NIH and WHI: time for a mea culpa and steps beyond

Abstract

The termination of the estrogen-progestin arm of the Women's Health Initiative (WHI) 5 years ago was abrupt and poorly planned. It has also become manifestly clear that the reporting at that time of the balance of risk and benefit for perimenopausal and early postmenopausal women was grossly exaggerated. Subsequent WHI publications including subanalyses of original data suggest a persistent pattern of over-reading of results toward a negative bias. The initial 2002 conclusion of the WHI investigators that harm was greater than benefit appears to be the result of several factors. One was the failure to recognize that initiation of therapy by decade of age or time since menopause was highly relevant; the WHI committee aggregated all outcome data into one group, even though in their demographic description they had the ability to investigate by age. An overhanging question is, therefore, what did they know, and when did they know it? Another factor was the utilization of a nonvalidated index termed the "global health index" that inexplicably assumed for comparison sake that all diseases were equivalent, for example, that a stroke was equivalent to a hip fracture in morbidity, mortality, and impact on quality of life. Although not a study about menopause, the data were extrapolated to all peri- and postmenopausal women. Despite the overall positive outcome of their results for women aged 50 to 60 years, most particularly those receiving estrogen-only therapy, the WHI investigators have irrationally maintained a defense of their misinterpretations of 2002. It is time for the National Institutes of Health and the WHI investigators to issue a final overall report that is clear and based on their actual results and not their personal interpretations. There is too much relevant and important information within the WHI to allow the overall study to continue to be perceived as biased to the detriment of both the National Institutes of Health and the study itself.