select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

expert reaction to study looking at preconception alcohol consumption in men and incidence of birth defects

A study published in JAMA Pediatrics looks at paternal drinking before pregnancy and fetal birth defect risk.


Prof Rod Mitchell, Professor of Developmental Endocrinology, MRC Centre for Reproductive Health, University of Edinburgh, and Consultant Paediatric Endocrinologist, Royal Hospital for Children and Young People, Edinburgh, said:

“This prospective population-based study in a large population report an association between alcohol consumption in the father and birth defects in the children.  They report a higher rate of birth defects in children born to fathers who had consumed alcohol (defined as at least once per week) when planning a partner pregnancy within 6 months.  The concept of paternal factors impacting on the health of offspring has been demonstrated in a number of animal studies that report paternal environmental and lifestyle factors may result in effects in the offspring.

“However, there are important limitations to this study that should be taken into account when interpreting the data:

“In this study, only 0.1% of children were reported to have birth defects, which is much lower than would be expected in the general population (3%).  This suggests that not all birth defects are identified in this study, which could affect the validity of the conclusion.  This under-reporting of birth defects may relate to the fact that birth defects were reported by the parents rather than being based on clinical registry data.

“The study was unable to determine the amount or frequency of alcohol consumption by the father.  In addition, the duration and timing of consumption in relation to conception is not described.  This is particularly important when trying to determine whether exposure to alcohol could lead to the development of birth defects.

“This study is not able to determine whether an association between paternal alcohol consumption and birth defects in the children would be due to effects on sperm as this was not investigated.

“An association does not prove causation and there are several other important environmental and lifestyle factors that could also account for such an association, including maternal and paternal factors.  Whilst the study did account for some confounders such as smoking and maternal age, other important factors to consider are stress, diet and obesity, all of which are shown to have potential to affect the offspring in animal studies.

“Whilst caution should be applied when interpreting this study, the overall benefits of moderating alcohol consumption in terms of fertility and general health are well-recognised from other previous studies.

“Despite the limitations of this study, it serves as a reminder to consider paternal exposures in studies that investigate outcomes in the offspring.”


Prof Allan Pacey, Professor of Andrology, University of Sheffield, said:

“There is growing interest in whether aspects of paternal lifestyle can have an impact on the health of any children born.  However, this is an immensely difficult area to study because the data collected is often retrospective and it usually relies on the man being able to accurately recall aspects of his lifestyle.  There is also the issue of disentangling any lifestyle factors that are shared by both partners.  It’s also difficult to determine cause and effect from correlation.

“As such, this is an interesting study looking at the pre-pregnancy alcohol intake of a large number of couples in China trying for a baby.  Given the number of patients involved, the study probably has a lot of statistical power.  However, unfortunately, it still only shows a correlation rather than causation.  It is not clear whether alcohol consumption is the key factor linked to birth defects or whether alcohol is a surrogate for another factor which the authors have not measured (e.g. diet or occupation).  Moreover, the report itself is rather brief and it’s not possible for the reader to be certain what variables the team have controlled for and how they have done that.”


Prof Kevin McConway, Emeritus Professor of Applied Statistics, The Open University, said:

“This is very far from being the first study that has provided some evidence of a correlation between alcohol consumption by fathers around the time of conception and various risks for the baby.  But the findings of the studies are not straightforward to interpret, and this study is no exception to that general rule.  A difficulty is that this study, and other studies in humans, have been observational.  (There have also been some findings from studies in animals, but it’s always difficult to be sure how far their results might apply to humans.)  In an observational study like this one, the researchers don’t try to make potential fathers drink alcohol or not – instead they observe what happens without interfering with it.  So, in this study, they recorded data from a large number of couples (over half a million) on whether the parents drank alcohol before conception, along with various other characteristics of the parents, and then they followed them up to the end of resulting pregnancies and recorded whether the baby had birth defects.  They found that the chance of a birth defect was greater in couples where the father had drunk alcohol before conception.

“The problem is that there are bound to be many differences between couples where the father drank alcohol before conception, and those where the father didn’t drink.  Any combination of those other differences might be the cause of the differences in the risk of birth defects, rather than the father’s use of alcohol.  Because the study took place in China, one particular possible alternative explanation seems unlikely: in many populations, one might find quite a strong association between the father’s and the mother’s use of alcohol before conception – so that an explanation for the finding might be that, if the father drank alcohol before conception, the mother would be more likely to drink alcohol before conception too, and the increased risk of birth defects could be because of the mother’s drinking.  (The evidence from previous studies that certain patterns of alcohol drinking by mothers can cause increases in the risk of birth defects is pretty solid.)  However, in China, principally for cultural reasons, women are very much less likely than men to drink alcohol.  That was certainly the case in this study, where almost one in three fathers drank alcohol around the time of conception, but only one in thirty mothers did.  But there are many other ways that the couple where the father drank alcohol might be different from the couples where the father did not drink, in a way that might affect the risk of birth defects.  It’s possible to make statistical adjustments to try to allow for other differences, where data are available, and indeed these researchers did make several such adjustments (including adjusting for whether the mother drank before conception, even though that’s unlikely to have made much difference) – though only one of them (adjusting for whether the father smokes during the pregnancy) related to the father directly.  However, one can never be confident that these adjustments have dealt with everything important, and adjustments can’t be made at all for differences on which the researchers have no data.  Perhaps they had no data on other possible risk factors that applied to the fathers before conception, for example.  So while the results are compatible with alcohol drinking by the father before conception actually causing the increased risk of birth defects, they absolutely can’t prove that that is the pattern of cause and effect.  It might work in a different way, partially or entirely.  The statistical methods that the researchers used are appropriate, but there’s no way they could have got round this limitation of observational studies.

“There are other limitations of what can be found, resulting from the sort of data that the researchers collected.  For instance, the data on the father’s alcohol consumption didn’t record how much they drank, only how often they drank (with those who said they drank at least once a week being counted as alcohol drinkers).  If it really is the father’s alcohol consumption causing birth defects, you’d expect that the risk would increase with the amount that the father drank – but the researchers couldn’t investigate that.  And you might also expect the risk to be greater for a father who gets very drunk every other week, than for a father who has one small drink once a week – but for this study, the first wouldn’t count as a drinker and the second would.

“It’s worth pointing out that the risk of birth defects was very low, overall, for these couples.  For every 10,000 couples who didn’t consume alcohol before conception, about 10 pregnancies resulted in birth defects.  After making the statistical adjustments, the estimated number in 10,000 couples where the father did drink before conception increased by about 3 to around 13, though there’s a bit of statistical uncertainty about that and the increase could be somewhere between 1 and 6 in 10,000 pregnancies.  So not a large increase in what’s a pretty small risk anyway – and we don’t know how much, if any, of the increase is actually caused by the fathers’ drinking.

“Personally, I’m long past the age where I’ll be planning any more babies.  If I were younger, this new study would be unlikely, on its own, to affect my decision on whether to avoid drinking alcohol before conception.  Looking at the combined picture from a range of previous studies, not this one alone, I might take a cautious view and not drink alcohol, but that’s because it’s a possible risk even if it hasn’t been proven beyond doubt, and I tend to be cautious about possible risks about important things like my children’s health.

“Apart from the general limitations that I’ve described, there’s another aspect that could make the new study less relevant to people from non-Chinese populations.  Studying possible effects of alcohol on health in Chinese populations, as opposed to some other populations, can have two advantages.  One is the strong cultural tendency of Chinese women not to drink, as I’ve mentioned, which means that comparisons of men and women can be particularly informative.  Another is that, in Chinese and some other East Asian populations, a large proportion of people carry a genetic variant in a single gene that can make the after-effects of alcohol particularly unpleasant for them.  As a result, people with that variant are very much less likely to drink alcohol.  This provides opportunities for researchers, who can get a better handle on what might cause what in observational studies involving alcohol, if they work with Chinese or other East Asian populations and can use data on whether each individual has this genetic variant or not.  That has been done in several studies previously, using a method called mendelian randomisation, though not (as far as I know) on pregnancy outcomes.  But the genetic difference may make it less likely that the results from this new study can be applied to other populations, like the one I’m from, where the genetic variant is less common.”



‘Association of Preconception Paternal Alcohol Consumption With Increased Fetal Birth Defect Risk’ by Qiongjie Zhou et al. was published in JAMA Pediatrics at 16:00 UK time on Monday 19 April 2021.



Declared interests

Prof Rod Mitchell: “None.”

Prof Allan Pacey: “Chairman of the advisory committee of the UK National External Quality Assurance Schemes in Andrology, Editor in Chief of Human Fertility, Trustee of the Progress Educational Trust (Charity Number: 1139856) and Trustee of the British Fertility Society (Charity Number:1075661) (all unpaid).”

Prof Kevin McConway: “I am a Trustee of the SMC and a member of its Advisory Committee.  However, my quote above is in my capacity as an independent professional statistician.”

in this section

filter RoundUps by year

search by tag