EXPERT REACTION: HRT menopause treatment may be linked to a higher risk of dementia

This is an interesting addition to ongoing research around menopause and its treatment.

Unfortunately, it raises more questions than it answers and caution needs to be applied in interpreting its results.

The major conclusion is that women taking menopausal hormone therapy, be that, combined oestrogen and progesterone, oestrogen only, or progesterone only, was associated with an extra four cases per 1000 women of dementia being diagnosed, compared to women who did not take hormones.

The particular hormones used in this population study, and their delivery systems, are not necessarily the same hormones and delivery systems we use in Australia.

The Danish women studied took mostly oral preparations and used synthetic progestogens whereas Australian women are more likely to use native progesterone and the transdermal route for administration.

There were also differences in the women who had dementia diagnosed when it came to education level, household income, living alone, high blood pressure, diabetes and thyroid disease which can conceivably be a cause of bias.

These results confirm the need for ongoing research into how best to manage menopausal symptoms but should not be used as a reason to stop or change treatments we offer in Australia.

It may be a timely reminder for menopausal women in Australia to schedule an appointment with their GP or gynaecologist to discuss menopausal symptoms and how best to treat them in their individual case.

Around 14% of Australian women take menopausal hormone therapy (MHT) for the treatment of troublesome menopausal symptoms, and most will take combined (estrogen and progestogen) preparations. In the past, MHT was recommended for the prevention of chronic diseases including dementia.

This nationwide study from Denmark provides important new information showing an increased risk of all cause dementia with combined MHT. This was observed with both continuous combined and sequential MHT preparations and risk increased with longer duration of use. These findings are similar to those reported in a large randomised controlled trials of older postmenopausal women taking MHT.

However, in this new study even women using MHT in the years immediately after menopause had an increased dementia risk. MHT is an effective treatment for hot flushes and night sweats but is not indicated for other reasons.

Strengths of this study include the large sample size and capture of all prescriptions for MHT from 1995. Limitations include the exact type of dementia and that women taking estrogen-only MHT (following hysterectomy) were not included.

This is an observational study so cannot prove that MHT causes dementia. However, the findings are concerning and reinforce recent guidance from the United States Preventive Task Force (2022) that HRT should not be used for the prevention of chronic conditions including dementia.

This paper reports on a large observational study of national registry data from Denmark that examined the link between HRT and incident dementia. Focusing on combined oral estrogen and progesterone, the authors found an increased risk of dementia in women who had ever taken HRT. When this was broken down by duration of exposure, there was an increase risk of all cause dementia of about 20%  when taking HRT for one or more years which appeared to increase to around 40% risk when taking HRT for eight to 10 years.

Previous research on this association has reported mixed findings (both increased and decreased risk of dementia associated with HRT).

There are some important factors to consider when interpreting the results of this study.  The sample included women who were aged 50-60 in the year 2000. Therefore the types of HRT that the study are based on were those prescribed in Denmark during this period and there may be factors associated with the patterns of prescribing at that time which do not apply currently. A strength was that the database captured nearly the entire population.

The study only examined orally administered HRT and not dermal so results cannot be generalised to dermal administered HRT. Overall, while the study findings are of significant concern, there are some caveats that need to be considered.

The follow-up interval meant that the study identified dementia cases occurring with a median age of 70 which is far younger than would usually be seen if the cohort had been followed up for longer. The usual median age for dementia diagnosis is in the 80s. The results may indicate earlier onset of dementia in women who took HRT in Denmark. The authors also report low rates of Alzheimer’s disease of 26% whereas in the population one would expect 60-70% of dementia cases to be diagnosed with Alzheimer’s disease if the sample were follow-up up into the 80s. This may reflect the way diagnoses were recorded in the registry. The authors note that the findings may also be explained by indication bias. This means the symptoms that lead to women taking HRT may also be linked to increased risk of dementia. The overall differences in the expected rates of dementia suggest that there are some significant biases in this study which may be due to the methodology. Further work is needed to clarify this important issue.

The observational study of Pourhardi et al published in the BMJ reports the associations between the use of combined menopausal hormone therapy (MHT: estrogen with a synthetic progestin) and the likelihood of subsequent dementia.

The average use of MHT was just under four years, virtually all use was oral therapy (90%) and the majority took a formulation of oral estrogen with oral norethisterone (83.5% of those who developed dementia and 80.6% of the control group). So, the findings are fundamentally driven by this specific combination of therapy that included a specific synthetic progestin therapy.

The authors say they excluded women who had had ever had a hysterectomy. Women who have had a hysterectomy only require estrogen therapy without a progestin. They say that in their analysis 'Compared with never users of menopausal oestrogen-progestin therapy, systemic or vaginal oestrogen only treatment, and perimenopausal progestin only therapy, ever users of menopausal oestrogen-progestin treatment were more likely to develop all cause dementia…'. So this gives rise to the question as to whether estrogen-only users were also included in the comparator group as stated, and if so, what proportion of the comparator group were using estrogen-only therapy is unclear.

Around the year 2000 in Denmark estrogen-only therapy comprised about 38% of MHT use (excluding vaginal estrogen only therapy). So why did the researchers not report on the association between estrogen only therapy and dementia risk? This might indicate whether the observed risk might be attributable to the progestin component when prescribed with estrogen.

As for all observational studies, despite statistical adjustments for variables such as weight or education etc, women who choose to take MHT are different from women who do not. The main difference is symptoms. So, while multiple adjustments have been made in the analysis for dementia risk factors such as living alone and education, no consideration was given to the reasons why women take MHT. 

*    Women who take MHT usually do so for vasomotor symptoms (hot flushes and sweats), sleep disturbance and/or mood symptoms. 
*    Hot flushes are associated with reduced blood vessel function and brain-specific blood flow, and poor sleep and low mood are both established risk factors for dementia. 

SO, the real elephant in the room for this analysis is: 
is the observed risk of dementia in MHT users is due to the use of oral synthetic MHT or due to the reasons why women ever took MHT? This has been over-looked.

As the study included women aged 50-60 years in 2000, just before the publication of the Women’s Health Initiative (WHI) Study that raised the concern of MHT use and breast cancer - specifically estrogen plus progestin use and breast cancer, many of the women in this study are likely to have stopped MHT because of this perceived risk, but still had symptoms. 

The conclusion that even short term use increased dementia risk may be because women either stopped MHT for fear of breast cancer after the WHI data was published, or were told by their doctor to stop, and had untreated ongoing symptoms of flushes, low mood and poor sleep that increased their subsequent risk of dementia.

Longer term users would have most likely continued to use MHT, despite warning of breast cancer risk because of the severity of their symptoms. 

So before drawing the conclusion that oral oestrogen with norethisterone causes a greater risk of dementia, one must consider whether MHT use is simply a surrogate marker for the postmenopausal symptoms that might directly contribute to dementia - hot flushes and night sweats, and poor sleep.