Abstract
Fear and loathing persist
As the post-Women’s Health Initiative (WHI) era limps towards the end of its second decade, a new report has found that hormone therapy continues to be underused in ‘subgroups for whom MHT [menopausal hormone therapy] is often recommended’. 1 The WHI hangover, it seems, is proving tough to shake off. The study, based on 17 years of survey data from the prospective Study of Women’s Health Across the Nation (SWAN), first found - not-surprisingly - that initiation of hormone therapy (HRT) declined from 8.6% (in the study population) pre-WHI to 2.8% post-WHI, which was accompanied by a corresponding decline in continuation (from 84% to 62%). The most common reasons cited for not starting hormones were found mainly to reflect guidance issued in the immediate aftermath of the first WHI reports from regulators and professional heart groups. But troublingly, say the authors, despite the huge overhauls of the original WHI findings since 2002, ‘younger women and those with more vasomotor symptoms’ are still being denied treatment on the back of those widely adopted original recommendations.
Indeed, these outdated but ongoing HRT scares keep cropping up. In 2017 a report from the US Preventive Services Task Force - albeit much criticised - recommended ‘against the use of combined estrogen and progestin for the primary prevention of chronic conditions in postmenopausal women’. 2 Several studies in the years after the WHI also attributed - again controversially - a decline in breast cancer incidence to the universal decline in HRT use. 3 Indeed, even in 2014 a modelling study from the Hutchinson Institute for Cancer Outcomes Research reported that the consequence of some 4.3 million women no longer taking combined HRT in the USA was 126,000 fewer breast cancer cases (and an increase in 126,000 fractures). 4
However, JoAnn Manson, one of the original WHI investigators, has recently cast doubt on many of these cause-and-effect claims emerging from the WHI data as cheerleader for the study. In an editorial for the journal Menopause and on the back of her 2017 follow-up all-cause mortality study of WHI subjects, she proposed that the only way to assess whether a population change in exposure might be causally related to a population change in disease ‘is to examine trends over a longer period of time’. 5 And that is just what was done in her landmark study of 2017. 6 This study, looking only at deaths from any cause up to 18 years after the start of HRT and up to 12 years after it was stopped, found that the risk of dying before the end of follow-up in 2014 was the same for women who had taken HRT as for those who had taken the placebo. This was true for all-cause rates and for cause-specific (including cardiovascular diseases and cancers). Notably, the hazard ratios were even lower for HRT users in the younger (50–59 years) age range. These results had already been anticipated in several individual revision studies of the WHI data. 7 , 8 Indeed, it was one of these studies which gave rise to the ‘timing hypothesis’ and the view current in most (but not all) guidelines, that HRT is safe and effective for most menopausal women with troublesome symptoms if they start treatment before the age of 60 and/or within ten years of menopause. While this guidance is quite clear in statements from NICE, the International Menopause Society, the North American Menopause Society, the British Menopause Society and many more, it seems inexplicable that so many women - especially younger symptomatic women - are denied a treatment whose safety perception remains shaped by findings from a largely discredited study of women from an age group older than their own.
Footnotes
No VTE risk from transdermal HRT in large observational study
The 2015 menopause guidelines of the UK’s National Institute for Health and Care Excellence (NICE) acknowledged that the background risk of venous thromboembolism (VTE) is increased by hormone therapy (HRT) but that this risk is ‘greater for oral than transdermal preparations’. Indeed, added NICE, the risk with transdermal HRT given at standard therapeutic doses is no greater than baseline population risk. As a result, NICE recommended transdermal HRT for women deemed at a higher risk of VTE (which included those with a BMI over 30).
NICE went no further in its recommendations, leaving to later research questions on dose and formulation - as well as the modifying effects of age and BMI. Now, however, many of these questions have been answered by a large case-control study based on UK prescription data, which included 80,000 women with a first VTE and nearly 400,000 women without. 1 Some 7% of the women had a prescription for HRT during the 90 days prior to the VTE event.
Firstly, as expected, results showed that oral HRT was indeed associated with a significantly higher risk of VTE than no treatment (OR 1.40 for estrogen only and 1.73 for combined). Moreover, estradiol had a lower risk than conjugated equine estrogen for estrogen-only and combined preparations. The highest risk for VTE in the oral preparations was conjugated equine estrogen combined with medroxyprogesterone acetate (OR 2.10), the formulation, incidentally, of the Women’s Health Initiative trial and generating the same hazard ratio for VTE. The lowest risk combination found in the case-control study was with estradiol and dydrogesterone.
Secondly, and also as expected, there was no increased risk of VTE associated with any transdermal preparation - prompting the investigators to conclude that ‘transdermal treatment appears to be underused, with the overwhelming preference still for oral preparations’. The authors thus support the recommendations of NICE - that doctors should give greater consideration to transdermal HRT for symptomatic women deemed to be at risk of VTE.
Vitamin D fails to show benefit in non-skeletal conditions
Vitamin D, the sunshine vitamin, is a hot topic right now. The latest chapter in what seems an ever-running saga began in 2016 when Public Health England proposed that everyone in the UK should consider taking 10 mcg of vitamin D supplement every day during the autumn and winter, a time of insufficient sunlight to synthesise vitamin D. They also recommended that people at risk of vitamin D deficiency - such as those living in care homes - should take supplements all year round.
However, just months later - and in stark contradiction to the PHE’s urgings - a meta-analysis of randomised trials showed that vitamin D supplementation alone does not improve musculoskeletal outcomes, with ‘ongoing uncertainty’ in other non-musculoskeletal outcomes. 1 These results echoed advice from the US Preventive Services Task Force which in 2013 had also recommended against vitamin D and calcium supplementation for fracture prevention in otherwise healthy postmenopausal women - a conclusion which largely remains in place in the latest version. 2 Importantly, the 2016 meta-analysis concluded that most people can achieve adequate levels of vitamin D from moderate exposure to sunlight and diet (oily fish, red meat, eggs . . . ).
There is little doubt, as evident in many trials and meta-analyses, that vitamin D is essential to regulate levels of calcium and thereby prevent osteoporosis - and there is also good evidence that one reason for the disturbed calcium balance which comes with ageing is inadequate vitamin D levels in the elderly. Indeed, almost all the body’s calcium is stored in the skeleton as hydroxyapatite, which provides skeletal strength and is a source of calcium for multiple calcium-mediated functions.
However, research into the beneficial role of vitamin D has now progressed far beyond the skeleton and in this helter-skelter saga has been seen in heart failure, cancer, Alzheimer’s disease, diabetes, chronic pain, and now superbugs. So there was much interest recently when supplements of vitamin D3 (along with the marine omega-3 fatty acid eicosapentaenoic acid (EPA)) were tested in two trials in the primary prevention of cancer and cardiovascular disease. 3 , 4 The studies, conjoined as the Vitamin D and Omega-3 trial (VITAL), were said to be among the largest RCTs to examine these associations in a diverse population of 26,000 men and women (with a mean age of 67 years).
VITAL’s two primary outcome measures were invasive cancer of any type and major cardiovascular events, but, after follow-up of more than five years, neither those who received vitamin D3 or the omega-3 showed any significantly lower incidence of either outcome than those on placebo. The results seemed especially disappointing for vitamin D - with no benefit even in those who appeared to be deficient at the trial’s outset. However, speaking in a media podcast, VITAL’s first investigator, the ubiquitous JoAnn Manson from the Harvard Medical School, said there was ‘a signal’ in the vitamin D’s findings for a potential benefit in reducing cancer death - though there was no significant benefit in the incidence of cancer. 5
These underwhelming results left Manson somewhat perplexed about their public health implications. But for those already taking the supplements over the counter - ‘and there are large numbers of them,’ she said, ‘we don’t find clear reasons to stop taking them.’ But for those just thinking about it, she added, ‘it’s maybe best to stay tuned for additional research’.
Still no encouragement for aspirin in the primary prevention of CVD
The aspirin story rumbles on. In my last installment of News and Views I reported the results of two trials (ASPREE and ARRIVE) which each found no benefit, especially in the elderly, from prophylactic aspirin in the primary prevention of cardiovascular disease. Indeed, there may even be harm, in the greater risk of bleeding in the brain and digestive tract.
These results have now been echoed in a new meta-analysis of trials, which included both ASPREE and ARRIVE and is described as the first systematic review for ‘five or six years’. 1 In fact, the results showed that aspirin use was associated with a lower risk of cardiovascular events (death, non-fatal myocardial infarction, non-fatal stroke, HR 0.89) than no aspirin. However, as in the recent trials it was also associated with an increased risk of major bleeding (HR 1.43). A total of 13 trials were included in the analysis, involving more than 160,000 subjects (male and female). Their median age was 62 years.
As last year, the results were widely reported in the press, with further warnings about the safety of aspirin for the primary prevention of cardiovascular disease. And again, the authors of this report - both from London - describe aspirin in primary prevention as ‘controversial’, a controversy reflected in different guidelines. However, these authors remain non-committal, recognising aspirin’s ‘modest’ cardiovascular benefit but one ‘equally balanced by major bleeding events’. This was evident in the study’s complex calculation of ‘absolute’ risk in addition to relative. This risk did appear to be modified according to the level of baseline cardiovascular risk, which explains why the authors conclude that ‘the decision to use aspirin for primary prevention may need to be made on an individual basis, accounting for the patient’s risk of bleeding and their views on the balance of risk vs benefit’.
Presently, UK and European guidelines (including NICE) tend not to recommend aspirin for primary prevention, but some of the US guidelines still recommend it based on age or cardiovascular risk, and there are a large number of patients in the US - as surely in Europe - still taking aspirin for primary prevention.
Early menopause and the effects of under-nourishment during pregnancy
Epidemiologist Professor David Barker from the University of Southampton was one of the first to recognise - as in the Barker hypothesis - a link between periconceptional conditions and disease in later life. He proposed that intrauterine growth retardation and low birth weight have a causal relationship with many of the origins of cardiovascular disease in middle age. 1 This theory of ‘fetal origins’ has been somewhat underlined by studies from the Netherlands, where evidence of disease risk was found as a consequence of the Dutch hunger winter, a five month famine in 1944–1945. Pregnancies exposed to poor nutrition during the early gestation phase have since been linked to an increased risk of coronary heart disease, glucose intolerance and accelerated ageing in the offsprings’ later life. The same risks were not observed for pregnancies exposed to famine during late gestation.
Now, in a recent study from China, this early life exposure to famine has been further implicated in the risk of reproductive ageing and early menopause. 2 The authors write that people in several communities in China were affected by famine in the years 1959 to 1961, and, between 2011 and 2012, almost 3000 of them who were pregnant during these famine years were enrolled in the study. Results showed first that the women themselves exposed to prenatal famine had a significantly higher risk of early menopause (defined as natural menopause <45 years, OR 1.59), with a non-significant trend of higher risk of premature ovarian failure (natural menopause <40 years) than unexposed women. And secondly, as an echo of the Barker hypothesis, prenatal fetal famine exposure was associated with a higher risk of premature ovarian failure (OR 2.07), with a non-significant trend of early menopause (OR 1.37), than found in those unexposed to prenatal famine.
The authors - quite rightly - say their study helps a better understanding of early menopause and premature ovarian insufficiency, which nevertheless remains poorly understood. A genetic link seems now beyond dispute, especially in Fragile X syndrome. There are also indications of an immune response, but otherwise the main but dimly recognised risk factors remain age and family history.