Methods and Limitations
We estimated the duration of diabetes using data from CDC’s National Health Interview Survey (NHIS) of the National Center for Health Statistics (NCHS). Conducted continuously since 1957, the survey provides information on the health of the civilian noninstitutionalized population of the United States. The multistage probability design of the survey has been described elsewhere.1 Adult respondents were asked whether a health professional had ever told them they had diabetes. To exclude gestational diabetes, women were asked whether they had been told they had diabetes other than during pregnancy. Adults who reported being diagnosed with diabetes were then asked at what age they were diagnosed. Duration was calculated by subtracting age at diagnosis from current age. The analyses were limited to adults aged 18–79 years because age at diagnosis of diabetes was not collected for people aged 17 years and younger, and there were limited data available on age at diagnosis of diabetes among adults aged 80 years and older.
We reported both medians and means because of the skewness of the distribution of diabetes duration. Medians of diabetes duration were presented by sex, age, and race/ethnicity. People of Hispanic origin may be of any race. The race groups included people of both Hispanic and non-Hispanic origin. Boxplots were constructed to display the distribution of duration for each year. Minimum values were always zero (indicating duration was less than 1 year) and maximum values were calculated as the 75th percentile plus 1.5 times the interquartile range (75th percentile–25th percentile). Three-year averages were used to improve the precision of annual estimates.
People with diabetes may have diabetes many years before diagnosis.2 Therefore, the National Health Interview Survey (NHIS) may underestimate the true duration of diabetes. With these cross-sectional data it is impossible to determine the causes of an increase or decrease in diabetes duration. Duration change can be influenced both by survival and incidence.
- Botman SL, Moore TF, Moriarity CL, Parsons VL. Design and estimation for the National Health Interview Survey, 1995–2004. National Center for Health Statistics, CDC. Vital Health Stat. 2000;2(130).
- Harris MI, Klein R, Welborn TA, Knuiman MW. Onset of NIDDM occurs at least 4-7 years before clinical diagnosis. Diabetes Care 1992;15(7):815–819.