Medical practice

New clue to early menopause

“A test could warn women of early menopause,” according to The Guardian. The newspaper reported that the test might aid the “growing number of women who put off children until their thirties but then find that they cannot conceive”.

The news is based on a study examining particular genetic variations that previous research had linked to early menopause, occurring before the age of 45. The study has compared the DNA of more than 2,000 women who had experienced early menopause with that of women who had menopause later than 45 years. It found that four particular genetic variations might account for part, but not all, of the risk of early menopause.

The researchers themselves have said that testing for these variations cannot predict whether a woman would have early menopause or not because other unknown regions of DNA are also likely to affect menopause. Further research is also needed to determine how these variations affects the function of the surrounding DNA. Nevertheless, this research is a useful early step in developing tools for predicting early menopause.

Where did the story come from?

The study was carried out by researchers from Peninsula Medical School and funded by The Institute of Cancer research. The study was published in the peer-reviewed medical journal Human Molecular Genetics.

The newspapers tended to suggest that a genetic test would be readily available based on this research. However, this preliminary research has demonstrated that more knowledge about the genetics of early menopause is needed before such a test could accurately predict who is likely to have an early menopause. Even if a test were developed in the future, as with all genetic testing, the risks and benefits would have to be carefully considered to ensure that each person could make a fully informed decision about whether to have such a test.

What kind of research was this?

This was a case control study that looked at the genes of women who had experienced menopause before the age of 45, and compared them to the genes of matched control women. The researchers considered menopause at this age to be early, as the average age at which menopause occurs (in Caucasian populations) is reportedly 51 years. They also estimate that 5% of women experience menopause before age 45.

The researchers say that women become infertile approximately 10 years before menopause, which can occur at any age between 40 and 60 years of age. They said that current methods for predicting menopause can only do so just prior to the onset of menopause. The researchers were interested in seeing whether it was possible to make an earlier prediction of when a woman would be likely to experience menopause, providing her with an estimate of when she would be most likely to be fertile and able to have a child.

The researchers focused on four regions of DNA that had been associated with early menopause in previous genome wide association studies.

What did the research involve?

The researchers selected 2,118 women who went through natural menopause prior to the age of 46 years, and 1,261 ‘control’ women who had menopause after the age of 45. These women were taking part in the Breakthrough Generations Study (BGS) - a separate prospective cohort study launched in September 2004, which investigated the environmental, behavioural, hormonal and genetic causes of breast cancer. This cohort had not been used in previous genome wide association studies of early menopause.

The participants gave a blood sample for genetic analysis. They were also asked questions about their menstrual histories. Natural menopause was defined as absent menstruation for at least six months without known cause. Women were excluded from the study if their periods had stopped due to pregnancy, breastfeeding, surgery, use of hormonal contraceptives or other medical treatment. Women were excluded if they had a medical condition that could cause their periods to stop, such as polycystic ovary syndrome.

For each woman who had experienced early menopause (a case), the researchers selected a control woman who had her menopause after 45 and was of a similar age and ethnicity. These matched controls had also been recruited into the study from a similar source and at a similar time.

The researchers looked at four regions of DNA on chromosome 20, 6, 19 and 5 to look for differences in the sequences of DNA (called single-nucleotide polymorphisms SNPs) between the participants of the study. They first compared the women who had experienced early menopause with women who had gone through menopause after 45. Then they divided the cohort of early menopause women into those who had menopause before the age of 40 (premature ovarian failure, or POF) and women with menopause between 40 and 45 years.

What were the basic results?

The researchers found that common genetic variants of the sequence on chromosome 19 and 20 affected the age of menopause in all of the women. A genetic variation (SNP) on chromosome 19 was associated with a reduction in menopause age of three months, whereas a SNP on chromosome 20 was associated with a rise in menopausal age of 11 months. Using a statistical model, the researchers estimated that in the control women (i.e. normal-onset menopause), variations in all four chromosome regions explained 1.4% of the variation in menopause age.

Women with early menopause were more likely to possess each of the risk SNPs. The likelihood was between 13% to 85% greater than in the women who had menopause after 45. The researchers then looked at whether women had two copies of the risk SNPs (known as homozygous) or whether they had two different forms of the SNP region of DNA (heterozygous). They found that only 3% of women were homozygous for all four risk variants. Of these 97 women, 66 (68%) were in the early menopause group and 31 (32%) were in the control group.

The lowest number of risk SNPs (two or three) was seen in 4.5% of the women. When the researchers compared these women to the 3% of women who were homozygous for all four risk variants (i.e. they had eight risk SNPs) the women with the lowest number of risk SNPs were four times less likely to have early menopause (odds ratio 4.1, 95% confidence interval 2.4 to 7.1).

There were 260 women who had premature ovarian failure and had experienced menopause before 40. The chances of having both a risk SNP and having POF were approximately the same, as were the chances of having a risk SNP and having early menopause. However, the researchers say that as there were a small number of women with POF, their calculations in this area may not have had the statistical power needed to detect any true differences.

The researchers created a model relating the influence of risk SNPs to the risk of early menopause. They then applied statistical tests to assess how well the model could discriminate between a case (a woman with early menopause) and a control. In this test, a score of 1 means that the model can discriminate between all cases and controls perfectly. A score of 0.5 means that the model has no predictive power. The researchers found that the model based on the four risk SNPs had a score of 0.6.

How did the researchers interpret the results?

The researchers say that four common genetic variants found in genes on chromosomes 19, 20, 6 and 5 affect the likelihood that a woman will have early menopause. However, they say that the discriminative power of these genetic variations is limited, meaning that by looking at these genetic variations alone it would be difficult to predict whether a woman would have early menopause. The researchers suggest that as more gene variations are discovered, they may be useful in predicting reproductive lifespan.


This was well-conducted research, which confirmed previous findings that associated genetic variations in regions of DNA on four chromosomes are associated with early menopause. However, there is still not enough information to predict who will experience early menopause based on their genes.

Although the research has found high-risk regions of DNA associated with early menopause, the research did not determine the functions of the genes in these regions. Further research is needed to assess whether these genetic variations may affect proteins involved in reproductive processes.

It is also worthwhile to consider environmental and lifestyle factors that may affect fertility and menopause, such as smoking and body mass index. The researchers found that their results were not affected when adjusting their analysis to take smoking status into account. However, there may have been other environmental factors that may have influenced (confounded) the results but were not accounted for.

This research is a useful, if early, step in developing a predictive profile for women who would be likely to have early menopause.

NHS Attribution