Normal Standards for QT and QT Subintervals Derived from a Large Ethnically Diverse Population of Women Aged 50 to 79 Years (the Women’s Health Initiative [WHI])†

Available normal standards for rate-adjusted QT intervals in women are based on samples that include only whites, and no normal standards are available for QT subintervals. This study derived normal limits from percentile distributions for QT as well as QT and T-wave subintervals in 22,311 participants in the Women’s Health Initiative (WHI), including 19,059 white, 1,771 African-American, 819 Hispanic, 82 American Indian, and 580 Asian women. Excluded were women with cardiovascular disease or who were using cardioactive drugs at baseline and cardiovascular morbidity or death during the subsequent mean 6.3-year follow-up. Normal limits for QT adjusted by Bazett’s formula were strongly rate dependent, invalidating their use in practical applications. QT adjusted as a linear function of RR (QTrr) or by power functions of RR with exponent 0.5 (QTsqr) or 0.42 (QT0.42) using an appropriate regression function produced rate-invariant upper and lower normal limits for rate-adjusted QT. Adjusted QT is preferable to adjusted JT because the latter requires the incorporation of QRS duration as a covariate with RR. Normal limits were also derived for T-wave subintervals. Normal limits of QTrr in Asian women were 10 ms longer than in other ethnic groups. In conclusion, QT adjusted for rate as a linear function of RR is preferable to JT and other QT subintervals in the evaluation of QT prolongation. The adaptation of considerable revisions of the currently used limits for prolonged QT in women is suggested, with an additional race-specific adjustment in Asian women. Bazett’s formula is inappropriate for testing new drugs or other applications.