New Mexico Diabetes Statistics Tables (1995)
Rebecca K. Olson
Two pieces of information are needed in order to estimate the number of cases of diabetes in New Mexico: 1) New Mexico population estimates and 2) prevalence rates of diabetes for specific population groups. Once these numbers are obtained, the number of cases of diabetes can be calculated by multiplying the prevalence rates times the population.
New Mexico Population Estimates
New Mexico population estimates for 1990-1994 were obtained from the Population Distribution and Population Estimate Branches, Bureau of the Census, U.S Department of Commerce through the Data Bank, Bureau of Business and Economic Research in Albuquerque. The 1990 population estimates were based on the 1990 census. The postcensal estimates (1991-1994) were based on births, deaths and measures of migration since the 1990 census. The population estimates are broken down by county, 5-year age groups (ages 0-4, 5-9... .80-84, 85 and over), sex (male, female), and race/ethnicity (White non-Hispanic; White Hispanic; Black; American Indian, Eskimo, and Aleut; Asian and Pacific Islander; and Total Hispanic) The "Total Hispanic" category was ignored in this analysis because almost 100 percent of Hispanics in New Mexico are White Hispanic.
New Mexico population estimates for 1995 were obtained from the Bureau of Vital Records and Health Statistics, Public Health Division, New Mexico Department of Health. These estimates were only broken down by county, age, and sex. In order to calculate age-sex-race-specific population estimates within counties, it was assumed that each race subgroup increased or decreased from the 1994 estimates by the same proportion as the age-sex-county subgroup of interest. For example, if the number of 30 to 34-yearold females in Rio Arriba county was 107~ higher in 1995 compared to 1994, each race subgroup within that age-sex subgroup in Rio Arriba county was multiplied times 107% to obtain 1995 population estimates.
Diabetes Prevalence Estimates
Age- and sex-specific diabetes Prevalence rates for adult Whites, Blacks and Asians were obtained from the 1976-1980 National Health and Nutrition Examination Survey (NHANES 11)1. NHANES II is unique in that it measured both diagnosed and undiagnosed diabetes in a representative sample of the U.S. population. Prevalence rates for previously diagnosed and previously undiagnosed diabetes were broken down by sex, age (20-44 yr, 45-54 yr, 55-64 yr, and 6574 yr), and race (White and Black) . NHANES II sampled insufficient numbers of persons from other races to calculate accurate prevalence rates for them. I used prevalence rates for all races combined to calculate the number of cases of diabetes among Asians since there were no better estimates available. This probably underestimates the number of diabetics among Asians. Prevalence rates for Hispanics were obtained from the 1982-1984 Hispanic Health and Nutrition Examination Survey (HHANES) which measured diagnosed and undiagnosed diabetes among Mexican Americans, Cubans, and Puerto Ricans in the U.S.2 Prevalence rates for previously diagnosed and previously undiagnosed diabetes were broken down into the same sex and age categories as the rates from NHANES II. Hispanics in New Mexico are genetically most like Mexican Americans, having Spanish and Amerindian ancestry and little African ancestry compared with Puerto Rican or Cuban Hispanics. 3 Therefore, I used the prevalence rates given for Mexican Americans to calculate the number of cases of diabetes among Hispanics in New Mexico. [The Third National Health and Nutrition Examination Survey was conducted between 1988 and 1994. Updated prevalence rates of diagnosed and undiagnosed diabetes broken down by sex, age, and race (White, Black, Mexican American) from this survey should be available in the coming year.]
NHANES II and HHANES did not measure undiagnosed diabetes among children (age 19 or younger). Most cases of diabetes among children are assumed to be insulin-dependent diabetes mellitus (IDDM). The prevalence of IDDM among American children has been estimated at 1.7 per 1,000.4 This prevalence rate was used as the prevalence rate of diagnosed diabetes for White, Black, Asian, and Hispanic children (age 19 or younger) in New Mexico. Since IDDM is diagnosed soon after onset of illness, the prevalence rate for undiagnosed diabetes for White, Black, Asian, and Hispanic children (age 19 or younger) in New Mexico was assumed to be zero.
Prevalence rates for diagnosed diabetes among American Indians were obtained from Indian Health Service (IHS) for the years 1990 to 1995. The prevalence rates were broken down by sex and age group (ages 0-14, 15-19, 20-24 60-64, 65-74, 75 and over) Prevalence rates were obtained for two IHS administrative areas: 1) Albuquerque Area, which includes the Ute Tribe in Southern Colorado and all tribes in New Mexico with the exception of the Navajo and 2) Navajo Area, which includes the Navajo and Hopi Indian Reservations. Prevalence rates from the Navajo Area were used to calculate the number of diagnosed diabetics among American Indians in McKinley and San Juan counties; prevalence rates from the Albuquerque Area were used for the remaining counties in New Mexico.
IHS does not have prevalence rates for undiagnosed diabetes among American Indians in New Mexico. It has been argued that the percentage of undiagnosed diabetics is lower among American Indians compared to the population studied in NHANES II because health care is free for American Indians and IHS providers are aware of the increased risk of diabetes among American Indians.5 Sugarman and Percy conducted a diabetes prevalence study among adult Navajo Indians in the Teec Nos Pos Chapter in northeastern Arizona and found that only 25% of diabetics were previously undiagnosed.6 However, this is probably an underestimate because fasting or random blood glucose was used to diagnose diabetes rather than a formal glucose tolerance test. An IHS physician working with tribes in the Midwest informed me that studies there' have found that the proportion of undiagnosed diabetes ranged from approximately 19 to 30 percent. In this analysis, I assumed that one third of diabetics among adult American Indians were undiagnosed. IDDM is rare among American Indians and is mainly found in individuals with significant non-Native American ancestry.4 IHS providers are now seeing non-insulin dependent diabetes mellitus (NIDDM) in Native American teenagers. I assumed that only diabetes under the age of 10 represented IDDM and that the prevalence of undiagnosed diabetes was zero for this age group. I assumed that diabetes in children over the age of 10 mainly represented NIDDM and that the proportion of undiagnosed diabetes in these older children was the same as for adults--one third.
Valway S, Freeman W, Kaufman S, Welty T, Helgerson SD, Ghodes D. Prevalence of diagnosed diabetes among American Indians and Alaska Natives, 1987: estimates from a national outpatient data base. Diabetes Care 1993;16:271-276.
Sugarman J, Percy C. Prevalence of diabetes in a Navajo Indian community. AJPH 1989;79:511-513.