health and Nutrition Survey, including nutrient intakes and the primary food sources of nutrients.

MATERIALS AND METHODS

Interviewers administered a single 24-h diet recall to all participants. Thirty-five pregnant women and three individuals who were unable to respond to the dietary interview were excluded from this analysis. Interviews were conducted Monday through Friday and were distributed approximately evenly during the months from October 1991 through December 1992.

Interviewers were members of the Navajo community who were fluent in both Navajo and English. Staff of the National Center for Health Statistics trained both the interviewers and the Indian Health Service (IHS)⁴ public health nutritionists who supervised them in the interview techniques used in the Third National Health and Nutrition Examination Survey (NHANES III, National Center for Health Statistics 1994). Training for the dietary recall included instruction in standardized probes for completeness, portion size estimation, preparation techniques, brand names, recipes and the use of nutritional supplements. The interviewers assisted participants in estimating portion sizes by referring to two- and three-dimensional models and standard measuring and serving utensils. They also examined the household's measuring and serving utensils and food packages.

Twenty-four hour recall data were recorded on interview forms for later entry into the Food Intake Analysis System (FIAS version 2.3) software (University of Texas Health Science Center at Houston, Houston, TX). FIAS uses a subset of the USDA Nutrient Data Base for Food Consumption Surveys, release 6 (U.S. Department of Agriculture, Human Nutrition Information Service 1991). To code all foods reported in the 24-h recalls, the data base was supplemented with six Navajo food ingredients: blue corn meal, juniper ash water, Mormon tea, two varieties of Navajo tea [Ephedra sp. and Thelasperma spp. (Wolfe et al. 1985)], and North American pine nuts (ESHA II nutrient database, 1987, ESHA Research, Salem, OR). Navajo recipes were incorporated into the database by modifying existing FIAS recipes. Representative recipes for fry bread, Navajo tortillas, blue corn mush with and without ash, kneeldown bread, Navajo cake and sumac pudding were obtained from informants who prepared them frequently; these recipes were used for all individuals who reported those foods. Other added recipes included a variety of soups, stews and mixed dishes that were coded separately for each household that reported them.

The 24-h diet recalls were coded by two of the interviewers who were trained by one of the supervising nutritionists. All computerized diet records were then reviewed against the original data recording forms by the supervising nutritionist, a Navajo coder, and a coding and software consultant. Any data entry errors that could not be unambiguously corrected from the original record were coded according to default definitions (e.g., the diameter of a fry bread, the composition of a Navajo taco) that were based on the expertise of the supervising nutritionist and the Navajo coder. Error and outlier reports from the FIAS software were reconciled with the original record.

The FIAS software nutrient data output was converted to a SAS data set (SAS Institute, Cary, NC) for analyses. We present weighted means, standard errors and 15th, 50th and 85th percentiles for energy and 22 nutrients. Sample weights were assigned according to the sampling design described by White et al. (1997), and weighted analyses were performed with SUDAAN software (Shah 1991). Because of the highly skewed distributions of many nutrients, sex and age-group comparisions were based on nonparametric Wilcoxon statistical tests (Siegal 1956).

We compared the median intake of nutrients to sex- and age-specific recommended dietary allowances (RDA, National Research Council 1989b). Because our age strata differed slightly from those of the RDA, we used the RDA for 15- to 18-y-olds for participants age 12-19 in our study, the RDA for 25- to 50-y-olds for participants age 20-59, and the RDA for 51 y for participants age 60 and older. To the extent that the RDA for some nutrients differ by age group, application of these RDA values to evaluate intake of our participants <15 y and between 50 and 59 y may result in some misclassification. We assessed this by comparing the median value for the affected age strata against the lower RDA value as well.

The software we used did not permit us to disaggregate mixed dishes into components; thus some foods could not be counted separately in food group analysis. Mixed dishes coded as recipes, such as soups, stews or meatloaf, were classified according to the constituent food contributing the largest proportion of energy and of most other nutrients. For example, mutton stew as prepared by most Navajos derives most of its energy and nutrients from mutton; thus it was classified with meats. Some other foods are coded as single items in the data base. For example, the calcium in cheese on fast-food cheeseburgers was included in estimates of calcium intake, but because a fast-food cheeseburger is coded as a single item in the software, it was attributed to the cheeseburger rather than to the dairy product category in the food group analysis. Pizza is also coded as a single item in the data base. Similarly, small amounts of fruits and vegetables used in mixed dishes were included in the calculation of nutrient intakes but were not tabulated separately as fruits and vegetables in the food group analysis. However, the mixed dishes represent categories of food or food groups as defined and consumed by the Navajo.

The only exceptions to this were the Navajo taco and the mutton sandwich. A Navajo taco is a piece of fry bread covered with chili beans, usually prepared with ground beef, then topped with shredded lettuce, grated cheese, salsa, tomatoes, green chili and sometimes chopped onions. A mutton sandwich is a Navajo tortilla wrapped around grilled mutton, sometimes served with slices of hot pepper and condiments. Because the composition of Navajo tacos and mutton sandwiches may vary, the interviewers attempted to elicit detailed descriptions of each one reported; these were coded as individual ingredients. Default codes for Navajo tacos and mutton sandwiches, based on those served in restaurants or by market vendors on the Reservation, were used if necessary.

The software permitted us to count the number of times participants reported eating specific foods, but we were not able to express these counts in terms of the standard servings of the Food Guide Pyramid (U.S. Department of Agriculture 1995). Each food reported therefore represents a serving as consumed by the participant, but this may be different than a standard serving.

To describe the food sources of nutrients, we present the percentage contribution of each food group to the total consumption of macronutrients and to the total consumption of five micronutrients that have consistently been found to be low in Navajo diets in the past, i.e., vitamin A, vitamin C, folate, calcium and iron.

RESULTS

Intakes of energy, carbohydrate and saturated fat varied significantly by age group among both men and women (Table 1). Among men, intakes of protein, vitamin E, vitamin C, calcium and phosphorous varied significantly with age group. Among women, intakes of total fat, vitamin A, -carotene, vitamin C, folate, calcium and phosphorous decreased significantly with increasing age group (Table 2). Participants age 60 and older consumed diets that were denser in protein, total fat, fiber, niacin, thiamin, iron and magnesium per 1000 kcal than younger participants (P < 0.01 for all comparisons, data not shown), but consumed diets less dense in carbohydrate, vitamin C and calcium (P < 0.001 for all comparisons). There were no statistically significant differences in nutrient intake by sex within age groups, except among participants age 40-59. Men age 40-59 consumed more energy, carbohydrate and protein than women (Table 1) and more phosphorous and zinc (Table 2) (P < 0.01).

Table 1. Intake of energy and macronutrients by sex and age group among participants in the Navajo Health and Nutrition Survey, 1991-1992 [View Table]

Table 2. Intake of micronutrients by sex and age group among participants in the Navajo Health and Nutrition Survey, 1991-1992 [View Table]

Among both men and women, the median intakes of vitamin A, vitamin E, vitamin B-6, folate, calcium and magnesium were below the RDA in all age groups (Table 2). The median vitamin C intake was below the RDA for men and women age 20 and older. Median thiamin intake was below the RDA for men age 60 and older and women under age 60. Median riboflavin intake was below the RDA for men under age 20 and men age 60 and older, and for women over age 20. Median iron intake was below the RDA for all women under age 60.

Intakes relative to the RDA for five nutrients consistently found to be low among the Navajo in the past (vitamin A, vitamin C, folate, calcium and iron) are shown in Figure 1. Median intakes of vitamin A among men and among women >20 y were particularly low relative to the RDA, as were intakes of calcium for both sexes and all age groups. The median intake of vitamin C decreased significantly with age among both men and women. Median iron intake was below the RDA for women <60 y.

Table 3. Food sources of macronutrients among participants in the Navajo Health and Nutrition Survey, 1991-1992 [View Table]

Table 4. Food sources of selected micronutrients among participants in the Navajo Health and Nutrition Survey, 1991-1992 [View Table]

Meats, excluding mutton, and grain products, excluding Navajo grains, desserts and snacks, each contributed 15% to the energy intake (Table 3). Meats, excluding mutton, also contributed 29% of protein, 20% of total fat, 20% of saturated fat and 21% of iron intake (Tables 3 and 4). Grains, excluding Navajo grains, desserts and snacks, contributed 15% of carbohydrate, 11% of vitamin A, 19% of folate, 15% of calcium and 23% of iron (Tables 3 and 4). Restaurant and fast-food meals accounted for only 3% of the foods mentioned by participants, but contributed 8% of the energy, 11% of the total fat, 13% of the saturated fat and 11% of the calcium (Tables 3 and 4). Foods of low nutritional value such as soft drinks, desserts and snacks, sugars and candy, and fats and oils counted separately (e.g., spread on bread, dressing on salad) together contributed 15% of energy intake and 11% of total fat intake. Because the category of soft drinks included fortified fruit-flavored beverages, they contributed 19% of the vitamin C and 9% of the calcium consumed.

Fruits and vegetables were each consumed less than once per person per day. Although fruits and vegetables contributed 25% of the fiber in the diets, 34% of the vitamin A, 57% of the vitamin C and 31% of the folate (Tables 3 and 4), their infrequent consumption resulted in low overall intakes of these nutrients (Table 2). Dairy products, rich in calcium, were consumed less than once per person per day. Only 19% of the dairy prodcuts mentioned were identified as lower fat items (skim, 1% or 2% milk, ice milk). Dairy products contributed 26% of the calcium in the diets (Table 4), but infrequent consumption led to low overall intake of calcium. Although juniper or other plant ash used in the preparation of blue corn dishes is a traditional source of calcium in the Navajo diet, it contributed only 1% of the calcium reported on a population basis. For 10 individuals who consumed blue corn dishes with ash, however, ash was a major source of calcium (average 475 mg per person, range 266-1127 mg).

DISCUSSION

Another limitation is that we were not able to separate the components of some mixed dishes and prepared foods to assess their individual contributions to the diet. The nutrient contents of component foods of a recipe were captured in the nutrient intake analysis, but could not be separated in the analysis of food sources of nutrients. This applied chiefly to dairy products and fruits and vegetables that were components of mixed dishes. However, the misclassifications that resulted are likely to be of small magnitude because this happened relatively infrequently, most often with restaurant and fast-food entrees that are coded as a unit rather than as component foods and that accounted for only 3% of the items coded.

Because interviews were conducted primarily on Monday through Friday, very few diet recalls were available for Fridays and Saturdays. Dietary intakes may vary substantially on weekdays compared with weekends in many segments of the U.S. population (Gibson 1990). We found no differences between Sundays and the other days of the week among our participants, but cannot assess possible differences on Fridays or Saturdays. However, many Navajo, particularly those in rural areas, make their living as farmers, herders or artisans and are less likely to be affected by weekday/weekend distinctions in lifestyle than are some other groups. Future dietary surveys among the Navajo should collect data on weekend days as well as weekdays to explore this possible source of dietary variability.

The potential for inaccurate estimates of median intakes of nutrients below the RDA occurred among participants age 12-14 and among those age 51-59 for whom the RDAs are slightly lower than the reference values applied to our age strata. However, if the lower RDA values had been applied to these age strata, median intakes would still be below the RDA, with the exception of niacin among men age 11-14, folate among men and women age 11-14 and iron among women age 40-59.

Our survey indicates that the current diet of the Navajo exceeded the recommended fat content of 30% of energy and saturated fat content of 10% of energy, even with the likely underestimate of saturated fat (National Research Council 1989a). Although median cholesterol intakes were consistently less than the recommended 300 mg/d, cholesterol intakes may have been underestimated in this analysis, as was saturated fat.

Some segments of the Navajo population continue to be at risk of inadequate intakes of several key nutrients, including those that have been found to be inadequate among the Navajo in the recent past, namely, vitamin A, vitamin C, folate, calcium and iron (Butte et al. 1981, Sandstead et al. 1956, Stockell et al. 1956, University of Pittsburgh 1969). Although intakes of vitamin C among teenagers and of iron among women of reproductive age appear to be higher than they once were (Butte et al. 1981, University of Pittsburgh 1969), intake of vitamin C is still inadequate among most Navajo and intake of iron is inadequate among women under age 60. Participants age 60 and older in our sample had diets particularly low in vitamin A, vitamin C, folate and calcium. The elderly may eat different foods than younger adults because they consume less total energy, but their diets are denser in protein, fat, iron and some other nutrients, although less dense in vitamin C and calcium. This is consistent with the observations of IHS dietitians that many elderly Navajo report diets of limited variety.

There are many ways to enhance the intakes of key nutrients, but dietary interventions must be feasible and consistent with existing food preferences. Three major limitations affect dietary choice among the Navajo: cost, availability and shelf life. Most Navajo families have limited cash resources and purchase food on the Reservation where selection is limited (Ford and Harris 1988). Many families shop infrequently and do not have refrigeration; thus perishable items must be consumed quickly. Because fortified ready-to-eat cereals supply many nutrients, require no refrigeration, and have a long shelf-life, they could be a valuable addition to Navajo diets. Cereals were reported very infrequently in this survey, perhaps because they are relatively expensive. Fruits and vegetables are also dense sources of many micronutrients but they were consumed, on the average, less than twice a day. Although some kinds of fresh fruits and vegetables are available and affordable on the Reservation, many fresh produce items may be unavailable, very expensive or too perishable for families who shop infrequently and lack refrigerators. Canned fruits and vegetables are available and adaptable to Reservation shopping and storage patterns.

Dairy products are not widely consumed among the Navajo, in part because lactose intolerance is common and in part because dairy products are perishable. The survey participants reported using canned and powdered milk more frequently than fresh milk. A few participants reported preparing Navajo tortillas and fry bread with nonfat dry milk, thus increasing the calcium content from 40 to 144 mg per tortilla, but even this use of milk may be problematic for those who are lactose intolerant. Traditional blue corn meal foods with ash provided substantial amounts of calcium for the few participants who ate them. In spite of the low reported calcium intakes of the participants, IHS diagnoses of hip fracture and other sequelae of osteoporosis are rare (Indian Health Service unpublished data for Navajo Area, 1992-1994).

The median folate intake among Navajo women was consistently below the current RDA, and the median folate intake of Navajo women under age 40 was less than half of the most recent recommendation by the U.S. Public Health Service (1992) that women of reproductive age who are capable of becoming pregnant consume 400 µg folic acid/d to reduce the risk of neural tube birth defects. The low folate intake among Navajo women of reproductive age is of special concern because of the Navajo's high average fertility (Navajo Nation 1993) and the relatively high prevalence of neural tube defects among Native American populations (Wilcox and Marks 1994). Folate intake could be increased by greater consumption of fortified ready-to-eat cereals, oranges and orange juice. Canned and boxed juices with extended shelf-life could be adaptable to the shopping and consumption patterns of the Reservation. These beverages contain more nutrients than soft drinks, which contribute calories but few nutrients to the diet.

Until about the middle of the twentieth century, the Navajo lived an active lifestyle and ate a wide variety of gathered and cultivated foods in addition to the staple mutton. An increasingly sedentary lifestyle and erosion of traditional dietary patterns, with increasing dependence on relatively few refined, processed and nutrient-poor foods may be major factors in the increases in obesity and chronic diseases documented among the Navajo in the past half century. Diet is an important behavioral risk factor for all chronic diseases and is susceptible to intervention within the limits of the resources available to a population.

ACKNOWLEDGMENTS

Data on the nutrient content of blue corn meal were obtained from The University of Georgia College of Agricultural and Environmental Sciences, Department of Food Science and Technology. Analysis of juniper ash water was provided by the U.S. Department of Agriculture, Agricultural Research Service, Food Composition Laboratory.