Platelet Counts Vary by Ethnicity, Sex, and Age: Analysis of NHANES III Data.

Background: The possibility of variation in platelet count by age and genetic background has not been examined. Identification of subpopulations with elevated platelet counts, within what is traditionally considered a normal range, may identify a group with excessive morbidity or mortality. Furthermore, platelet count differences may suggest populations in which genetic polymorphisms in regulatory proteins such as the thrombopoietin receptor influence platelet production. We hypothesized that there were differences in platelet count by ethnicity, sex and age not explained by environmental factors. Objective: To demonstrate differences in mean platelet counts by ethnicity, sex, and age while controlling for variables known to influence platelet count.

Methods and Design: We used data from the National Health, Nutrition and Examination Survey III (NHANES III), which is a multistage probability sample of the United States population with data collected between 1988 and 1994. Using appropriate weighting for the complex sampling design, the geometric mean platelet count was calculated for the total population and the population stratified by ethnicity, sex, and age, while controlling for C-reactive protein, white blood cell count, iron-deficiency, serum folate, markers of alcohol intake, presence of hepatitis B or C antibodies, and diabetes mellitus. Other potential influences, such as medications, were found not to affect the predicted counts and not included in the models.

Results: The lowest mean platelet counts were among whites (259 K/ml [95% C.I. 255–264 K/ml]) and the highest were in non-Hispanic blacks (275 K/ml [95% C.I. 270–280 K/ml]) with Mexican-Americans having intermediate values (266 K/ml [95% C.I. 261–272 K/ml]), when controlled for age and sex. Older men and women of each ethnicity consistently had lower mean platelet counts, with 60–69 years olds having mean counts approximately 7 K/ml lower than young adults (p=0.015) and 70–90 year olds having mean counts 19 K/ml lower than young adults (p<0.001). Even with controlling for iron deficiency, women had significantly higher platelet counts than men (273 K/ml [95% C.I. 269 – 278]) versus 251 K/ml [95% C.I. 245 – 256])

Conclusion: Mean platelet counts differ by ethnicity, sex, and age and these differences are not explained by covariates known to influence platelet count. This suggests that genetic influences on the platelet count are operative. These data also suggest that there may be a hormonal regulation of platelet count. The decline with age may reflect decreased stem cell function with age; alternatively, there may be a survival advantage to having a lower platelet count. Longitudinal studies of platelet count with aging are indicated.