Evaluation of Fasting Plasma Glucose as Screening Test for NIDDM in Older Adults: Rancho Bernardo Study

Abstract
Objective: To examine the efficiency of fasting plasma glucose (FPG) as a screening test for non-insulin-dependent diabetes mellitus (NIDDM). RESEARCH AND METHODS DESIGN: A population-based evaluation was made of FPG as screening test for NIDDM in an upper middle-class white community of Rancho Bernardo, California. NIDDM was defined by 2-h postchallenge plasma glucose (PCPG) level ≥ 11.1 mM, the cutoff point recommended by the World Health Organization. Participants comprised a population-based sample of 1851 men and women 50–79 yr of age that represented 80% of surviving participants surveyed between 1972 and 1974 for the Lipid Research Clinic Prevalence Study. Those with insulin-dependent diabetes were excluded. Results: Analyses were stratified by age after logistic regression indicated that FPG and age (but not gender) were significantly related to probability of disease. As FPG cutoff points increased, sensitivity and percentage of the population to be recalled for confirmation decreased, whereas specificity and positive predictive value increased. Negative predictive value was consistently in the 90% range. Specificity did not change with age. In contrast, at virtually every FPG cutoff point, sensitivity decreased with increasing age. For example, at FPG ≥ 6.7 mM, sensitivity was 65.6% for those 50–64 yr of age and 40.0% for those 65–79 yr of age. At FPG ≥ 7.2 mM, these sensitivities were 46.9 and 28.5%, respectively. Positive predictive value increased with increasing age, reflecting the increasing prevalence of NIDDM with age. Conclusions: Poorer sensitivity with increasing age reflects the fact that the numerator of the sensitivity equation is not affected by age (mean FPG did not vary significantly between age-groups), whereas the denominator increases with age (mean PCPG increased from 6.6 mM for subjects 50–64 yr of age to 8.2 mM for subjects 65–79 yr of age). Nevertheless, because the clinical significance of increasing PCPG with age in older adults is unknown, age-specific screening criteria probably are not warranted.