Childhood disadvantages and the timing of the onset of natural menopause in Latin America and the Caribbean

ABSTRACT

The aim of this study is to investigate the association of early life factors with the timing of the onset of natural menopause in Costa Rica and Puerto Rico. We use Cox proportional hazard models to estimate the risk of the onset of menopause. Our results suggest that socioeconomic disadvantages, as expressed by difficulties attending school due to economic hardships or parents never living together, increase the risk of the onset of natural menopause among Puerto Rican women. Among Costa Rican women, early life nutrition, estimated using anthropometric measures, is related to the timing of the onset of natural menopause.

KEYWORDS:

• Aging
• menopause
• childhood disadvantages

Notes

¹ Spanish acronym for “Costa Rica: Estudio de Longevidad y Envejecimiento Saludable.” Data available at http://ccp.ucr.ac.cr/creles/.

² Data available at http://prehco.rcm.upr.edu/.

³ León (2004) and Rosero-Bixby, Dow, and Fernández (2013).

⁴ The CRELES study does not ask age at first marriage. However, because women from the birth cohorts included in the sample married early, we use as proxy an indicator of whether the respondent have been ever married or in union.

⁵ Other variables like primary occupation and position at job of self and spouse were not included in the analysis due to the huge amount of missing responses. The same is true for income. McEniry (2014) estimated that the median household income, among individuals born before 1945, is 26 and 10 (expressed as country median per capita income divided by U.S. median household income per capita PPP [purchasing power parities] 2000 international dollars) for Puerto Rico and Costa Rica respectively. However, the proportion of people in poverty is around 23% in Costa Rica (World Bank, 2007) and around 46% in Puerto Rico (Bishaw & Glassman, 2016).

⁶ The PREHCO study does not ask respondents about age at menarche nor about age at first birth.

⁷ The CRELES study does not ask the date the respondent started taking estrogen, if she ever did. Therefore, the variable HT just takes into account whether the respondent ever had HT, which could easily have been after the onset of menopause. The PREHCO study does not ask respondents about the use of contraceptives.

⁸ The CRELES study does not ask the age at surgery; therefore we censored these cases at the age when menses stopped. This procedure may result in women censored at the age of the onset of natural menopause, thus limiting the time of exposure, or in women inflating the time of exposure to the risk of natural menopause. If a simple hysterectomy was conducted (at least one ovary is conserved), the ovarian function may continue for some time after surgery (WHO, 1996), then the question on the age at last menstruation is always pertinent.

⁹ As our goal is to study the contribution of childhood disadvantages to the risk of the onset of natural menopause, for the sake of clarity, Tables 5 to 8 do not include results for the control variables.

¹⁰ Table 9 only shows results for the estimates that were statistically significant in the unadjusted or in the adjusted models (Tables 5, 6, 7, and 8).