In the NHANES III combined study population of 7832 women aged 25 and older, there were 7596 women with information on their menstrual periods in the past 12 months (representing an estimated 79 million US women). After eliminating subjects reporting other reasons for apparent menopause (ie. current use of oral contraceptives, surgical menopause, chemotherapy or radiation, current pregnancy, breast feeding) and/or missing FSH information, the final study population consisted of 5029 women aged 25 and older. Based on the study smoking and menopausal classifications described above, the majority of women (57%) in the study population were nonsmokers who experienced secondhand smoke; 54% were pre-menopausal (see ). A slight majority (52%) of women were not employed, of whom the majority were post-menopausal, while 74% of the employed women were predominantly pre-menopausal ().
All US Women classified by menopausal status with smoking status and employment status
To examine mean age at menopause, there were 1825 post-menopausal women aged 25 and older with a valid date for age at last menstrual period (see ). Smokers reported a statistically significant earlier mean age (47.17 yrs) at last menstrual period compared to both nonsmokers with SHS exposure (48.59 yrs) and nonsmokers with no SHS exposure (48.55 yrs). There was no statistically significant difference between the age at last menstrual period between nonsmokers with SHS exposure (48.59 yrs) and nonsmokers with no SHS exposure (48.55 yrs). We also examined mean age at menopause among working and non working women. Using the NCHS occupational criteria, although not statistically significant, Service workers experienced the earliest age (46.95 yrs) at menopause, while White collar workers experienced the highest age (48.75 yrs) even compared to unemployed US women. Using the NIOSH industry sector NORA criteria and taking small sample sizes into account, although not statistically significant, Manufacturing workers experienced the earliest age (47.32 yrs), while Wholesale and retail trade workers reported the oldest age (49.17 yrs) of menopause even compared to unemployed US women (48.28 yrs).
Mean age at menopause of all US Women classified by age at last menstrual period with smoking status and NCHS Occupational and NORA Industrial Sector status
In order to examine the odds of earlier age at menopause, there were 2935 women 25-50 years of age. This number was further reduced to 2812 when women with “other” race/ethnicity (ie. Native American and Asian subpopulations), oophrectomy, and missing data in the regression predictors were eliminated from the analysis. The logistic regression models included: age, BMI, education, smoking exposure, and employment; the models evaluated the odds of being early post-menopausal respectively compared to pre-menopausal women. Interactions were examined; due to the significant interactions between race-ethnicity and smoking status, additional stratification of the data was performed by the 3 race-ethnicity subgroups: Whites, Blacks, and Hispanics.
In the logistic regression modeling, as would be expected, increasing age was a significant risk factor for earlier age at menopause for all 3 race-ethnicity subpopulations (see ). For all 3 race-ethnic subpopulations, there was an increased odds of earlier age at menopause with both smoking and with secondhand smoke exposure. Among Blacks, there was a statistically significant 12 times increased odds of earlier age at menopause with tobacco smoke exposure when comparing smokers to nonsmokers with no SHS exposure, and 6 times increased odds compared to nonsmokers with SHS to nonsmokers with no SHS exposure. The odds for earlier age at menopause between smokers and nonsmokers with SHS exposure was elevated 1.8 times, but not significantly. Among Whites, smokers had 1.8 times significantly increased odds compared to nonsmokers with SHS exposure; there was a 2.3 times increased odds for smokers compared to nonsmokers with no SHS, and a 1.3 times increased odds for nonsmokers with SHS compared to nonsmokers with no SHS exposure, although neither was significant. Hispanics also had significant 6.8 times increased odds of earlier age at menopause for smokers compared to nonsmokers with no SHS exposure and a 19 times increased odds for nonsmokers with SHS compared to nonsmokers with no SHS exposure. However, unlike Blacks and Whites, when Hispanic smokers were compared to nonsmokers with SHS exposure, there was a significantly decreased odds of earlier menopause, and Hispanic non smokers with SHS had an increased odds for earlier menopause than Hispanic smokers as well as Hispanic non smokers.
Logistic regression model evaluating the odds of earlier age at menopause among the study population of US women age 25-50 years
In this logistic model controlling for several variables, BMI was not a significant risk factor for earlier age at menopause for any of the race-ethnic groups. Educational level was not a significant risk factor for earlier age at menopause, although for all 3 race-ethnic groups, there appeared to be an increased odds of earlier age at menopause with fewer years of education. Finally, although not statistically significant, employment appeared to protect against the odds of earlier age at menopause for Blacks and Hispanics, but not for Whites.