News and views

Severity of vasomotor symptoms linked to hip fracture risk

A prospective observational study which followed up almost 30,000 women from the Women's Health Initiative has found a correlation between the incidence of vasomotor symptoms and hip fracture. ¹ Women with moderate/severe vasomotor symptoms were found to have lower bone mineral density (BMD) over time at the femoral neck and lumbar spine than women without symptoms, and an increased rate of hip fractures, over an average 8.2 years of follow-up.

This, say the authors, is only the second time that menopausal symptoms have been studied in the context of fracture incidence, and so far the association between the two has been unknown, even though – as they unsurprisingly suggest – ‘lower estradiol levels in women with hot flashes may partly explain the associations we found’.

The subjects finally included in the analysis were 23,573 women from the WHI trials not using hormone therapy (either self-reported or because of randomisation), and an additional 4867 from a sub-set cohort whose BMD was recorded at baseline. Of these, 65% reported having no vasomotor symptoms, 24% reported mild symptoms, and 10% reported moderate/severe symptoms.

After adjustment for age, BMI, smoking and education, the hazard ratio for hip fracture among women with moderate/severe symptoms at baseline was 1.78 (vs no symptoms, 95% CI, 1.20–2.64), with symptom severity inversely associated with BMD during follow-up. However, there was no association between symptoms and vertebral fracture.

As a further explanation for the findings, the investigators suggest that balance may play a role – although the extent of symptoms did not correlate with rates of arm or wrist fracture. But, they say, ‘our analysis does suggest that impaired physical functioning may partially explain the association between moderate/severe [symptoms] and hip fracture’.

Given the strength of their study – large numbers, long follow-up, detailed and reliable baseline data – the investigators seem justified in suggesting that ‘women with vasomotor symptoms may benefit from greater attention to healthy lifestyle habits to maintain bone health’.

Another follow-up report from the WHI has this time dismissed the notion that weight gain protects against fracture in postmenopausal women. A post hoc analysis of more than 120,000 healthy women enrolled in the original trials and observational studies and followed up until 2013 found that weight gain, weight loss and intentional weight loss were all associated with increased incidence of fracture. ² Only stable weight, apparently, was risk-free.

Each year subjects were weighed and asked to report fractures of the upper and lower limb and central body. Changes in body weight were categorised as stable (a change of less than 5% from baseline), weight loss (a decrease of 5% or more) and weight gain (an increase of 5% or more since initial examination). Overall, 66% had stable weight, 15% lost weight and 19% gained weight.

Results at the third annual assessment showed that, during an average of 11 years’ follow-up, weight loss was associated with a 65% greater increase in hip fracture than stable weight (RR 1.65), a 9% increase in upper limb fracture (RR 1.09), and a 30% increase in central body fracture (RR 1.30).

Compared with stable weight, unintentional weight loss was associated with an increased risk of hip and vertebral fractures, while intentional weight loss was associated with an increased risk of lower limb fractures, but a decreased risk of hip fracture.

Six healthy habits to reduce female mid-life cardiovascular disease

The Nurses' Health Study rumbles on, almost 40 years after it was first set up to test the long-term effects of oral contraceptives. Nurses were the chosen ones for the original cohort because they would likely ‘be able to respond with a high degree of accuracy to brief, technically-worded questionnaires and motivated to participate in a long term study’. The Nurses' Health Study II – funded like the original by the US National Institutes of Health – was a 1989 extension, to study oral contraceptives, diet and lifestyle in a population now younger than the original cohort. Both cohorts included more than 100,000 nurses followed up by regular questionnaire. Then in 2010 a third Nurses' Health Study was announced as a web-based exercise ‘more representative of nurses’ diverse backgrounds’ with an aim to look closely at health outcomes related to lifestyle, fertility, pregnancy and environment.

The latest report, however, comes from the Nurses' Health Study II and concludes dramatically that the adoption of six healthy lifestyle habits could actually prevent 73% of coronary heart disease (CHD) cases in women, and 46% of diabetes, hypertension and hyperlipidaemia. ¹ The six healthy habits – deduced from a 20-year follow-up of 69,000 final subjects – were not smoking, maintaining a low BMI (defined as 18.5–24.9 kg/m²), being physically active for at least 2.5 h a week, watching no more than 7 h television per week, and eating a healthy diet. Also included was consumption of up to one alcoholic drink a day.

The cohort's original 88,940 women were 27–44 years old at entry in 1991 and had no cardiovascular disease (CVD) or diabetes, and most of them no hypertension or hypercholesterolaemia. Those who went on to develop a risk factor for CVD were on average 46.8 years old.

The analysis showed that if all women in the cohort had stuck faithfully to the six lifestyle habits, 93% of the cases of diabetes, 57.0% of cases of hypertension and 40% of the cases of hypercholesterolemia would not have occurred.

In multivariable-adjusted models, non-smoking, a normal BMI, exercise, and a healthy diet were independently and significantly associated with lower CHD risk. Compared with women with no healthy lifestyle factors, the hazard ratio for CHD in women adopting the six lifestyle habits was 0.08 (95% CI, 0.03–0.22). Approximately 73% of CHD cases were attributable to poor adherence to a healthy lifestyle. Similarly, 46% of clinical CVD risk factor cases were attributable to a poor lifestyle.

‘To competent clinicians on the front lines of primary and primordial prevention, these results will not be terribly surprising,’ said Dr Donna K Arnett with some understatement in an accompanying editorial. ² Exercise, activity, diet, weight, tobacco use and alcohol consumption ‘have long been important behavioral targets’ in campaigns to lower CVD risk on an individual and population level.

And these campaigns have had some success in some population groups. Female mortality from CVD in some (but not all) European countries and the USA has steadily declined over the past 40 years, but there remain concerns in certain groups, notably 35–44-year old women in whom CVD mortality rates are not falling. Analysts speculate that rising rates of obesity and diabetes are now offsetting reductions in smoking, hypertension, and hypercholesterolaemia.

Thus, as the accompanying editorial states, ‘if the recent trend of rising CHD in young women is indeed true and not a mere blip, then this report is both timely and … encouraging’. Of course, the means to reverse this trend – like the six healthy habits – are well known. ‘So all that remains,’ said the editorial with barely concealed resignation, ‘is the task of successfully convincing young adults not to smoke, to exercise more, and to eat and drink prudently.’

Endocrine disrupting chemicals associated with earlier menopause

An association between early age at menopause and exposure to ‘endocrine disrupting chemicals’ has been found in an analysis of the US National Health and Nutrition Examination Survey (NHANES), with data on more than 30,000 women collected between 1999 and 2008. ¹ However, in mitigation at the outset, the investigators describe as ‘challenging’ any study of the effect of occupational and environmental chemicals, since exposure levels vary by region and are much dependent on elusive demographic and socioeconomic factors.

Nevertheless, endocrine disruptors have long been putatively associated with multiple reproductive parameters – in both males and females. Exposure to such compounds as phthalates and pesticides has been linked to reduced sperm concentrations, earlier age of puberty, declines in female fertility and increased rates of pregnancy complications. Other reports have implicated endocrine disruptors in cardiovascular disease risk factors, such as hypertension.

This complex study analysed levels of endocrine disrupting chemicals in 31,000 women in the NHANES, none of whom were pregnant, breastfeeding or using hormonal contraception (and with no history of bilateral oophorectomy or hysterectomy). The final study sample of 1442 menopausal women who had had laboratory assessment, said the investigators, ‘represents a population of 8,872,966 menopausal women across the USA’. Exposures were defined by serum lipid and urine creatinine measures, and the chemicals stratified according to long (>1 year), short and unknown half-lives. The women were screened for a total of 111 chemicals, including phytoestrogens and polychlorinated biphenyls (PCBs), whose manufacture is now banned in many countries but which where once widely used in paints, electrical components and coolants.

There were 15 major culprit chemicals – nine PCBs, three pesticides and two phthalates – and complex analysis and adjustment showed that women with high levels of exposure had mean ages of menopause 1.9 to 3.8 years earlier than women with lower levels. Exposed women were also up to six times more likely to be menopausal than non-exposed women. The 15 endocrine disruptors, the investigators say, ‘warrant closer evaluation because of their persistence and potential detrimental effects on ovarian function’, and especially as an ‘earlier menopause can alter the quantity and quality of a woman’s life’, with implications in the development of cardiovascular diseases and osteoporosis.

The study also found (in 14 of the 15 chemicals) a dose–response relationship, ‘suggesting that increasing levels of environmental exposures to these chemicals could affect ovarian function’. One explanation for this effect, the authors write, is that endocrine disruptors ‘slowly damage the follicular pool, leading to earlier menopause’, or ‘may cause failure to establish a maximal oocyte pool in the case of exposure in utero, or may increase follicular recruitment, leading to premature depletion of the follicular pools as is seen in some primary ovarian insufficiency models’.

They further add that many of the 15 culprits (especially PCBs) are now banned from many countries, but ‘other related chemicals are still produced globally and are pervasive and persistent’.

These US findings came at the same time as one chemical under much scrutiny as an endocrine disruptor, bisphenol A, was given a clean bill of health by the European Food Standards Authority (in January). Although the EU and US authorities have banned the use of bisphenol A in baby bottles, it remains widely used in the manufacture of food contact materials such as plastic tableware and can coatings. However, current exposures from diet or other sources, says the EFSA, are ‘considerably under the safe level’.

Menopause through the crystal ball

As personalised medicine steps tentatively into the future, responsibility for the health, quality of life and longevity of postmenopausal women will increasingly rest on preventative therapy, predicted Roger Lobo, a former President of the American Society for Reproductive Medicine and Board member of the International Menopause Society. ¹ And an important component of that prevention, he said, if prevention therapies are needed, will be menopausal hormone therapy.

Lobo's forecast, delivered as a plenary lecture during last year's IMS congress and now presented as a paper in the Society's journal Climacteric, envisages a ‘prescription for healthy lifestyle and mental activities to prevent dementia, [which] will include specific exercises and diets depending on prediction models’. In symptomatic women, Lobo adds that the choice of hormone therapy is ‘clear and straightforward’, with younger women likely to gain additionally from a reduced risk of osteoporotic fracture.

In younger women at the onset of menopause, he notes, there are ‘compelling’ observational and trial data showing that estrogen decreases cardiovascular disease and mortality. ‘The consistency of the entire body of literature available to us, particularly with long-term data – women followed for 10 years or more – makes the argument more acceptable,’ Lobo writes. ‘It also needs to be appreciated that, apart from lifestyle changes, there is no other validated primary prevention therapy to prevent CVD in women.’

However, the key to Lobo's future consultations is individualisation: ‘Continuous ongoing assessment and the use of new diagnostic tools for efficacy and risks, and therapeutic adjustments as necessary, will ensure the best welfare of postmenopausal women.’

HRT associated with modest risk increase in a relatively rare but serious cancer

A Lancet report in February from the ‘Collaborative Group on Epidemiological Studies of Ovarian Cancer' caused more ruffled feathers among the professionals than among the public. ¹ For, despite the investigators' usual soundbites for the press (‘The increased risk may well be largely or wholly causal'), most newspapers seemed somewhat underwhelmed by the results. Indeed, the UK's Daily Mail, quoting the new chairman of the International Menopause Society, Rod Baber, reported that the study ‘at worst suggests a very small increase in risk with HRT' and that women should not stop their HRT based on these results. NHS Direct, not usually the greatest fan of popular journalism, conceded that ‘refreshingly, most of the UK media resisted running the study as a scare story’.

The study, funded by the Medical Research Council and Cancer Research UK and organised by the Cancer Epidemiological Unit in Oxford, was an analysis of individual participant data from 52 epidemiological studies and 21,488 women with ovarian cancer. Results showed that those who used hormone therapy (HRT) for just a few years were about 40% more likely to develop ovarian cancer than those who had never taken it (RR 1ċ43, 95% CI 1ċ31--1ċ56). Women who reported stopping after five years or longer remained at elevated risk for some time, but those who reported stopping after less than five years seemed to be at no increased risk.

‘For women who take HRT for 5 years from around age 50, there will be about one extra ovarian cancer for every 1000 users and one extra ovarian cancer death for every 1700 users,' explained study co-author Professor Sir Richard Peto from the Oxford group. With a concession to practicality, Peto later told the BBC: ‘If it were me and I had really bad [menopausal] symptoms, I'd worry more about those than any possible risk.'

His co-author, Professor Dame Valerie Beral, also from the Oxford group, said: ‘The definite risk of ovarian cancer even with less than 5 years of HRT is directly relevant to today's patterns of use - with most women now taking HRT for only a few years - and has implications for current efforts to revise UK and worldwide guidelines.'

Heather Currie, joint editor of this journal and Chairman Elect of the British Menopause Society, said that the study ‘does not prove causation, particularly when it is stated that the incidence of ovarian cancer decreases with time after stopping HRT'. She too, emphasising the ‘extremely small' absolute risk, said that women currently taking HRT ‘should not be concerned by this report'. Dr Currie was speaking on behalf of the BMS and the UK's Royal College of Obstetricians and Gynaecologists. Even Cancer Research UK, one of the study's sponsors, said, despite the investigators ‘causal' warnings, that ‘it is ultimately up to women to decide whether to use HRT or not'.

The Oxford report notes that current guidelines say little about ovarian cancer and HRT, no doubt influenced by results from the Women's Health Initiative (WHI), which in 2002 found no statistically significant, or indeed any, association. However, following the decline in HRT use following that first WHI report, ovarian cancer rates were found to fall in the USA, and there were a small number of observational studies suggesting a positive association.^2,3

In recognising a trend over time in HRT use and ovarian cancer incidence, the former investigators noted that this ‘strong temporal association, although not proving a causal role of hormones in ovarian carcinogenesis, suggests that future analytic research supporting cancer control efforts should clarify the role of hormonal exposures on the development and behavior of subtypes of ovarian cancer'. ² And this, presumably, is what the Oxford investigators set out to achieve.

Vasomotor symptoms typically last seven years: is a short-term approach appropriate?

An ongoing observational study has found that menopausal vasomotor symptoms typically lasted for more than seven years for more than half of the subjects, and persisted for 4.5 years after the final menstrual period. ¹ Health care professionals, say the authors, ‘should counsel women to expect that frequent [symptoms] could last more than 7 years, and they may last longer for African American women'.

The advice follows a report from the Study of Women's Health Across the Nation (SWAN), a multiracial/multiethnic observational study of the menopausal transition among 3302 women enrolled at seven US sites who completed a median of 13 visits between 1996 and 2013. The analysis included 1449 women with frequent vasomotor symptoms.

Women who were premenopausal or early perimenopausal when they first reported frequent symptoms had the longest total duration (median greater than 11.8 years) and persistence. Women who were postmenopausal at the onset of symptoms had the shortest total duration after a final menstrual period (median of 3.4 years).

Compared with women of other racial/ethnic groups, African American women reported the longest total symptom duration (median of 10.1 years) and Japanese and Chinese women had the shortest.

An accompanying editorial described the results as ‘important information', and asked if ‘women's health or quality of life and can be readily addressed by short-term approaches'.