Strengths and limitations
The main strengths of the FMF study are first, prospective examination of a large population of women with singleton pregnancies attending for routine pregnancy care at 11-13 weeks’ gestation, second, we recorded maternal and pregnancy characteristics that have previously been reported to be associated with development of hypertensive disorders of pregnancy, third, we used a recently updated definition of PE which requires the development of hypertension with either proteinuria or renal insufficiency, hepatic dysfunction, thrombocytopenia, neurological complications or pulmonary edema,28 and fourth, we carried out multiple logistic regression analysis and found that significant independent contribution to PE, in addition to racial origin, was provided by maternal age, weight, height, method of conception, smoking, history of chronic hypertension or diabetes, parity, family history of PE, previous pregnancy affected by PE or birth of small for gestational age neonate.
A limitation of the FMF study is that racial origin was classified into five broad categories and it is likely that there would be variations in outcome in subgroups within each category, such as different regions of Africa and between African and Caribbean women classified as black.
The main limitations of the study relate to the findings of the systematic review of the literature and meta-analysis. For example, 16 studies provided data on the comparison of the incidence of PE between black and white women and although in most the incidence in black women was higher, the heterogeneity between studies was 98%; furthermore, only two of the studies reported adjusted ORs and adjustments were made for very few of the maternal characteristics. Similarly, there were only seven studies reporting on South Asian women and five on East Asian women, by comparison with white women and only two of these studies reported adjusted ORs. Consequently, although the combined data included more than 800 thousand black women and more than two million white women, the meta-analysis does not provide useful information on the true contribution of black race to the prediction of PE because of the heterogeneity between studies and the lack of adjustment for confounders in most of the studies; the same is true for women of South Asian and East Asian racial origin.