American Journal of Epidemiology
Adolescent, Adult, African Americans/statistics & numerical data, African Continental Ancestry Group, Analysis of Variance, European Continental Ancestry Group/statistics & numerical data, Female, Humans, Hypertension/epidemiology, Male, Middle Aged, Prevalence, Questionnaires, Review, United States/epidemiology, United States/ethnology
In this integrative review, the authors analyzed 25 studies on hypertension prevalence among black and white adults (1960-1991). The authors made the following inferences: 1) both female (2.59 vs. 1.77) and male (2.20 vs. 1.38) black/white hypertension prevalence ratios have diminished by approximately a third over the past three decades; 2) response rates were significantly lower among the more recent surveys (i.e., 1976 or later, mean 69.2 percent (standard deviation (SD) 6.9) vs. 1960 to 1975, mean 86.1 percent (SD 9.1)); and 3) these two trends are directly associated--response rates may account for a third (women, R2 = 0.362) to nearly a half (men, R2 = 0.469) of the variability in black-white hypertension differentials. These findings suggest that although respondent-based research has found black and white adults in the United States to be increasingly similar in hypertension prevalence, the same may not be true of the entire adult population (responders and nonresponders). The apparent diminishment over time in the black-white hypertension gap is as likely to be a methodological artifact allied with declining response rates as a true parametric phenomenon resultant from substantive factors such as enhanced treatment effectiveness among blacks.
Gorey, Kevin M.. (1998). Secular trends in the United States black/white hypertension prevalence ratio: potential impact of diminishing response rates. American Journal of Epidemiology, 147 (2), 95-102.