Article Text

PDF

Extended report
Menopause, postmenopausal hormone use and risk of incident gout
  1. A Elisabeth Hak1,2,
  2. Gary C Curhan2,
  3. Francine Grodstein2,
  4. Hyon K Choi2,3,4
  1. 1Department of Internal Medicine, Division of Clinical Immunology, Erasmus MC University Medical Centre, Rotterdam, The Netherlands
  2. 2Channing Laboratory, Department of Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, Massachusetts, USA
  3. 3Department of Medicine, Division of Rheumatology, Arthritis Research Centre of Canada, Vancouver General Hospital, University of British Colombia, Vancouver, Canada
  4. 4Section of Rheumatology and the Clinical Epidemiology Unit, Boston University of School of Medicine, Boston, Massachusetts, USA
  1. Correspondence to Dr Hyon Choi, Section of Rheumatology and the Clinical Epidemiology Unit, Boston University of School of Medicine, 650 Albany Street, Suite 200, Boston, MA 02118, USA; hchoius{at}bu.edu

Abstract

Objective To prospectively study the relation between menopause, postmenopausal hormone use and risk of gout, since female sex hormones have been postulated to decrease gout risk among women.

Methods In the Nurses' Health Study, the association between menopause, age at menopause, postmenopausal hormone use and risk of self-reported physician-diagnosed incident gout among 92 535 women without gout at baseline was examined. Multivariate proportional hazards regression analysis was used to adjust for other risk factors for gout such as age, body mass index, diuretic use, hypertension, alcohol intake and dietary factors.

Results During 16 years of follow-up (1 240 231 person-years), 1703 incident gout cases were recorded. The incidence rate of gout increased from 0.6 per 1000 person-years in women <45 years of age to 2.5 in women ≥75 years of age (p for trend <0.001). Compared with premenopausal women, postmenopausal women had a higher risk of incident gout (multivariate-adjusted relative risk (RR)=1.26; 95% confidence interval (CI) 1.03 to 1.55). Among women with a natural menopause, women with age at menopause <45 years had a RR of 1.62 (95% CI 1.12 to 2.33) of gout compared with women with age at menopause 50–54 years. Postmenopausal hormone users had a reduced risk of gout (RR=0.82; 95% CI 0.70 to 0.96).

Conclusion These prospective findings indicate that menopause increases the risk of gout, whereas postmenopausal hormone therapy modestly reduces gout risk.

View Full Text

Statistics from Altmetric.com

Despite the increased incidence of gout in recent years,1 and the substantial prevalence particularly in the older female population,2 the risk factors for gout in women remain ill-defined. Sex hormones have been postulated to be associated with the risk of gout among women. Serum urate levels, which are closely associated with gout, increase substantially with age in women, whereas among men urate concentrations are not significantly different between middle- and older age.3,,5 In two surveys, a rise in urate levels among women was detected after middle age with a subsequent plateau.3 4 Given the coinciding time periods, investigators inferred that this observation may be due to hormonal changes accompanying menopause.3 4 Another survey, however, showed increasing uric acid levels with age from the fourth till the eighth decade and over in women.5 Nonetheless, small cross-sectional studies have suggested that the vast majority of women with gout were diagnosed after menopause.6,,8 Furthermore, postmenopausal hormone therapy reduces uric acid levels among women.9 10 Our recent data from a nationally representative sample of US women indicate that menopause is independently associated with higher serum uric acid levels, whereas postmenopausal hormone use is associated with lower uric acid levels among postmenopausal women.11 Together, these reported findings suggest that menopause may increase the risk of gout, whereas postmenopausal hormone use may reduce gout risk among women. To date, no study has directly investigated these relations.

To investigate these issues, we prospectively examined the association between menopause, age at menopause and postmenopausal hormone use and incident gout in the Nurses' Health Study.

Subjects and methods

Nurses' Health Study

The Nurses' Health Study is a cohort study of 121 700 female US nurses aged 30–55 years in 1976 at study inception.12 The study was designed to prospectively examine the relationship of lifestyle factors with chronic diseases, particularly cancer and cardiovascular disease. Information about diseases, lifestyle and health practices was collected via biennial mailed questionnaires. The follow-up rate exceeded 90% for each 2-year period.

Assessment

On each questionnaire, menopausal status was determined by asking whether the participants' menstrual periods had ceased permanently and, if so, at what age and for what reason (occurring naturally or after radiation therapy or surgery). If menopause was due to surgery, the subject was asked to report the number of ovaries removed. Self-reported type of menopause and age at time of menopause in this cohort were highly accurate compared with medical records.13 From baseline onwards, all participants were asked whether they used postmenopausal hormone therapy.

Women were defined as menopausal from the self-reported time of natural menopause or bilateral oophorectomy (surgical menopause). Women with indeterminable age at menopause were women who underwent hysterectomy without bilateral oophorectomy, women with incomplete availability of data on the extent of pelvic surgery, women with permanent cessation of menses due to radiation therapy and women who did not report the age of onset of menopause. Women with indeterminable age at menopause were considered menopausal when they reached the age at which natural menopause occurred in 90% of the cohort (54 years for smokers and 56 years for non-smokers).14

Information on smoking status (never, past or current) was assessed at baseline and updated every 2 years. Body mass index was computed for each 2-year time interval using the most recent weight in kilograms divided by height in metres squared as reported at baseline. A study participant was considered to have hypertension from the time it was first reported on the questionnaire. Similarly, at baseline and every 2 years thereafter, the participants provided information on regular use of drugs, including use of diuretic agents. To assess dietary intake, a self-administered semiquantitative food frequency questionnaire (SFFQ) was used that inquired about the average use of more than 130 foods and beverages, including alcohol, during the previous year.15 16 The baseline SFFQ was completed in 1980, and updated every 4 years. The reproducibility and validity of this dietary questionnaire in this cohort have been previously reported.15 16

Starting with the 1982 questionnaire, and biennially up to 1988, participants were asked whether they had received a diagnosis of gout from a doctor, and if so, the date of first diagnosis. In 2002, the question was posed again and women were then asked to report gout diagnosis since 1996. In 2004, we inquired about gout diagnosis between 2002 and 2004. Therefore, information regarding self-reported physician-diagnosed gout and its diagnosis date was available from 1980 to 1988, and from 1996 through 2004.

Study population

After we excluded cases of prevalent gout and women with indeterminable age at menopause in 1980, 92 535 women were included in the analyses of menopause and risk of incident gout for the 1980–8 period. Since the question asking for gout occurrence from 1996 to 2002 was posed in the 2002 questionnaire, we restricted the analyses for the 1996 period onwards to women answering the 2002 questionnaire. Thus, after excluding prevalent gout cases from 1988 to 1996, we resumed our follow-up among 64 966 women for the analyses from 1996 onwards.

The analysis of age at menopause and risk of gout was restricted to women who had experienced natural menopause. Women who reported use of postmenopausal hormones were excluded from this analysis since use of postmenopausal hormones causes withdrawal bleeding and therefore makes it difficult for women to identify their exact age at menopause. As of 1980, and subsequently from 1996 onwards, 14 715 and 16 317 women were included in the analysis, respectively.

The analysis of postmenopausal hormone use and incident gout included all postmenopausal women. Women with missing data on postmenopausal hormone use or use of hormones other than oral oestrogen with or without progestin (eg, unknown type or vaginal preparations) were excluded. As of 1980, and subsequently from 1996 onwards, 26 520 and 52 976 women were included in the analysis, respectively. Participants who later reached menopause during the follow-up were included beginning at the time of their menopause. We repeated our analysis after excluding women with a history of cardiovascular disease (myocardial infarction, angina or revascularisation) or cancer (except non-melanoma skin cancer) at baseline and on subsequent questionnaires, because these conditions may have caused them to alter their use of hormones.

Statistical analysis

We computed the person-time of follow-up for each woman as the interval between the date on which the 1980 questionnaire was returned and the date of the first report of physician-diagnosed gout, death or return of the 1988 questionnaire, and subsequently as the interval between the date of return of the 1996 questionnaire and diagnosis date of gout, date of death or end of the study period, whichever came first. Person-time was allocated to menopausal status or postmenopausal hormone use status according to the 1980 questionnaire and then updated with each subsequent biennial questionnaire up to 1988, and subsequently from 1996 onwards.

Person-time for each exposure category was accumulated, and incidence rates were calculated by dividing the number of events by person-time of follow-up in each category. We used proportional hazards models to determine the independent association between menopause and postmenopausal hormone use and risk of incident gout while simultaneously adjusting for other risk factors. The Anderson–Gill data structure was used to handle time-varying covariates efficiently,17 where a new data record is created for every questionnaire cycle at which a participant is at risk, with covariates set to their values at the time that the questionnaire is returned. In these analyses, we used time-varying information from each 2-year interval to analyse the risk of gout in the next 2-year cycle from 1980 up to 1988, and subsequently from 1996 onwards. Using the SFFQ,15 16 we calculated intake of alcohol (g/day), meat, seafood, total dairy and coffee (servings/day), total calories, energy-adjusted fructose (g/day) and vitamin C (mg/day).

Risk factors for gout included in the multivariate proportional hazards models were age (continuous), body mass index (<21, 21–22, 23–24, 25–27.5, 27.6–29, 30–31, or ≥32), smoking (never, past, current), use of diuretics (thiazide or furosemide; yes or no), history of hypertension (yes or no), alcohol intake (non-drinker, ≤4.9 g/day, 5.0–9.9 g/day, 10.0–14.9 g/day, 15.0–29.9 g/day, or ≥30.0 g/day), daily average intake of meats, seafood, dairy, coffee, energy-adjusted fructose and vitamin C (each in quintiles) and total energy intake. In the analyses of postmenopausal hormone use and risk of gout among menopausal women, we also controlled for age at menopause (continuous) and type of menopause (natural, surgical, other). All analyses were performed using SAS software, version 9.1 (SAS Statistical Software).

The Partners Health Care System institutional review board approved this study. Return of a completed questionnaire was accepted by the review board as implied consent.

Results

During 16 years of follow-up (1 240 231 person-years), we documented 1703 incident self-reported physician-diagnosed gout cases. The incidence rates of gout per 1000 person-years according to age categories (<45, 45–49, 50–54, 55–59, 60–64, 65–69, 70–74 and ≥75 years) were 0.6, 1.1, 1.4, 1.5, 1.2, 1.7, 2.2 and 2.5, respectively (p for trend <0.001).

The characteristics of the participants included in the analysis on menopause and risk of gout are shown in table 1. Surgical menopausal women had the highest body mass index, reported use of diuretics and hypertension most often and tended to consume less alcohol. Intake of meat, seafood and dairy food was not consistently different between premenopausal and postmenopausal women. The consumption of meat and dairy products tended to be lower, whereas the consumption of seafood and vitamin C tended to be higher, in 1996 than in 1980.

Table 1

Age-standardised distribution of characteristics among women participating in the Nurses' Health Study by menopausal status

The risk of incident gout according to menopausal status is shown in table 2. Menopause was associated with an increased risk of gout (age-adjusted relative risk (RR)=1.33, 95% CI 1.08 to 1.63). After adjusting for other covariates including body mass index, use of diuretics, hypertension, smoking and intake of alcohol, meat, seafood, dairy, coffee, fructose, vitamin C and total energy the RR was slightly attenuated. When we examined type of menopause, we found that the risk of incident gout tended to be higher for surgical than natural menopause (table 2).

Table 2

Relative risk (RR) of self-reported physician diagnosed incident gout according to menopausal status among women participating in the Nurses' Health Study

Among women who had experienced natural menopause and had never used postmenopausal hormones, women with age at menopause <45 years had a multivariate-adjusted RR of 1.62 (95% CI 1.12 to 2.33) of incident gout compared with women with age at menopause 50–54 years (table 3).

Table 3

Relative risk (RR) of self-reported physician diagnosed incident gout according to age at natural menopause among women who never used postmenopausal hormones in the Nurses' Health Study

The age-adjusted relative risk of gout among current postmenopausal hormone users (oestrogen alone or with progestin) compared with postmenopausal women who had never used postmenopausal hormones was 0.81 (95% CI 0.70 to 0.94) (table 4). Multivariate adjustment including age at menopause, type of menopause, body mass index, use of diuretics, hypertension, smoking and intake of alcohol, meat, seafood, dairy, coffee, fructose, vitamin C and total energy did not influence this result. The results were not materially different for users of oestrogen alone (n=238, multivariate-adjusted RR=0.86; 95% CI 0.72 to 1.02) or oestrogen with progestin (n=73, multivariate-adjusted RR=0.74; 95% CI 0.57 to 0.96). We found no association between duration of postmenopausal hormone use among current users (p for trend=0.37) or interval since last postmenopausal hormone use among past users and risk of gout (p for trend=0.69). Exclusion of women with a history of cardiovascular disease or cancer did not materially affect the results.

Table 4

Relative risk (RR) of self-reported physician diagnosed incident gout according to postmenopausal hormone use among postmenopausal women in the Nurses' Health Study

Discussion

In this large prospective study of female registered nurses, we found that menopause was associated with a 26% increased risk of gout. This increase in the risk of gout tended to be more evident among women with surgical menopause and among women with younger age at natural menopause. We also found that postmenopausal hormone use was associated with an 18% lower risk of incident gout among postmenopausal women. These associations were independent of dietary and other risk factors for gout such as age, body mass index, diuretic use and hypertension. This study provides the first prospective evidence for the long-suspected hypothesis that menopause and postmenopausal hormone use affect the risk of gout among women.

A biological mechanism that has been postulated to underlie the relation between menopause, postmenopausal hormone use and the risk of gout is the effect of oestrogens and progesterone on a more efficient renal clearance of urate.18,,20 In contrast to men who attain their ‘adult levels’ of serum urate around the time of puberty, women persist with ‘juvenile levels’ of serum urate throughout most of young adulthood and do not reach their maximum urate levels until after menopause. Thus, incident gout among women has been speculated to be simply a reflection of degree and duration of female hormonal influence. We found the highest risk of gout among women experiencing surgical menopause, in whom loss of ovarian hormone production is sudden and complete,21 providing support for a causal role of female sex hormone depletion. Age of menarche and period of relative oestrogen deficiency (eg, illness, athletic training or eating disorders) might also affect the risk of incident gout in women, although these factors were not a focus of this study. In parallel with our results, among women enrolled in the Heart and Estrogen-Progestin replacement Study, treatment with postmenopausal hormones resulted in a slight lowering of serum uric acid levels compared with placebo (0.2 mg/dl at 1 year of follow-up).10 However, the risks of hormone replacement therapy, including breast cancer and stroke,22 should dampen the potential utility of this option as a means to lower serum uric acid levels and the risk of gout.

Although we observed an independent association between menopause and risk of gout, gout incidence increased before menopause and continued to increase until old age (≥75 years). This sharp increase of gout incidence with age among women corresponds with results from the UK General Practice Research Database.23 The gout incidence among women reported in this UK study using primary care physician data was comparable with the self-reported physician-diagnosed gout incidence in our population (up to 2.7 vs 2.5 per 1000 person-years in the oldest-age groups, respectively). These incidence rates of gout among elderly women have been reported to be approximately half of those among men.23

Strengths and limitations of our study deserve comment. A major strength of our study is its prospective design with biennial update of menopause, postmenopausal hormone use and other covariates, which reduces misclassification of these exposures. It has been shown, for example, that the validity of reported age at menopause, which can only be determined in retrospect, increases with decreasing number of years since menopause.24 Furthermore, the high level of education and the strong health interest of the women in our cohort favourably influenced the quality and accuracy of information reported. Participants of our study are not a random sample of US women, so our findings may not be directly generalisable to the entire US population. However, it seems unlikely that the basic biological relationships among our participants would differ from the population in general. A limitation of our study is that the end point was self-reported physician-diagnosed gout. Since these self-reports were not verified, the incidence rates presented are likely to be overestimates.25 However, any misclassification introduced by using self-reported physician-diagnosed gout is likely to be non-differential with regard to menopause and postmenopausal hormone use, and may thus only have led to an underestimate of the association between menopause, postmenopausal hormone use and risk of gout. Notably, our results with incident gout as the outcome agreed well with the recent data on the outcome of serum uric acid levels based on a nationally representative study of US women.11

In conclusion, our findings provide prospective evidence that menopause increases the risk of gout among women, whereas postmenopausal hormone therapy modestly reduces gout risk among postmenopausal women. The impact of menopause on the risk of gout may be larger among women with surgical menopause and younger age at natural menopause.

Acknowledgments

AEH is the recipient of an Erasmus MC Fellowship (Erasmus MC University Medical Centre, Rotterdam, The Netherlands) and has been supported by the Foundation ‘Vereniging Trustfonds Erasmus Universiteit Rotterdam’, The Netherlands. The authors gratefully acknowledge the participants in the NHS for their continuing cooperation and thank Elaine Coughlan-Gifford for her assistance.

References

View Abstract

Footnotes

  • Funding This research was supported by grant AR056042 from the National Institutes of Health.

  • Competing Interests None.

  • Ethics approval This study was conducted with the approval of the Partners Health Care System Institutional Review Board, Boston, USA

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Patient consent Obtained.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.