Menopause and HIV infection: age at onset and associated factors,ANRS CO3 Aquitaine cohort

Abstract

The aim of the paper is to describe the characteristics of postmenopausal HIV-infected women and to investigate the factors associated with an earlier onset of menopause in a hospital-based cohort. Information was collected using a self-administered questionnaire. A Cox model was used to determine factors associated with menopause. Among the 404 women who completed the questionnaire, 69 were naturally postmenopausal at the time of the study (median age at onset: 49 years, premature menopause <40 years: 12%). The onset of menopause was studied among the 41 women still menstruating at the enrolment in the cohort, and who experienced menopause during follow-up. African origin (hazard ratio [HR] = 8.16; 95% confidence interval [CI] = 2.23–29.89) and history of injecting drug use (IDU) (HR = 2.46; 95% CI = 1.03–5.85) were associated with an increased risk of earlier menopause. Women with a CD4 cell count <200 cells/mm³ tended to reach menopause earlier (HR = 2.25; 95% CI = 0.94–5.39). Earlier occurrence of menopause seems to be associated with factors already reported in HIV-negative women (IDU, ethnicity) and with HIV-related immunodeficiency.

Keywords

HIV infection menopause women

INTRODUCTION

In France, as in other industrialized countries, women represent approximately 30% of the patients infected with HIV.^1,2 Since 1996, the use of combination antiretroviral therapy (cART) for the care of HIV-infected patients has improved substantially the life-expectancy of this population.³ Thus, the median age of patients increased dramatically during the last decades: in France, around one out of four of those living with HIV is aged over 50.⁴ In the USA, according to the Centers for Disease Control and Prevention (CDC), the estimated proportion of HIV-infected patients aged >50 years increased from 5.4% in 2001 to 28.3% in 2007.^5,6 Thus, the proportion of women reaching the age of menopause is increasing over time and such women are at higher risk of antiretroviral therapy (ART)-associated cardiovascular disease and bone loss because of estrogen deficiency.^7–11

Limited knowledge is available regarding menopause in HIV-infected women. Among the different studies conducted so far, few have investigated the onset of menopause and its associated factors.^12–16

The objective of this study was to describe the characteristics of the postmenopausal women within a cohort of HIV-infected patients and to investigate the factors associated with an earlier onset of menopause in this population.

METHODS

Study population

The study was conducted among women from the ANRS CO3 Aquitaine cohort, a prospective open cohort of HIV-infected individuals.¹⁷ As menopausal status is not routinely collected in this cohort, a self-questionnaire was administered to all women attending outpatient clinics during the study period. This questionnaire allowed us to collect data on age and type of menopause and to select women who were still menstruating at the time of enrolment in the cohort. All women from the Aquitaine cohort, seen at least once between April 2007 and February 2008, and who accepted the principles of this study received the self-administered questionnaire. Women with reading difficulties or who could not understand the questions were assisted by medical staff.

To describe the characteristics of the postmenopausal women, we studied all women reporting natural menopause at the date of the administration of the questionnaire.

To investigate the factors associated with an earlier onset of menopause, we only considered the subgroup of women who were premenopausal (still menstruating) at enrolment in the cohort and who reached menopause during follow-up.

Study variables

Menopause was defined as 12 consecutive months without menstruation, in accordance with the World Health Organization (WHO) definition.¹⁸ The date of the onset of menopause was determined at the end of these 12 months of amenorrhoea. Menopause was considered as ‘natural’ if it had occurred spontaneously, and ‘secondary’ if it was related to surgery (oophorectomy, hysterectomy) or to anti-cancer chemotherapy. Women with secondary menopause were excluded from the analysis. Patients reporting a 12-month amenorrhoea that could be explained by external causes (oral contraceptive, progesterone-containing intrauterine system, menstrual cycle disorders or pregnancy) were not considered as postmenopausal. Premenopausal women included patients who reported any menstruation during the last 12 months.

The menopausal status (premenopause, natural menopause or secondary menopause), the age and type of menopause were determined based on the responses provided by patients. Premature menopause was defined as menopause occurring before the age of 40, according to the WHO definition.¹⁸

The characteristics studied in all naturally postmenopausal women were the age at menopause, HIV transmission group, geographic origin and education level.

The following variables were used to determine factors associated with an earlier onset of menopause in the subgroup of women still menstruating at enrolment: geographic origin, tobacco consumption, history of injecting drug use (IDU) history of ART, HIV plasma viral load (VL) and CD4 count, time since HIV diagnosis, clinical stage of the disease, co-infection with hepatitis C, education level, age at first menstruation and year of inclusion in the cohort.

Statistical analysis

The median age at menopause was estimated according to self-reported onset of menopause in all women.

The probability of being postmenopausal at age 40, 45 and 50 years was estimated using the Kaplan–Meier method among women who had not experienced menopause at the time of enrolment in the cohort. Point estimates were reported with their 95% confidence interval (CI). A Cox proportional hazards model was used to study the factors associated with the onset of menopause among the cohort of women still menstruating at baseline. The point date was the date of administration of the questionnaire. The event of interest was the onset of natural menopause and data were then censored for women reporting a surgical menopause before the point date.

The variables collected at the time of enrolment in the Aquitaine cohort (geographic origin, IDU, CD4 count, time since HIV diagnosis) or those considered as intrinsic characteristics of the patients (education level, age at first menstruation) were used as fixed-effect covariates.

We also considered time-dependent variables taking into account the change in exposure of patients during the follow-up. Tobacco consumption and stage of the disease were used as time-dependent covariates with a unique point change. Patients reporting tobacco consumption or classified at the AIDS stage at enrolment or during follow-up were considered as ‘exposed’ since that date and until the end of follow-up. Finally, ART, plasma VL measured at each follow-up visit and the lowest CD4 count since enrolment in the cohort (CD4 count nadir) were considered as time-dependent covariates allowing multiple changes, and were collected repeatedly during follow-up in order to assess the existence of an association between exposure to these variables and an earlier onset of menopause.

Variables associated with the onset of menopause (P ≤ 0.25) in the univariate analysis were introduced in a multivariate model. The final model was also adjusted on the period of enrolment in the Aquitaine cohort (<1996, 1996–2001, >2001) according to critical years in the improvement of medical care of HIV-infected patients in France. For each of the explanatory variables, the hypothesis of the proportionality of risks was assessed by testing the interaction between time and the variable considered. The analysis was conducted using the SAS 9.1 software (SAS Institute Inc, Cary, NC, USA).

RESULTS

Study population

Between 15 April 2007 and 29 February 2008, among the 884 women attending one of the participating study wards, 441 were offered and accepted the self-administered questionnaire. Among these 441 patients, six were included by mistake as their informed consent for enrolment in the Aquitaine cohort had not yet been signed, 10 were misidentified and thus their questionnaire could not be analysed, and 21 were excluded from analysis due to incomplete data. Finally, data from 404 women were analysed (Figure 1).

Figure 1

Menopause study profile, ANRS CO3 Aquitaine Cohort, April 2007–February 2008. *Missing data: menopausal status: n = 7; type of menopause: n = 4; and age at menopause: n = 10. Misidentification: n = 10

Representativeness of the study sample

Compared with the 480 women attending the study sites during the same period and not included in the study, the 404 women included were less frequently of African origin (14.6% versus 21.3%; P = 0.011) and more often treated with ART (83.2% versus 75.6%; P = 0.006). Conversely, no significant difference was observed between the two groups of patients regarding the stage of HIV disease (16.3% at the AIDS stage among the participants versus 19% among non-participants; P = 0.31), the transmission group (IDU among 20.3% of participants versus 19.0% of non-participants; P = 0.62) and the age at the time of the survey (43 years [interquartile range {IQR} = 37–47 years] versus 42 years [IQR 37–48 years]; P = 0.73). At this time, the median CD4 count of the participants (516 cells/mm³ [IQR = 367–676] did not differ from that of the non-participants (473 cells/mm³ [IQR = 336–679]; P = 0.08). The proportion of undetectable plasma VL (<50 copies/mL) among participants (80%) was not significantly different from that observed among non-participants (78%; P = 0.36).

Characteristics of naturally postmenopausal women

Among the 404 women aged 19–79 years who responded to the self-administered questionnaire, 98 (24.3%) met the postmenopausal criteria: 29 of them reported secondary menopause and were excluded from the analysis (Figure 1). Among the 69 women reporting a natural menopause, the median age at menopause was 49 years (IQR 40–50); 12% of them had reached menopause before the age of 40 (premature menopause), 22% between the ages of 40–45, 28% between 46–50, and 38% after the age of 50. The vast majority of naturally postmenopausal women were of Caucasian origin (89.9%) and 10.1% were of African origin. Only 43.3% of those women interviewed had received secondary or higher education. The main routes of HIV transmission were heterosexual contact (69.6% of the naturally postmenopausal women), IDU (21.8%); 4.3% had been infected after a contaminated blood transfusion. The HIV transmission route remained undetermined for 4.3% of the patients.

Factors associated with an earlier onset of menopause

Among the 404 participants, data from 362 women still menstruating at baseline and followed-up until the date of the survey or the onset of menopause were used for this analysis (Figure 1). These women had a median age of 31 years (IQR 26–37) at their enrolment in the cohort, and were followed-up for a median 8.8 years (IQR 3.3–13.9). Among them, 41 reported a natural menopause during follow-up.

The probability of reaching menopause was 4.5% (95% CI = 2.5–8.0) at the age of 40, 11.3% (95% CI = 7.2–17.5) at the age of 45 and 38.9% (95% CI = 26.3–54.9) at the age of 50. After adjustment for the other co-variates (hepatitis C virus co-infection not taken into account because of 45% of missing data), women of African ethnicity and those with a history of IDU had a higher risk of earlier onset of menopause (hazard ratio [HR] = 8.16; 95% CI = 2.23–29.89 for the geographic origin and HR = 2.46; 95% CI = 1.03–5.85 for IDU). Immunosuppression characterized by a CD4 lymphocyte count <200 cells/mm³ at the time of enrolment tended to be associated with an earlier onset of menopause (HR = 2.25; 95% CI = 0.94–5.39; P = 0.069) compared with women with a CD4 count ≥350 cells/mm³ (Table 1). Tobacco consumption, ART use, a lower CD4 count nadir, a high VL measured during follow-up, the transition to AIDS stage and the enrolment period in the cohort were not identified as factors associated with the onset of menopause.

Table 1

Factors associated with an earlier onset of natural menopause among women in the ANRS CO3 Aquitaine cohort (n = 362)

	Total*	Postmenopausal patients	Univariate analysis		Adjusted analysis^†
	n	n = 41	HR	P	HR	95% CI	P
Geographic origin
Caucasian	306	36	1.00	0.052	1.00	2.23–29.89	0.001
African	56	5	2.91	0.052	8.16	2.23–29.89	0.001
Tobacco consumption ^‡
No	120	16	1.00	0.098	1.00	0.88–3.94	0.107
Yes	231	25	1.78	0.098	1.86	0.88–3.94	0.107
Intravenous drug use at enrolment
Never	281	27	1.00	0.053	1.00	1.03–5.85	0.042
Current or previous	81	14	1.96	0.053	2.46	1.03–5.85	0.042
ART treatment ^‡
Yes	333	34	1.00	0.903	–	–	NI
No	29^§	7	0.95	0.903	–	–	NI
CD4 level at enrolment in the cohort (cells/mm³)**
>350	220	24	1.00	0.041	1.00
200–349	73	6	1.10		1.11	0.43–2.87	0.932
<200	62	9	1.76		2.25	0.94–5.39	0.069
Not documented	7	2	8.19		–	–	0.069
CD4 count nadir (cells/mm³)^‡††
>350	78	11	1.00	0.370	–	–	NI
200–349	132	9	0.55
<200	150	20	0.89
Undetectable viral load ^‡
Yes	321	25	1.00	0.870	–	–	NI
No	32	11	1.06	0.870	–	–	NI
Time since HIV diagnosis
≥1 year before enrolment	160	14	1.00	0.245	1.00
<1 year before enrolment	202	27	1.48	0.245	2.11	0.92–4.86	0.079
AIDS stage ^‡
No	302	34	1.00	0.383	–	–	NI
Yes	60	7	1.46	0.383	–	–	NI
Education level
≥High school	201	22	1.00	0.896	–	–	NI
Primary or secondary	150	19	1.04	0.896	–	–	NI
Age at first menstruation (years)
<13	128	12	1.00	0.888	–	–	NI
≥13	225	28	1.05	0.888	–	–	NI

HR = hazard ratio; CI = confidence interval; ART = antiretroviral therapy; NI = not included in the multivariate mode

*Missing data: tobacco consumption: n = 11; viral load: n = 9; education level: n = 11; age at first menstruation: n = 9

^†Multivariate model on 344 women without missing data, adjusted on the enrolment period in the cohort (<1996, 1996–2001 and >2001). HR = 95%, CI = 95%

^‡Time-dependent variable measured during follow-up

^§Twenty-nine patients did not receive ART during follow-up

**Or first value measured during the three months following enrolment if data missing at enrolment

^††Missing data = 2 (CD4 count not documented before the onset of the event)

DISCUSSION

Improvements in the care of HIV-infected patients have led to more HIV-infected women facing menopause and its complications in France, as in all industrialized countries.

In HIV-negative women, menopause seems to occur around the ages of 50–52 years whether in the USA,^19,20 France²¹ or other European countries.^22,23 According to several studies conducted among HIV-infected women, menopause could start earlier in this population, around the ages of 46–47 years.^13,16,24 This difference could be partly explained by the existence of several predicting factors of an early onset of menopause that would be more common among HIV-infected women.^14,25 Tobacco use, drug addiction or a low socioeconomic status often associated in the literature with an earlier menopause are also classical risk factors observed among HIV-infected women.^{16,19,20,26,27}

In our study, the median age at onset of menopause was 49 years (IQR = 40–50 years), which seems not much different than what has been reported in non-infected women. However 12% of the naturally postmenopausal women reached menopause before the age of 40, which corresponds to ‘premature menopause’ according to the WHO definition.¹⁸ The prevalence of premature menopause among these HIV-infected women thus seems to be higher than among the general population: 6.3% of the 1994 postmenopausal women of the American Study of Women's Health Across the Nation (SWAN) cohort and 1.8% of the 15,253 women in an Italian study.^28,29 Other studies conducted among HIV-infected women also showed high rates of premature menopause in spite of small study samples: 23.3% in the Fantry American study published in 2005 (7/30 women) and 26% among the women from the Menopause Study (16/62 women).^15,16 Such data are not available from the French reference study conducted in 2005 among the general population, but among 624 postmenopausal women, only 7% documented an age at last menstruation before 45 years, which suggested an even lower premature menopause rate.²¹

Our study showed the existence of some factors influencing the age at onset of menopause. Drug users were 2.5 times more likely than other women to begin their menopause. Consistently, two US studies conducted in 2005 among HIV-infected women showed a significant association between the use of drugs and an increased risk of earlier onset of menopause.^15,16 We also found a much younger age at onset of menopause among women of African origin. Geographic origin is not one of the factors classically investigated in the literature. However, in 1997, a study conducted in the USA among non-infected women found a high risk of earlier onset of menopause (HR = 4.34; CI = 2.42–7.79) among Afro-American patients, after adjusting for factors such as parity, tobacco use or socioeconomic level.¹⁹

In our study, a severe level of immunosuppression at the time of enrolment in the Aquitaine cohort (CD4 < 200 cells/mm³) tended to be associated with an earlier onset of menopause (HR = 2.25; 95% CI = 0.94–5.39). Although not collected at the onset of menopause, this factor could be indicative of a more advanced stage of HIV infection at enrolment, related to a late access to care of some women. This association between low CD4 count and earlier menopause has already been observed within the cohort of HIV-infected women of the American Menopause Study in 2006.¹⁶

Several studies conducted among HIV-negative women^20,22,23,26 have suggested that tobacco consumption was associated with an earlier onset of menopause. Among the women participating in the Aquitaine cohort, as well as within several studies conducted among HIV-infected women, the role of tobacco was not significantly established.^12,16 The high prevalence of smoking in our patients and within the Aquitaine cohort overall, and its frequent association with IDU, could explain the lack of significant association of this variable with the age at onset of menopause.³⁰ Similarly, low levels of education and early age at first menstruation were not statistically associated with earlier menopause in our cohort, conversely to other studies conducted among non-infected women.^19,20,23,31

In our study, the main variables of interest, including the age at natural menopause, were documented retrospectively using a self-administered questionnaire. This tool was elaborated in order to be easily and rapidly completed by the patients during outpatient contacts. However, this data collection strategy may have affected the quality of the responses, and potentially led to recall bias, in particular if the event was old. When completing the questionnaire, women had to self-evaluate their menopausal stage guided by specific instructions, i.e. menopause was defined as ‘12 months without menstruation’. A French study showed a good concordance between the information (menopausal status and type of menopause) collected directly among the patients and those documented within the medical record.³² However, it is possible that this self-evaluation method used in our study may have induced errors in the classification of our patients within the different groups of ‘menopause’, ‘natural menopause’ and ‘secondary menopause’.

Although participation in the study was systematically proposed to all women attending outpatient clinics, only 50% of them accepted the self-administered questionnaire; this may have led to participation bias. On the one hand, the prevalence of menopause may have been overestimated because women having accepted to participate may have been those most concerned by the topic; on the other hand, the under-representation of African patients in our study may have underestimated the prevalence of menopause. Nevertheless, the comparison between the women enrolled in the study (n = 404) and the remaining female patients seen over the same period in the participating wards (n = 480) seems to indicate that our study sample is quite representative of all women of the cohort.

We used the definition of natural menopause proposed by WHO.¹⁸ Cases of menopause based on the responses to the self-administered questionnaire were not confirmed biologically. Even if the diagnosis of menopause is clinical above all, measurement of follicle-stimulating hormone (FSH) could have fully ascertained the diagnosis, specifically within this population of HIV-infected women. Among the HIV-infected women of the Women's Interagency HIV Study, 50% of the 120 patients reporting an absence of menstruation for at least 12 months were not really postmenopausal after testing their FSH levels.¹² We could not propose this validation test within our study. Several studies have shown a high prevalence of prolonged amenorrhoea among HIV-infected women: indeed, women with hypothalamic amenorrhoea due to HIV may have a return of menses, especially with ART.^12,33–35 In our study, where menopause was evaluated at a single time point, this could have lead to a slight overestimation of the prevalence.

To date, very few studies have investigated similarly the factors associated with earlier menopause among HIV-infected women. In order to better understand the link between HIV infection and menopause, complementary studies are warranted, which would require the inclusion of a larger number of women and a prospective gynaecological follow-up including hormonal estimations. For HIV-infected women already in care, it seems prudent to provide specific biological and gynaecological follow-up for those with risk factors for early onset of menopause, and to implement specific screening of bone and cardiovascular disorders in postmenopausal women, especially those with premature menopause.

References

Cazein

, Pillonel

, Le Strat

, Surveillance de l'infection à VIH-sida en France, 2007. Bulletin Epidémiologique Hebdomadaire 2008;45–46:434–43

UnAIDS. AIDS epidemic update 2009. See http://www.unaids.org/en/KnowledgeCentre/HIVData/EpiUpdate/EpiUpdArchive/2009 (last checked 4 October 2010)

Antiretroviral Therapy Cohort Collaboration. Life expectancy of individuals on combination antiretroviral therapy in high-income countries: a collaborative analysis of 14 cohort studies. Lancet 2008;372:293–9

Costagliola

. Epidemiology of HIV infection in France and Europe [French]. Rev Med Interne 2008;29 (Suppl 3):S266–8

US Centers for Disease Control and Prevention. Cases of HIV infection and AIDS in the United States. HIV/AIDS Surveillance Report. 2002:14. See http://www.cdc.gov/hiv/surveillance/resources/reports/2002report/ (last checked 4 October 2010)

US Centers for Disease Control and Prevention. Cases of HIV infection and AIDS in the United States and Dependent areas. HIV/AIDS Surveillance Report. 2009:19. See http://www.cdc.gov/hiv/surveillance/resources/reports/2007report/ (last checked 4 October 2010)

Gordon

, Kannel

, Hjortland

, McNamara

. Menopause and coronary heart disease. The Framingham Study. Ann Intern Med 1978;89:157–61

, Grodstein

, Hennekens

, Age at natural menopause and risk of cardiovascular disease. Arch Intern Med 1999;159:1061–6

Kritz-Silverstein

, Barrett-Connor

. Early menopause, number of reproductive years, and bone mineral density in postmenopausal women. Am J Public Health 1993;83:983–8

10.

Brown

, Qaqish

. Antiretroviral therapy and the prevalence of osteopenia and osteoporosis: a meta-analytic review. AIDS 2006;20:2165–74

11.

Friis-Moller

, Sabin

, Weber

, Combination antiretroviral therapy and the risk of myocardial infarction. N Engl J Med 2003;349:1993–2003

12.

Cejtin

, Kalinowski

, Bacchetti

, Effects of human immunodeficiency virus on protracted amenorrhea and ovarian dysfunction. Obstet Gynecol 2006;108:1423–31

13.

Clark

, Cohn

, Jarek

, Perimenopausal symptomatology among HIV-infected women at least 40 years of age. J Acquir Immune Defic Syndr 2000;23:99–100

14.

Conde

, Pinto-Neto

, Costa-Paiva

. Age at menopause of HIV-infected women: a review. Gynecol Endocrinol 2008;24:84–6

15.

Fantry

, Zhan

, Taylor

, Sill

, Flaws

. Age of menopause and menopausal symptoms in HIV-infected women. AIDS Patient Care STDS 2005;19:703–11

16.

Schoenbaum

, Hartel

, Lo

, HIV infection, drug use, and onset of natural menopause. Clin Infect Dis 2005;41:1517–24

17.

Binquet

, Chene

, Jacqmin-Gadda

, Modeling changes in CD4-positive T-lymphocyte counts after the start of highly active antiretroviral therapy and the relation with risk of opportunistic infections: the Aquitaine Cohort, 1996–1997. Am J Epidemiol 2001;153:386–93

18.

Utian

. Menopause-related definitions. Advances in Fertility and Reproductive Medicine, 18th World Congress on Fertility and Sterility, April International Congress Series, 2004:133–8

19.

Bromberger

, Matthews

, Kuller

, Wing

, Meilahn

, Plantinga

. Prospective study of the determinants of age at menopause. Am J Epidemiol 1997;145:124–33

20.

Gold

, Bromberger

, Crawford

, Factors associated with age at natural menopause in a multiethnic sample of midlife women. Am J Epidemiol 2001;153:865–74

21.

Rozenbaum

. Ménopause et règles en 2001: Résultats de l'enquête CLIMATS. Reprod Hum Hormones 2002;15:11–22

22.

Kaczmarek

. The timing of natural menopause in Poland and associated factors. Maturitas 2007;57:139–53

23.

Parazzini

. Determinants of age at menopause in women attending menopause clinics in Italy. Maturitas 2007;56:280–7

24.

Ferreira

, Pinto-Neto

, Conde

, Costa-Paiva

, Morais

, Magalhaes

. Menopause symptoms in women infected with HIV: prevalence and associated factors. Gynecol Endocrinol 2007;23:198–205

25.

Kojic

, Wang

, Cu-Uvin

. HIV and menopause: a review. J Women's Health 2007;16:1402–11

26.

Torgerson

, Avenell

, Russell

, Reid

. Factors associated with onset of menopause in women aged 45–49. Maturitas 1994;19:83–92

27.

, Treloar

, Pandeya

, Predictive factors of age at menopause in a large Australian twin study. Hum Biol 1998;70:1073–91

28.

Progetto Menopausa Italia Study Group. Premature ovarian failure: frequency and risk factors among women attending a network of menopause clinics in Italy. BJOG 2003;110:59–63

29.

Luborsky

, Meyer

, Sowers

, Gold

, Santoro

. Premature menopause in a multi-ethnic population study of the menopause transition. Hum Reprod 2003;18:199–206

30.

Benard

, Tessier

, Rambeloarisoa

, HIV infection and tobacco smoking behaviour: prospects for prevention? ANRS CO3 Aquitaine Cohort, 2002. Int J Tuberc Lung Dis 2006;10:378–83

31.

Chang

, Kim

, Lee

, Premenopausal factors influencing premature ovarian failure and early menopause. Maturitas 2007;58:19–30

32.

Clavel-Chapelon

, Dormoy-Mortier

. A validation study on status and age of natural menopause reported in the E3N cohort. Maturitas 1998;29:99–103

33.

Chirgwin

, Feldman

, Muneyyirci-Delale

, Landesman

, Minkoff

. Menstrual function in human immunodeficiency virus-infected women without acquired immunodeficiency syndrome. J Acquir Immune Defic Syndr Hum Retrovirol 1996;12:489–94

34.

Clark

, Mulligan

, Stamenovic

, Frequency of anovulation and early menopause among women enrolled in selected adult AIDS clinical trials group studies. J Infect Dis 2001;184:1325–7

35.

Santoro

, Fan

, Maslow

, Schoenbaum

. Women and HIV infection: the makings of a midlife crisis. Maturitas 2009;64:160–4