Abstract
Objectives. This study assessed predictors of the onset of natural menopause in African American women.
Methods. We used mailed questionnaires to collect data at baseline in 1995 and during follow-up from Black Women’s Health Study participants. Cox proportional hazards regression was used to assess potential predictors—including experiences of racism—of the onset of natural menopause among 17 070 women aged 35 to 55 years and premenopausal in 1995.
Results. The hazard ratio (HR) was 1.43 for current smokers (95% confidence interval [CI] = 1.24, 1.66) and 1.21 (95% CI = 1.06, 1.38) for ex-smokers and significantly less for obese women and oral contraceptive users. Hazard ratios for most questions about racism were elevated by 10% to 30% but were not statistically significant.
Conclusions.Earlier onset of natural menopause among African American women is strongly associated with smoking and inversely associated with body mass index and oral contraceptive use.
A woman’s age at the onset of natural menopause is associated with several serious illnesses. For example, early menopause has been associated with an increased risk of osteoporosis1,2 and with an increased overall mortality.3,4 Cigarette smoking is the only factor that has been consistently linked to earlier natural menopause.5–13 Nulliparity,12,14–16 nonuse of oral contraceptives,12,16–18 low body mass index,10,19 and low educational status12,16,20 have also been associated with earlier menopause, but the evidence is inconsistent.
Almost all available data on determinants of age at natural menopause in American women have come from studies of White women. The exceptions are Bromberger et al.’s6 prospective study of age at natural menopause, which included 19 Black women and Gold et al.’s12 cross-sectional study of natural menopause, which included 4157 African American women. One of the findings of the Bromberger study was an association between increased stress and an earlier age of menopause among the Black women. In our report, we prospectively assessed the relation of demographic, reproductive, and health factors to age at natural menopause among US Black women, based on data from a follow-up study of 64 500 African American women. To pursue a hypothesis raised by the study of Bromberger et al.,6 we also assessed the relation of participants’ perceived experiences of racism to the risk of early menopause.
METHODS
Data Collection
In 1995, 64 500 US Black women aged 21 to 69 years from all areas of the United States enrolled in the Black Women’s Health Study by completing health questionnaires.21–24 The questionnaires were mailed to subscribers to Essence magazine, members of selected Black professional organizations, and friends and relatives of early respondents. The participants were sent follow-up questionnaires in 1997 and 1999: 82% completed the 1997 questionnaire and 80% completed the 1999 questionnaire.
The 1995 questionnaire asked about many factors, including current weight and height, history of oral contraceptive use, vigorous physical activity in the previous year, current and past cigarette smoking, parity, and age at menarche. With respect to menopause, the women were asked in 1995: “Have your menstrual periods stopped permanently (yes/ yes but have periods now due to female hormones/ no/ not sure).” The 1997 questionnaire asked: “Have you gone through menopause (yes/ yes but have periods now due to female hormones/ no but I am currently going through menopause/ no/ uncertain); If yes (gone through menopause), what was the age at menopause and what caused the menopause (natural menopause/ surgery/ medication, chemotherapy, radiation).” The 1999 questionnaire asked: “Have your menstrual periods stopped permanently (menopause) (yes/ yes but have periods now due to female hormones/ no/ uncertain/ never had periods).” If periods had ceased permanently, the women were asked the age at which their periods stopped and the reason (natural menopause/ surgery/ medications, chemotherapy, radiation). The 1997 questionnaire also asked questions about perceptions of experiences of racism. There were 5 questions about experiences in daily life: “In your day-to-day life, how often has it happened that you received poorer service than other people in restaurants or stores; people acted as if they thought you were not intelligent; people acted as if they were afraid of you; people acted as if they thought you were dishonest; people acted as if they were better than you.” There were 3 questions about unfair treatment: “Have you ever been treated unfairly due to your race in the area of: job (hiring, promotion, firing); housing (renting, buying, mortgage); police (stopped, searched, threatened).” These were taken from a set of questions used by Williams et al.25 in a study of racial differences in physical and mental health; a high score was positively related to several measures of poor health in that study. Finally, participants were asked how often they thought about their race, with a question developed by C. Jones.26
In our study, menopause was defined as permanent cessation of menstrual periods. To assess the degree of misclassification of natural menopause, we examined the 1999 questionnaire responses of the 253 women who reported a natural menopause on the 1997 questionnaire. Of these, 224 had completed the menopause section of the 1999 questionnaire; 88% reported having had a natural menopause, suggesting that misclassification of age at menopause was not substantial. The remaining 12% indicated that they were currently “going through menopause” or were “uncertain” about their menopausal status; these women were not considered to have undergone natural menopause. It was not possible to perform similar data cleaning for women who first reported the occurrence of natural menopause on the 1999 questionnaire because subsequent data on their menopausal status were not available, but there is no reason to expect different results.
Data Analysis
Our analytic sample comprised all Black Women’s Health Study participants who were premenopausal and aged 35 to 55 years at baseline (1995) and who had completed at least 1 follow-up questionnaire (1997 or 1999) (n = 17 070). Our analytic goal was to estimate the incidence of menopause in relation to selected baseline characteristics and perceived experiences of racism. All women were considered “at risk” for menopause until the occurrence of natural menopause, the end of their follow-up period in the study (maximum of 4 years), or the occurrence of a censoring event. Women were censored at the midpoint of the follow-up interval in which they reported hysterectomy (n = 670), bilateral oophorectomy (n = 553), medical menopause (n = 82), or use of hormone replacement therapy or oral contraceptives (n = 1646).
Proportional hazards regression models were used to derive hazard ratios (HRs) for age at natural menopause in relation to selected baseline characteristics and perceived experiences of racism, with adjustment for potential confounders.27 The hazard function for each variable was related to the conditional probability of reaching menopause at age t, given that a woman had not reached menopause by age t. We used the PHREG procedure in SAS (version 8.0; SAS Institute Inc, Cary, NC), with the “exact” option to handle ties, as age was measured to the nearest year.28
We assessed factors that, on theoretical grounds, may be associated with natural menopause or that have been shown to be associated with natural menopause in previous studies. Our final multivariate model included the following covariates measured at baseline: parity (nulliparous, 1, 2, 3, ≥ 4 births), age at menarche (<11, 11, 12–13, 14, ≥15 years), years of oral contraceptive use (never use, 1–4, 5–9, ≥10 years), unilateral oophorectomy, smoking status (never, past, current), years of education (<12, 12, 13–15, 16, ≥17), vigorous physical activity (none, < 5 hours per week, ≥ 5 hours per week), and body mass index (in kilograms per meter2) (< 20, 20–24, 25–29, ≥ 30). Women with missing information on any covariate (n = 760) were dropped from the final analysis. We also assessed the estimated number of ovulatory cycles, defined as follows: [(age at baseline – age at menarche) – (parity + years of oral contraceptive use)] ÷ (average cycle length/365); we did not have information on cycle length and used 28 for the average cycle length in the formula.29
The assumption of proportional hazards was evaluated for the final model by plotting the complementary log-log transform and assessing parallelism of the exposure-specific survival curves.28 To address the possibility that hormone replacement therapy may be related to the onset of menopausal symptoms and may differ by socioeconomic position,30 we conducted separate analyses28,31 in which the 724 women who initiated hormone replacement use before menopause were reclassified as having had the outcome event. We also ran models in which premenopausal women who used oral contraceptives (n = 922) were considered premenopausal throughout the follow-up period rather than being censored. Because these analyses yielded results similar to those with the original outcome definition, we present data only from the original analysis.
RESULTS
Baseline demographic, lifestyle, and health characteristics of the 17 070 participants who were premenopausal at the start of follow-up are presented in Table 1 ▶ according to their menstrual status at the end of follow-up. In total, 1323 women reached a natural menopause (reported by 253 women on the 1997 questionnaire and by 1070 on the 1999 questionnaire) and 12 796 women were still premenopausal at the end of follow-up. The remainder underwent either surgical menopause or medical menopause, or began taking hormones before the permanent cessation of menses. The mean age at natural menopause was 49.6 years (SD = 3.71); the median age was 50 years (interquartile range = 47, 52).
TABLE 1.
—Demographic, Lifestyle, and Health Characteristics of 17 070 Participants Who Were Premenopausal at Baseline, by Menopausal Status at Follow-Up: Black Women’s Health Study, United States, 1995–1999
| Menopausal Status at Follow-Up | |||||
| Baseline characteristic | Premenopausal (n = 12 796), No. (%) | Premenopausal, Hormone Use (n = 1646), No. (%) | Surgical Menopause (n = 1223), No. (%) | Medical Menopause (n = 82), No. (%) | Natural Menopause (n = 1323), No. (%) |
| Age at baseline, median (IQ) | 40 (37–43) | 41 (37–45) | 42 (38–45) | 44 (40–47) | 48 (45-50) |
| Smoking status | |||||
| Never smoker | 7495 (58.6) | 1003 (60.9) | 750 (61.3) | 47 (57.3) | 672 (50.8) |
| Past smoker | 2716 (21.2) | 361 (21.9) | 271 (22.2) | 19 (23.2) | 370 (28.0) |
| Current smoker | 2541 (19.9) | 271 (16.5) | 196 (16.0) | 16 (19.5) | 275 (20.8) |
| Missing | 44 (0.34) | 11 (0.67) | 6 (0.49) | 0 (0.0) | 6 (0.45) |
| Years of age at menarche | |||||
| <11 | 1382 (10.8) | 187 (11.3) | 148 (12.1) | 8 (9.8) | 152 (11.5) |
| 11 | 1977 (15.5) | 263 (16.0) | 197 (16.1) | 23 (28.0) | 196 (14.8) |
| 12–13 | 6783 (53.0) | 859 (52.2) | 669 (54.7) | 35 (42.7) | 699 (52.8) |
| 14 | 1336 (10.4) | 166 (10.1) | 108 (8.8) | 5 (6.1) | 133 (10.1) |
| ≥15 | 1257 (9.8) | 163 (9.9) | 96 (7.9) | 11 (13.4) | 136 (10.3) |
| Missing | 61 (0.5) | 8 (0.5) | 5 (0.4) | 0 (0.0) | 7 (0.5) |
| Parity | |||||
| Nulliparous | 3058 (23.9) | 440 (26.7) | 223 (18.2) | 18 (22.0) | 206 (15.6) |
| Parous | 9605 (75.1) | 1181 (71.8) | 983 (80.4) | 64 (78.0) | 1106 (83.6) |
| Missing | 133 (1.0) | 25 (1.5) | 17 (1.4) | 0 (0.0) | 11 (0.8) |
| Years of oral contraceptive use | |||||
| <1 | 5385 (42.1) | 522 (31.7) | 490 (40.1) | 34 (41.5) | 632 (47.8) |
| ≥1 | 7411 (57.9) | 1124 (68.3) | 733 (59.9) | 48 (58.5) | 691 (52.2) |
| Years of education | |||||
| <12 | 202 (1.6) | 15 (0.9) | 23 (1.9) | 1 (1.2) | 32 (2.4) |
| 12 | 1703 (13.3) | 211 (12.8) | 193 (15.8) | 16 (19.5) | 205 (15.5) |
| 13–15 | 4576 (35.8) | 548 (33.3) | 459 (37.5) | 26 (31.7) | 410 (31.0) |
| 16 | 2881 (22.5) | 404 (24.5) | 269 (22.0) | 13 (15.9) | 256 (19.4) |
| ≥17 | 3261 (25.5) | 445 (27.1) | 271 (22.2) | 26 (31.7) | 401 (30.3) |
| Missing | 173 (1.3) | 23 (1.4) | 8 (0.6) | 0 (0.0) | 19 (1.4) |
| Body mass index in kg/m2 | |||||
| < 20 | 530 (4.1) | 68 (4.2) | 25 (2.0) | 2 (2.4) | 39 (3.0) |
| 20–24 | 3957 (30.9) | 524 (31.8) | 336 (27.5) | 19 (23.2) | 392 (29.6) |
| 25–29 | 4041 (31.6) | 543 (33.0) | 429 (35.1) | 32 (39.0) | 451 (34.1) |
| ≥ 30 | 4078 (31.9) | 476 (28.9) | 414 (33.9) | 28 (34.2) | 413 (31.2) |
| Missing | 190 (1.5) | 35 (2.1) | 19 (1.5) | 1 (1.2) | 28 (2.1) |
| Unilateral oophorectomy | |||||
| No | 12 526 (97.9) | 1595 (96.9) | 1199 (98.0) | 80 (97.6) | 1290 (97.5) |
| Yes | 270 (2.1) | 51 (3.1) | 24 (2.0) | 2 (2.4) | 33 (2.5) |
| Vigorous physical activity | |||||
| None | 4734 (37.0) | 577 (35.1) | 487 (39.8) | 35 (42.7) | 593 (44.8) |
| < 5 h/wk | 6481 (50.6) | 852 (51.7) | 605 (49.5) | 43 (52.4) | 596 (45.1) |
| ≥ 5 h/wk | 1581 (12.4) | 217 (13.2) | 131 (10.7) | 4 (4.9) | 134 (10.1) |
Note. IQ = interquartile range: 25th to 75th percentile. Follow-up was at 4 years.
Hazard ratios are given in Table 2 ▶ as estimates of the incidence rate ratios for these factors in relation to the occurrence of natural menopause. Cigarette smoking was most strongly associated with early menopause: the hazard ratio was 1.43 (95% confidence interval [CI] = 1.24, 1.66) for current smokers and 1.21 (95% CI = 1.06, 1.38) for past smokers relative to never smokers. A dose response–relation with the duration and intensity of cigarette smoking was found, with the hazard ratio for pack-years of smoking increasing to 1.52 (95% CI = 1.30, 1.79) for the highest quartile (P for trend = .0048). High body mass index was associated with a reduced occurrence of menopause: the hazard ratio for women in the highest category, ≥ 30, was 0.78 (95% CI = 0.67, 0.90) relative to women with a body mass index of 20 to 24 years. The hazard ratio was also reduced for women who had used oral contraceptives for at least 1 year relative to women who used them for shorter periods or not at all (0.86 [95% CI = 0.77, 0.96]), but there was no further reduction associated with longer-term use. The hazard ratio for parous women relative to nulliparous women was 0.93, but the confidence interval included 1.0, and there was no trend by the number of births. The hazard ratio for unilateral oophorectomy was 1.17 (95% CI = 0.82, 1.66). Age at menarche, years of education, and vigorous physical activity were not related to the occurrence of menopause. The number of lifetime ovulatory cycles also was not associated with menopause (HR for ≥ 300 cycles relative to < 300 was 1.03 [95% CI = 0.87, 1.24]).
TABLE 2.
—Hazard Ratios and 95% Confidence Intervals for Occurrence of Natural Menopause, by Selected Baseline Characteristics: Black Women’s Health Study, United States, 1995–1999 (n = 17 070)
| Baseline characteristics | Crude HR (95% CI) | Adjusted HRa (95% CI) |
| Smoking status | ||
| Never smoker | 1.00b | 1.00b |
| Past smoker | 1.17 (1.03, 1.33) | 1.21 (1.06, 1.38) |
| Current smoker | 1.46 (1.27, 1.68) | 1.43 (1.24, 1.66) |
| Pack years | ||
| Never smoker | 1.00b | 1.00b |
| 1st quartile | 1.23 (0.98, 1.53) | 1.20 (0.95, 1.51) |
| 2nd quartile | 1.11 (0.92, 1.34) | 1.13 (0.93, 1.36) |
| 3rd quartile | 1.32 (1.11, 1.56) | 1.33 (1.12, 1.58) |
| 4th quartile | 1.50 (1.29, 1.75) | 1.52 (1.30, 1.79) |
| Test for trend: P < .0144 | Test for trend: P < .0048 | |
| Parity | ||
| Nulliparous | 1.00b | 1.00b |
| Parous | 0.93 (0.80, 1.08) | 0.93 (0.80, 1.09) |
| 1 birth | 0.98 (0.82, 1.17) | 0.96 (0.80, 1.15) |
| 2 births | 0.90 (0.76, 1.07) | 0.90 (0.76, 1.07) |
| 3 births | 0.93 (0.77, 1.13) | 0.94 (0.77, 1.14) |
| ≥ 4 births | 0.93 (0.76, 1.14) | 0.94 (0.76, 1.16) |
| Oral contraceptive use, y | ||
| <1 | 1.00b | 1.00b |
| ≥1 | 0.89 (0.80, 0.99) | 0.86 (0.77, 0.96) |
| 1–4 | 0.88 (0.77, 1.01) | 0.86 (0.74, 0.99) |
| 5–9 | 0.86 (0.73, 1.00) | 0.84 (0.71, 0.99) |
| ≥10 | 0.97 (0.82, 1.15) | 0.90 (0.75, 1.08) |
| Age at menarche, y | ||
| <11 | 1.07 (0.85, 1.35) | 1.07 (0.84, 1.36) |
| 11 | 0.95 (0.76, 1.18) | 0.97 (0.77, 1.21) |
| 12–13 | 0.94 (0.78, 1.13) | 0.95 (0.78, 1.14) |
| 14 | 0.92 (0.73, 1.17) | 0.88 (0.69, 1.12) |
| ≥15 | 1.00b | 1.00b |
| Education, y | ||
| <12 | 1.13 (0.78, 1.62) | 1.03 (0.69, 1.53) |
| 12 | 1.00 (0.84, 1.18) | 0.99 (0.83, 1.18) |
| 13–15 | 0.99 (0.87, 1.14) | 0.96 (0.83, 1.11) |
| 16 | 1.11 (0.95, 1.30) | 1.11 (0.95, 1.31) |
| ≥17 | 1.00b | 1.00b |
| Body mass index in kg/m2 | ||
| < 20 | 1.06 (0.76, 1.47) | 1.03 (0.74, 1.44) |
| 20–24 | 1.00b | 1.00b |
| 25–29 | 0.89 (0.78, 1.02) | 0.89 (0.77, 1.02) |
| ≥ 30 | 0.81 (0.70, 0.93) | 0.78 (0.67, 0.90) |
| Unilateral oophorectomy | 1.16 (0.82, 1.65) | 1.17 (0.82, 1.66) |
| Vigorous physical activity | ||
| None | 1.00b | 1.00b |
| < 5 h/wk | 0.99 (0.88, 1.11) | 0.97 (0.86, 1.10) |
| ≥ 5 h/wk | 1.02 (0.84, 1.23) | 0.96 (0.79, 1.17) |
Note. HR = hazard ratio; CI = confidence interval.
aAdjusted for all other variables in table, using parity (0, 1, 2, 3, ≥ 4 births), years of oral contraceptive use (never use, 1–4, 5–9, ≥10), and smoking status (never, past, current). Pack-years of smoking in quartiles were modeled separately.
bReference category.
The analyses were repeated separately according to age at the start of follow-up, 35 to 44 or 45 to 55 years, with the former providing information on very early menopause. Unilateral oophorectomy was strongly associated with age at menopause among the younger women (HR = 2.00 [95% CI = 1.09, 3.68]), but not among the older women (HR = 0.87 [95% CI = 0.54, 1.39]). Cigarette smoking, body mass index, and oral contraceptive use were associated with age at menopause in both groups.
Factors concerning perceived experiences of racism are presented in Table 3 ▶ according to menstrual status at the end of follow-up, and hazard ratios for the association between perceived experiences of racism and the occurrence of natural menopause are given in Table 4 ▶. For the 5 questions on how people acted toward the participant, most hazard ratios exceeded 1.0 but were not statistically significant. There was a significant association between risk of menopause and having experienced “people act as if I am not intelligent” once a month (HR = 1.32 [95% CI = 1.06, 1.65]). Hazard ratios for positive responses to questions about experiencing discrimination in the areas of employment, housing, or the police were all close to 1.0. For the question on how often the participant thinks about her race, the hazard ratio for the category “once a day” was significantly elevated (1.23 [95% CI = 1.02, 1.47]), but the estimate for the highest category (“constantly”) was only slightly elevated (1.05 [95% CI = .87, 1.27]). For all 9 questions concerning race and racism, the hazard ratio was elevated for the missing category. A comparison of women who did not answer 1 or more of the racism questions with those who answered all of the questions indicated that these groups were similar with respect to most of the study variables but had fewer years of education. We repeated the analyses of racism questions within subgroups defined according to age and educational status and found no evidence of interaction.
TABLE 3.
—Perceived Experiences of Racism of 17 070 Participants Who Were Premenopausal at Baseline, by Menstrual Status at Follow-Up: Black Women’s Health Study, United States, 1995–1999
| Racism Question | Premenopausal (n = 12 796), No. (%) | Premenopausal, Hormone Use (n = 1646), No. (%) | Surgical Menopause (n = 1223), No. (%) | Medical Menopause (n = 82), No. (%) | Natural Menopause (n = 1323), No. (%) |
| Did not complete 1997 questionnaire | 878 (6.9) | 82 (5.0) | 89 (7.3) | 7 (8.5) | 103 (7.8) |
| Receive poorer service than others | |||||
| <1/mo | 9807 (76.6) | 1315 (80.0) | 934 (76.4) | 60 (73.2) | 1053 (79.6) |
| 1/mo | 934 (7.3) | 118 (7.2) | 82 (6.7) | 5 (6.1) | 58 (4.4) |
| ≥1/wk | 1030 (8.0) | 112 (6.8) | 99 (8.1) | 9 (11.0) | 89 (6.7) |
| Missing | 147 (1.2) | 19 (1.2) | 19 (1.5) | 1 (1.2) | 20 (1.5) |
| People act as if I am not intelligent | |||||
| <1/mo | 8794 (68.7) | 1177 (71.5) | 810 (66.2) | 56 (68.3) | 919 (69.4) |
| 1/mo | 954 (7.5) | 126 (7.6) | 72 (5.9) | 4 (4.9) | 91 (6.9) |
| ≥1/wk | 2027 (15.8) | 242 (14.7) | 232 (19.0) | 14 (17.1) | 185 (14.0) |
| Missing | 143 (1.1) | 19 (1.2) | 20 (1.6) | 1 (1.2) | 25 (1.9) |
| People act as if they are afraid of me | |||||
| <1/mo | 9991 (78.1) | 1335 (81.1) | 964 (78.8) | 58 (70.8) | 1047 (79.1) |
| 1/mo | 578 (4.5) | 91 (5.5) | 40 (3.3) | 2 (2.4) | 48 (3.6) |
| ≥1/wk | 1202 (9.4) | 123 (7.5) | 117 (9.6) | 13 (15.9) | 103 (7.8) |
| Missing | 147 (1.1) | 15 (0.9) | 13 (1.1) | 2 (2.4) | 22 (1.7) |
| People think I am dishonest | |||||
| <1/mo | 10 345 (80.8) | 1376 (83.6) | 975 (79.7) | 66 (80.5) | 1078 (81.5) |
| 1/mo | 505 (4.0) | 69 (4.2) | 54 (4.4) | 1 (1.2) | 45 (3.4) |
| ≥1/wk | 875 (6.8) | 98 (5.9) | 87 (7.1) | 7 (8.5) | 74 (5.6) |
| Missing | 193 (1.5) | 21 (1.3) | 18 (1.5) | 1 (1.2) | 23 (1.7) |
| People think they are better than me | |||||
| <1/mo | 7818 (61.1) | 1053 (64.0) | 741 (60.6) | 46 (56.1) | 849 (64.2) |
| 1/mo | 1101 (8.6) | 145 (8.8) | 107 (8.7) | 5 (6.1) | 84 (6.3) |
| ≥1/wk | 2793 (21.8) | 344 (21.0) | 270 (22.1) | 22 (26.8) | 259 (19.6) |
| Missing | 206 (1.6) | 22 (1.3) | 16 (1.3) | 2 (2.4) | 28 (2.1) |
| I have been treated unfairly at job | |||||
| No | 7269 (56.8) | 938 (57.0) | 688 (56.3) | 41 (50.0) | 743 (56.2) |
| Yes | 4370 (34.1) | 588 (35.7) | 420 (34.3) | 32 (39.0) | 438 (33.1) |
| Missing | 279 (2.2) | 38 (2.3) | 26 (2.1) | 2 (2.4) | 39 (2.9) |
| I have been treated unfairly in housing | |||||
| No | 4648 (36.3) | 589 (35.8) | 420 (34.3) | 31 (37.8) | 493 (37.3) |
| Yes | 6790 (53.1) | 907 (55.1) | 662 (54.1) | 41 (50.0) | 658 (49.7) |
| Missing | 480 (3.7) | 68 (4.1) | 52 (4.3) | 3 (3.7) | 69 (5.2) |
| I have been treated unfairly by police | |||||
| No | 2913 (22.8) | 355 (21.6) | 231 (18.9) | 23 (28.1) | 276 (20.8) |
| Yes | 8535 (66.7) | 1149 (69.8) | 846 (69.2) | 51 (62.2) | 865 (65.4) |
| Missing | 470 (3.7) | 60 (3.6) | 57 (4.6) | 1 (1.2) | 79 (6.0) |
| How often I think about my race | |||||
| Never—1/yr | 1783 (13.9) | 243 (14.8) | 173 (14.1) | 15 (18.3) | 200 (15.1) |
| 1/mo—1/wk | 3177 (24.8) | 445 (27.0) | 308 (25.2) | 21 (25.6) | 285 (21.6) |
| 1/d | 3555 (27.8) | 441 (26.8) | 306 (25.0) | 19 (23.2) | 392 (29.6) |
| Constantly | 2984 (23.3) | 367 (22.3) | 288 (23.6) | 15 (18.3) | 279 (21.1) |
| Missing | 419 (3.3) | 68 (4.1) | 59 (4.8) | 5 (6.1) | 64 (4.8) |
Note. Follow-up was at 4 years.
TABLE 4.
—Hazard Ratios and 95% Confidence Intervals for Occurrence of Natural Menopause, by Perceived Experiences of Racism: Black Women’s Health Study, United States, 1995–1999
| Racism Question | Crude HR (95% CI) | Adjusted HRa (95% CI) |
| Receive poorer service than others | ||
| <1/mo | 1.00b | 1.00b |
| 1/mo | 0.89 (0.68, 1.17) | 0.91 (0.70, 1.19) |
| ≥1/wk | 1.00 (0.80, 1.24) | 1.00 (0.80, 1.25) |
| Missing | 1.10 (0.71, 1.72) | 1.11 (0.70, 1.78) |
| People act as if I am not intelligent | ||
| <1/mo | 1.00b | 1.00b |
| 1/mo | 1.32 (1.07, 1.64) | 1.32 (1.06, 1.65) |
| ≥1/wk | 1.14 (0.97, 1.33) | 1.11 (0.94, 1.31) |
| Missing | 1.57 (1.05, 2.33) | 1.71 (1.14, 2.26) |
| People act as if they are afraid of me | ||
| <1/mo | 1.00b | 1.00b |
| 1/mo | 1.19 (0.89, 1.59) | 1.15 (0.85, 1.56) |
| ≥1/wk | 1.13 (0.92, 1.38) | 1.09 (0.88, 1.35) |
| Missing | 1.10 (0.72, 1.68) | 1.27 (0.83, 1.95) |
| People act as if they think I am dishonest | ||
| <1/mo | 1.00b | 1.00b |
| 1/mo | 1.15 (0.85, 1.55) | 1.13 (0.82, 1.55) |
| ≥1/wk | 1.13 (0.89, 1.43) | 1.12 (0.88, 1.43) |
| Missing | 1.40 (0.93, 2.13) | 1.49 (0.97, 2.28) |
| People act as if they are better than me | ||
| <1/mo | 1.00b | 1.00b |
| 1/mo | 0.97 (0.78, 1.22) | 0.97 (0.77, 1.23) |
| ≥1/wk | 1.10 (0.96, 1.27) | 1.10 (0.96, 1.28) |
| Missing | 1.24 (0.85, 1.81) | 1.40 (0.95, 2.04) |
| Job discrimination | ||
| No | 1.00b | 1.00b |
| Yes | 1.01 (0.90, 1.14) | 0.98 (0.86, 1.10) |
| Missing | 1.22 (0.88, 1.69) | 1.28 (0.92, 1.79) |
| Housing discrimination | ||
| No | 1.00b | 1.00b |
| Yes | 0.97 (0.86, 1.09) | 1.00 (0.88, 1.13) |
| Missing | 1.17 (0.91, 1.50) | 1.25 (0.97, 1.61) |
| Police discrimination | ||
| No | 1.00b | 1.00b |
| Yes | 1.01 (0.88, 1.16) | 0.97 (0.84, 1.12) |
| Missing | 1.35 (1.05, 1.73) | 1.36 (1.07, 1.73) |
| How often I think about my race | ||
| Never–1/y | 1.00b | 1.00b |
| 1/mo–1/wk | 1.00 (0.83, 1.20) | 0.99 (0.82, 1.19) |
| 1/d | 1.22 (1.03, 1.44) | 1.23 (1.02, 1.47) |
| Constantly | 1.06 (0.88, 1.27) | 1.05 (0.87, 1.27) |
| Missing | 1.17 (0.88, 1.55) | 1.13 (0.84, 1.52) |
Note. HR = hazard ratio; CI = confidence interval.
aAdjusted for parity (0, 1, 2, 3, ≥ 4 births), age at menarche (< 11, 11, 12–13, 14, ≥ 15 y), years of oral contraceptive use (never use, 1–4, 5–9, ≥ 10), unilateral oophorectomy, smoking status (never, past, current), years of education (< 12, 12, 13–15, 16, ≥ 17), vigorous physical activity (none, < 5 h/wk, ≥ 5 h/wk), and body mass index (< 20, 20–24, 25–29, ≥ 30).
bReference category.
The questions on racism were asked at the first follow-up, in 1997. We assumed that responses to those questions represented the individual’s general experiences of racism rather than experiences at 1 particular time, and therefore we considered these variables to be “baseline” characteristics of the participants. To check this assumption, we repeated our analyses, beginning follow-up in 1997 rather than 1995, and found the results to be unchanged.
DISCUSSION
The median age at menopause in this study falls in the range of values reported in follow-up studies of White women.10,32 In the small study by Bromberger et al.,6 the median age at menopause in Black women was 49.3 years.
The permanent cessation of menstrual periods is thought to be due to the depletion of ovarian follicles.33 Follicles are depleted by ovulation and by atresia or degeneration. Previous studies indicate that current cigarette smoking leads to an earlier age at menopause5–13; results on past smoking are mixed.10,12,13 One possible mechanism is that tobacco smoke constituents such as polycyclic hydrocarbons may destroy primordial oocytes or may contribute to degeneration of follicles through lowered estrogen levels.34–36 Our results suggest that both current and past smoking result in an earlier menopause among US Black women. The effect was stronger for current smokers and was greatest for women who had the heaviest exposure to cigarette smoke, as estimated by total pack-years of smoking.
Factors related to the lifetime number of ovulatory cycles, and therefore possibly to age at menopause, include age at menarche, parity, and oral contraceptive use. We found no relation between age at menarche and age at menopause, in agreement with most previous studies on the topic.14,16 A number of studies have indicated that nulliparous women and oral contraceptive users have an earlier menopause,12,14–16 and some have suggested that age at menopause increases with increasing parity.15 We did not find an association with number of births and observed only a weak (nonsignificant) association with nulliparity. The use of oral contraceptives for 1 year or longer was associated with a later age at menopause, although we did not find a stronger association for long-term users. Cramer and Xu29 suggested that a measure of the lifetime number of ovulatory cycles may be a good predictor of age at menopause. We found no association between the number of ovulatory cycles and age at menopause. However, this measure was imprecise because we did not have information on the participants’ usual cycle length, which may contribute much of the variation in the number of lifetime ovulatory cycles.
Another reproductive variable, a history of unilateral oophorectomy without hysterectomy, has been linked with an early age at menopause in 2 previous studies.15,37 Such surgical procedures deplete a woman’s ovarian follicles and may predispose to a hypoestrogenic state. In our study, there was little association between unilateral oophorectomy and age at menopause overall, but a strong association among the younger women. The study by Cramer et al.37 was also a study of very early menopause—the average age at menopause was 42.2 years
Obese women had a reduced occurrence of early menopause in our study. Obesity has been associated with inadequate ovarian function, which could result in fewer ovulatory cycles and less depletion of follicles.38 In addition, the production of estrone in the adipose tissue of obese women may result in higher estrogen levels, which may lead to less follicular degeneration. Some previous studies found an earlier age at menopause among women with a low body mass index,10,19 but others did not.8,12,13,20
A prospective study of the determinants of age at menopause in 185 healthy US women found an earlier age at menopause for the African American women (n = 19) and also found that psychosocial stress was predictive of an earlier menopause among those women.6 In another study, by Harlow et al.,39 depression was associated with earlier menopause. It appears that stress activates the hypothalamic-pituitary-adrenal axis, leading to increased secretion of glucocorticoids with inhibition of release of gonadotropinreleasing hormone and suppression of ovulation.40,41 This mechanism is thought to explain the well-established link between stress and hypothalamic amenorrhea.42,43 However, it is not clear how this pathway would lead to an earlier menopause. Racism can act as a chronic stressor, and in recent years there has been increasing interest in the potential health effects of daily life experiences of racism.44–47 In our study, participants answered 9 questions related to experiences of racism; statistically significant associations were observed for 2. These could be chance associations given the number of possible outcomes for these questions and the lack of a “dose–response relation.” However, the hazard ratios for most categories of response for the questions were over 1.0. The largest and most consistent associations with the occurrence of menopause were for categories of missing response. Further research is needed to explore whether not answering sensitive questions like these may in some way be a marker for the experience of racism.
The data in our study were collected prospectively, minimizing the potential for recall bias and uncertainty about the timing of factors of interest in relation to the occurrence of menopause. The sample is much larger than that in only previous study to assess age at menopause prospectively in US Black women6 and larger than that in a crosssectional study that assessed natural menopause in Black women.12 As in other studies of age at menopause, the analyses were based on self-reports; random misclassification would have tended to weaken associations. Previous US studies have indicated a high degree of accuracy in the self-reporting of hysterectomy and oophorectomy.48,49 The usual definition of menopause is permanent cessation of menstrual periods, indicated by a lack of menstruation for at least 12 months. In our study, permanent cessation of periods was self-defined. To assess the degree of misclassification, we checked 1999 questionnaire responses for women who reported permanent cessation of menstruation due to natural causes on the 1997 questionnaire: the concordance was 88%, suggesting that misclassification of age at menopause was not substantial.
The vast majority of participants in the present study had completed high school or a higher level of education. Thus the study results are likely generalizable to the 85% of US Black women of the same ages who have at least a high school education.50 We cannot determine whether any of the factors under study, including racism, have a different relation to age at menopause among low-income Black women.
In summary, the present data from the Black Women’s Health Study indicate that the factor most strongly associated with age at menopause in Black women is smoking, in agreement with findings in White women. The results also suggest that young women who have a unilateral oophorectomy without hysterectomy are at increased risk of early menopause. The study provides equivocal data on whether experiences of racism accelerate the onset of menopause.
Acknowledgments
This work was supported by grant CA58420 from the National Cancer Institute.
We thank Drs. Camara Jones and David Williams for providing their questions on race and racism.
The study was approved by the institutional review boards of Boston University and Howard University. Participants indicated their informed consent by completing postal health questionnaires and providing written releases for medical records.
J. R. Palmer, L. Rosenberg, and L. L. Adams-Campbell designed the Black Women’s Health Study and participated in the writing of this article. L. A. Wise carried out all data analyses and participated in preparation of the article, particularly with regard to the Methods section. N. J. Horton acted as biostatistician.
Peer Reviewed
References
- 1.Cummings SR. Epidemiology of osteoporosis in women. In: Korenman SG, ed. The Menopause: Proceedings of the Serono Symposium on the Menopause, March 1989, Napa, CA. Norwell, Mass: Serono Symposium; 1990:1–15.
- 2.Kritz-Silverstein D, Barrett-Connor E. Early menopause, number of reproductive years, and bone mineral density in postmenopausal women. Am J Public Health. 1993;83:983–988. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Cooper G, Sandler DF. Age at natural menopause and mortality. Am J Epidemiol. 1998;8:229–235. [DOI] [PubMed] [Google Scholar]
- 4.Snowden DA, Kane RL, Beeson WL, et al. Is early natural menopause a biologic marker of health and aging? Am J Public Health. 1989;79:709–714. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.McKinlay SM, Brambilla DS, Posner JG. The normal menopause transition. Am J Hum Biol. 1992;4:37–46. [DOI] [PubMed] [Google Scholar]
- 6.Bromberger JT, Matthews KA, Kullerr LH, et al. Prospective study of the determinants of age at menopause. Am J Epidemiol. 1997;145:124–133. [DOI] [PubMed] [Google Scholar]
- 7.Togerson DJ, Avenell A, Russell IT, et al. Factors associated with onset of menopause in women aged 45–49. Maturitas. 1994;19:83–92. [DOI] [PubMed] [Google Scholar]
- 8.Brambilla DJ, McKinlay SM. A prospective study of factors affecting age at menopause. J Clin Epidemiol. 1989;42:1031–1039. [DOI] [PubMed] [Google Scholar]
- 9.Lindquist O, Bengtsson C, Hansson T, et al. Age at menopause and its relation to osteoporosis. Maturitas. 1979;1:175–181. [DOI] [PubMed] [Google Scholar]
- 10.Willett W, Stampfer MJ, Bain C, et al. Cigarette smoking, relative weight, and menopause. Am J Epidemiol. 1983;117:651–658. [DOI] [PubMed] [Google Scholar]
- 11.Cooper G, Sandler DP, Bohlig M. Active and passive smoking and the occurrence of natural menopause. Epidemiology. 1999;10:771–773. [PubMed] [Google Scholar]
- 12.Gold EB, Bromberger J, Crawford S, et al. Factors associated with age at natural menopause in a multiethnic sample of midlife women. Am J Epidemiol. 2001;153:865–874. [DOI] [PubMed] [Google Scholar]
- 13.Hardy R, Kuh D, Wadsworth M. Smoking, body mass index, socioeconomic status and the menopausal transition in a British national cohort. Int J Epidemiol. 2000;29:845–851. [DOI] [PubMed] [Google Scholar]
- 14.Whelan EA, Sandler DP, McConnaughey DR, et al. Menstrual and reproductive characteristics and age at natural menopause. Am J Epidemiol. 1990;131:625–632. [DOI] [PubMed] [Google Scholar]
- 15.Hardy R, Kuh D. Reproductive characteristics and the age at inception of the perimenopause in a British national cohort. Am J Epidemiol. 1999;149:612–620. [DOI] [PubMed] [Google Scholar]
- 16.Stanford JL, Harge P, Brinton LA, et al. Factors influencing the age at natural menopause. J Chronic Dis. 1987;40:995–1002. [DOI] [PubMed] [Google Scholar]
- 17.Garrido-Latorre F, Lascano-Ponce EC, LopezCarrillo L, et al. Age of natural menopause among women in Mexico City. Int J Gynaecol Obstet. 1996;53:159–166. [DOI] [PubMed] [Google Scholar]
- 18.Van Keep PA, Brand PC, Lehert P. Factors affecting the age at menopause. J Biosoc Sci. 1979;6(suppl):37–55. [DOI] [PubMed] [Google Scholar]
- 19.Sherman BM, Wallace RB, Bean JA, et al. The relationship of menopausal hot flushes to medical and reproductive experience. J Gerontol. 1981;36:306–309. [DOI] [PubMed] [Google Scholar]
- 20.Luoto R, Kaprio J, Uutela A. Age at natural menopause and sociodemographic status in Finland. Am J Epidemiol. 1994;139:64–76. [DOI] [PubMed] [Google Scholar]
- 21.Rosenberg L, Palmer JR, Rao RS, AdamsCampbell L. Risk factors for coronary heart disease in African-American women. Am J Epidemiol. 1999;150:904–909. [DOI] [PubMed] [Google Scholar]
- 22.Palmer JR, Rao RS, Adams-Campbell LL, Rosenberg L. Correlates of hysterectomy among African-American women. Am J Epidemiol. 1999;150;1309–1315. [DOI] [PubMed] [Google Scholar]
- 23.Palmer JR, Rao RS, Adams-Campbell LL, Rosenberg L. Height and breast cancer risk: results from the Black Women’s Heath Study. Cancer Causes Control. 2001;12:343–348. [DOI] [PubMed] [Google Scholar]
- 24.Adams-Campbell LL, Rosenberg L, Rao RS, Palmer JR. Strenuous physical activity and breast cancer risk in African-American women. J Natl Med Assoc. 2001;93:267–275. [PMC free article] [PubMed] [Google Scholar]
- 25.Williams DR, Yu Y, Jackson JS, Anderson NB. Racial differences in physical and mental health. Socio-economic status, stress, and discrimination. J Health Psychol. 1997;2:335–351. [DOI] [PubMed] [Google Scholar]
- 26.Cozier Y, Jones C, Palmer JR, Rao RS, Adams-Campbell LL, Rosenberg L. Perceptions and experiences of racism among African-American women [abstract]. Am J Epidemiol. 2000;151:S24. [Google Scholar]
- 27.Cox DR. Regression model and life tables (with discussion). J R Stat Soc B. 1972;34:187–220. [Google Scholar]
- 28.Allison PD. Survival Analysis Using the SAS System: a Practical Guide. Cary, NC: SAS Institute; 1995.
- 29.Cramer DW, Xu H. Predicting age at menopause. Maturitas. 1996;23;319–326. [DOI] [PubMed] [Google Scholar]
- 30.Marks NF, Shinberg DS. Socioeconomic status differences in hormone therapy. Am J Epidemiol. 1998;148:581–593. [DOI] [PubMed] [Google Scholar]
- 31.Prentice RL, Kalbfleisch JD, Peterson AV, Flournoy N, Farewell VT, Breslow NE. The analysis of failure times in the presence of competing risks. Biometrics. 1978;34:541–554. [PubMed] [Google Scholar]
- 32.McKinlay SM, Bifano NL, McKinlay JB. Smoking and age at menopause in women. Ann Intern Med. 1985;103:350–356. [DOI] [PubMed] [Google Scholar]
- 33.Ginsburg J. What determines the age at the menopause? BMJ. 1991;302:1288–1289. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Matikainen T, Perez GI, Jurisicova A, et al. Aromatic hydrocarbon receptor-driven Bax gene expression is required for premature ovarian failure caused by biohazardous environmental chemicals. Nat Genet. 2001;28:355–360. [DOI] [PubMed] [Google Scholar]
- 35.Khaw KT, Tazuke S, Barrett-Connor E. Cigarette smoking and levels of adrenal androgens in postmenopausal women. N Engl J Med. 1988;318:1705–1709. [DOI] [PubMed] [Google Scholar]
- 36.Longcope C, Johnston CC. Androgen and estrogen dynamics in pre- and post-menopausal women: a comparison between smokers and non-smokers. J Clin Endocrinol Metab. 1988;67:379–383. [DOI] [PubMed] [Google Scholar]
- 37.Cramer DW, Xu H, Harlow BL. Does “incessant” ovulation increase risk for early menopause? Am J Obstet Gynecol. 1995;172:568–573. [DOI] [PubMed] [Google Scholar]
- 38.Garner PR. The impact of obesity on reproductive function. Semin Reprod Endocrinol. 1990;8:32–43. [Google Scholar]
- 39.Harlow BL, Cramer DW, Annis KM. Association of medically treated depression and age at natural menopause. Am J Epidemiol. 1995;141:1170–1176. [DOI] [PubMed] [Google Scholar]
- 40.Berga SL. How stress can affect ovarian function. Contemp Obstet Gynecol. July 1993:87–94.
- 41.Yehuda R, Giller EL, Southwick SM, Lowy MT, Mason JW. Hypothalamic-pituitary-adrenal dysfunction in posttraumatic stress disorder. Biol Psychiatry. 1991;30:1031–1048. [DOI] [PubMed] [Google Scholar]
- 42.Liu JH. Hypothalamic amenorrhea: clinical perspectives, pathophysiology, and management. Am J Obstet Gynecol. 1990;163:1732–1736. [DOI] [PubMed] [Google Scholar]
- 43.Khoury SA, Reame NE, Kelch RP, et al. Diurnal patterns of pulsatile luteinizing hormone secretion in hypothalamic amenorrhea: reproducibility and responses to opiate blockade and an alpha 2–adrenergic agonist. J Clin Endocrinol Metab. 1987;64:755–762. [DOI] [PubMed] [Google Scholar]
- 44.Krieger N, Sidney S. Racial discrimination and blood pressure: The CARDIA study of young black and white adults. Am J Public Health. 1996;86:1370–1378. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Ren XS, Amick BC, Williams DR. Race/ethnic disparities in health: the interplay between discrimination and socioeconomic status. Ethn Dis. 1999;9:151–165. [PubMed] [Google Scholar]
- 46.Broman CL. The health consequences of racial discrimination: a study of African Americans. Ethn Dis. 1996;6:148–153. [PubMed] [Google Scholar]
- 47.Stancil TR, Hertz-Picciotto I, Schramm M, Watt-Morse M. Stress and pregnancy among AfricanAmerican women. Paediatr Perinat Epidemiol. 2000;14:127–135. [DOI] [PubMed] [Google Scholar]
- 48.Colditz GA, Stampfer MJ, Willett WC, et al. Reproducibility and validity of self-reported menopausal status in a prospective cohort study. Am J Epidemiol. 1987;126:319–325. [DOI] [PubMed] [Google Scholar]
- 49.Paganini-Hill A, Ross RK. Reliability of recall of drug usage and other health-related information. Am J Epidemiol. 1982;116:114–122. [DOI] [PubMed] [Google Scholar]
- 50.Educational Attainment in the United States. Washington, DC: US Department of Commerce; March 1998 (update). Current Population Reports, P20–543.