The ability of antimüllerian hormone (AMH) to predict age at menopause has been reported in several studies, and a decrease in AMH level has been found to increase the probability of menopause. The rate of decline varies among women, and there is also a variability of decline between women's cycles. As a result, individualized evaluation is required to accurately predict time of menopause. To this end, we have used the AMH trajectories of individual women to predict each one's age at menopause.
From a cohort study, 266 women (ages 20-50 y) who had regular and predictable menstrual cycles at the initiation of the study were randomly selected from among 1,265 women for multiple AMH measurements. Participants were visited at approximately 3-year intervals and followed for an average of 6.5 years. Individual likelihood of menopause was predicted by fitting the shared random-effects joint model to the baseline covariates and the specific AMH trajectory of each woman.
In total, 23.7% of the women reached menopause during the follow-up period. The estimated mean (SD) AMH concentration at the time of menopause was 0.05 ng/mL (0.06 ng/mL), compared with 1.36 ng/mL (1.85 ng/mL) for those with a regular menstrual cycle at their last assessment. The decline rate in the AMH level varied among age groups, and age was a significant prognostic factor for AMH level (P < 0.001). Adjusting for age and body mass index, each woman had her own specific AMH trajectory. Lower AMH and older age had significant effects on the onset of menopause. Individualized prediction of time to menopause was obtained from the fitted model.
Longitudinal measurements of AMH will enable physicians to individualize the prediction of menopause, thereby facilitating counseling on the timing of childbearing or medical management of health issues associated with menopause.
1Department of Biostatistics
2Minimally Invasive Surgery Research Center, Iran University of Medical Sciences, Tehran, Iran
3Department of Epidemiology and Public Health, The University of Nottingham, UK
4Reproductive Endocrinology Research Center
5Endocrine Research Center, Research Institute for Endocrine Sciences, Shahid Beheshti University of Medical Sciences, Tehran, Iran
6Department of Statistics & Actuarial Science, University of Waterloo, Waterloo, Ontario, Canada.
Address correspondence to: Mahmood Reza Gohari, PhD, Department of Biostatistics, Iran University of Medical Sciences, Tehran, Iran. E-mail: firstname.lastname@example.org
Received 15 September, 2015
Revised 26 January, 2016
Accepted 26 January, 2016
Funding/support: The work of Dr Chenouri was partially supported by the Natural Sciences and Engineering Research Council of Canada.
Financial disclosure/conflicts of interest: None reported.