Wednesday April 13 2011
Girls are starting their periods earlier, but it is uncertain why
Genetics, and not diet, has the biggest influence on when a girl starts her periods, according to the Daily Mail. The newspaper says that a new study has shown that family history has a greater effect than lifestyle and a girl’s environment, which were previously thought to play the biggest role.
The new study examined the age at which 26,000 women experienced their first period (the menarche) and analysed how menarche ages were linked between related participants. The results suggest that genetics explains 57% of the variation in the group’s menarche age, confirming that there are likely to be multiple genetic, lifestyle and environmental influences.
The study itself was large and well-conducted, using modelling techniques to work out the proportion of variation that was due to factors such as genes, childhood environment and lifestyle. It is worth noting that the estimate that 57% of variation was based on genetics applies to the group of participants tested, and it cannot be assumed that this will be the same in other groups or the wider population. Also, it is not correct to say diet is not involved. In this population 43% of the variation was explained by something other than genes, which could include dietary factors.
Where did the story come from?
The study was carried out by researchers from the Institute of Cancer Research in Sutton and London. It was funded by Breakthrough Breast Cancer, the Sir John Fisher foundation and the Institute of Cancer Research.
The study was published in the peer-reviewed medical journal Paediatric and Perinatal Epidemiology.
The Daily Mail’s coverage suggests that genetics is the most influential factor in determining the timing of first period. However, this research clearly suggests that multiple causes are likely to interact with each other. Its headlines and pictures also seem to dismiss the influence of diet, which may still play an important role.
What kind of research was this?
The data for this study came from the Breakthrough Generations Study. This is an ongoing cohort study started in 2003 looking primarily into the causes of breast cancer. So far it has enrolled more than 111,000 women from the UK.
The researchers explain that the age a woman has her first period is associated with the risk of chronic diseases including breast cancer and diabetes. Therefore, they say, a better understanding of what determines the age at menarche could, theoretically, improve the understanding of the causes of a number of diseases.
Factors such as childhood body size, exercise and social and economic variables have all been linked to the timing of menarche through previous research. In this study the researchers were interested in assessing the influence of genes on the timing of menarche. This is called heritability. They wanted to look at heritability beyond first degree relatives (mothers, sisters etc.) so that they could discount the influence of any similarities between relatives due to shared environment or behaviour, such as shared dietary habits. Previous studies have examined the heritability of menarche age by examining first degree relatives, but may have incorrectly attributed the effect to genetic influences when shared family dietary patterns may have been an explanation.
To date, this is the largest study to look at heritability of age at menarche, and the methods used are appropriate to the question these researchers were asking. The fact that the age at which identical twins in the study began their periods was so strongly linked also supports a genetic link, although, once again, a substantial part of the variation may also be due to environment and lifestyle factors.
What did the research involve?
In this modelling study the researchers identified the participants from the records of the Breakthrough Generations Study. They selected women if they had a first-degree relative (mother, daughter or sister) or second-degree relative (half sister, grandmother, granddaughter, aunt or niece) who was also a participant in the study. They excluded those women who no longer had an eligible relative and those who had a history of breast cancer (as this was mainly a breast cancer study there was a higher-than-usual proportion of these women). They also excluded women who had never had a period or had started their periods after the age of 20. This meant that they were left with data on almost 26,000 related women from the original population of 111,000 women enrolled in the cohort study.
The researchers sent all participants a questionnaire for them to complete themselves, which provided most of the data used in the analyses. The participants were asked about when they started their periods, their weight and height at the age of seven (recalled relative to other girls of the same age that they knew at that age) and the amount of exercise that they did as a child outside of school hours. The researchers also scored each participant for socioeconomic factors using a score based on their postcode and census data, with the scores ranging from 1 (greatest affluence) to 5 (lowest).
The statistical analysis used a standard technique called linear regression. Using this technique the researchers looked at the extent to which a younger relative’s age at menarche could be explained by the age at menarche of their older relative.
In their calculations the researchers adjusted for factors that may also influence the menarche, including weight at seven years, height at seven years, socioeconomic status, exercise and year of birth. Results were expressed as the difference in age at menarche (months) relative to the average age at menarche, associated with each year delay of menarche in an older relative.
What were the basic results?
On average, women were 46.4 years old at study entry (range 16 to 98 years), and the average age at menarche was 12.7 years.
The researchers then looked at how menarche age in women (given in months) related to menarche age in their older relatives (given in years). There were different strengths of the association demonstrated for the different pairs of related women. For example, a woman’s age at menarche was significantly delayed by:
- 7.2 months for each yearly increase in the menarche of her older identical twin
- 3.0 months for each yearly increase in the menarche of her older non-identical twin
- 3.3 months for each yearly increase in the menarche of her older sister
- 3.4 months for each yearly increase in the menarche of her mother
- 3.0 months for each yearly increase in the menarche of her paternal aunt
There were smaller delays associated with menarche delays for a maternal grandmother and maternal aunt and, based on small numbers, no significant difference between the ages at menarche of half-sister pairs or of paternal grandmother–granddaughter pairs.
Heritability in the study population was estimated as 0.57 (95% confidence interval 0.53 to 0.61). This means that 57% of the variation in age at menarche seen in this selected population could be attributed to cumulative genetic effects.
How did the researchers interpret the results?
Based on data from a large number of related pairs with different combinations of shared genetic, environmental and childhood factors, the researchers conclude that approximately half of the variation in age at menarche was attributable to additive genetic effects. They also say that the remainder is attributable to non-shared environmental effects.
This study was based on a large amount of paired data and provides a reliable estimate of the heritability of menarche age. Accurately estimating the heritability of a given trait is an important first step that can inform us of whether there is merit in performing epidemiological genetic analyses. These are costly and time-consuming as they must analyse participants’ DNA to look at which specific genes are associated with the trait being investigated.
The idea of heritability, while intuitively simple, is prone to misinterpretation and it is worth pointing out that these researchers also say that their result should be interpreted with caution because:
- It only explains the variation in age at menarche in a particular study population and so the estimate cannot be automatically generalised to other populations. For example, participants in this study were volunteer recruits without a history of breast cancer who agreed to be followed for a number of years. They could potentially have been less diverse in terms of environmental and behavioural factors than a random sample of women.
- Genetic studies have identified several gene loci (regions) that are associated with age at menarche, but these explain at most about 2% of the variation in menarcheal age. This means that the vast majority of the heritability of age at menarche is not due to the loci so far identified.
- The study did not look at the interactions between genes and environment, for example how a particular genetic makeup may predispose someone to the effects of diet or environmental influences. This is a further avenue for future research.
Overall, this study adds to an understanding of the degree to which shared familial factors and non-shared, environmental or behavioural influences contribute to the variation seen in age at menarche. It is premature to say that one causative factor has been identified.