Sense about Science ? equipping people to make sense of science and evidence
For the record
Heatwave advice to pregnant women
In June 2007, the Metro, The Times and Daily Mail ran a story based on a press release from the Royal College of Obstetricians and Gynaecologists (RCOG) which reported a “heatwave warning for pregnant women”. The articles were headlined, “Sun can damage unborn babies” (The Times) and “Pregnant women should avoid the sun” (Metro, Daily Mail). These articles stated that the research publicized by the RCOG, which used records of women who gave birth in Aberdeen in the 1950s, found that exposure to high temperatures in the first three months of pregnancy could lead to babies being born with a slightly lower birth weight. Some articles went on to link such babies with being prone to a low IQ and learning disabilities and as being more likely to exhibit behavioural problems at school.
Professor of obstetrics Andrew Shennan and clinical research fellow Dr Manju Chandiramani at the Maternal and Fetal Research Unit, King’s College London, and public health statistician Dr Christina Pagel of the Clinical Operational Research Unit, University College London respond below.
Summary of their main points
- There is no plausible reason why minor changes in temperature should influence birthweight because the mother’s core and baby’s temperature simply would not change. Even with an increase in temperature of 20 degrees Celcius, the change in birthweight is probably meaningless in terms of health, and is less than the difference between girls and boys at birth
- Environmental and dietary conditions were different in the 1950s—for instance at that time many pregnant women smoked, something which is known to significantly affect birthweight. The authors of the research make it clear that the effects of social class and birth order have a much higher impact on the eventual birthweight of the baby, making the presence or absence of a temperature effect immaterial. They do not claim that their study should be used in the way in which the Royal College of Obstetricians and Gynaecologists have used it
- The advice given in the press release is applicable in nearly all its points to anyone in hot weather. Pregnant women need not worry about the effect of temperature on their unborn babies’ health
Full comments from Dr Manju Chandiramani and Professor Andrew Shennan:
A study reviewing past records of over 12,000 individuals born in Scotland in the 1950s has suggested that pregnant women should protect themselves from relative temperature extremes, especially high temperatures in early pregnancy. The basis for this is that a 1C increase in a 10-day period in the first trimester was associated with a 5.4 g decrease in birthweight. The same increase in the third trimester was associated with a 1.3 g increase. Even if there was an increase in temperature of 20 degrees Celcius (a rare event in Scotland), a 108g change in birthweight is probably meaningless in terms of health, for example this is less than the difference between girls and boys at birth.
There is no plausible reason why minor changes in temperature should influence birthweight (the mother’s core and baby’s temperature simply would not change). The human body easily copes with changes in body temperature through behaviour (e.g. sleeping with no covers) and sweating. Although it is correct to warn pregnant women regarding sensible behaviour during a heat wave (i.e. drinking plenty of fluid, sunscreens), minor changes in ambient temperature are highly unlikely to influence the health of their unborn children. For example, long distance runners who may just be pregnant are advised to reduce training times to avoid prolonged high temperatures during fetal development.
As both an increase and decrease in weight were found with higher temperatures in this study, this may be a chance finding that is not real. Research that involves retrospective examination of data occurs all the time. On the rare event that it finds something statistically significant then a paper is written, even when there is no plausible scientific reason for the findings and it may be accepted by a journal. It is rare for this type of evidence to change practice, as doctors are rightly conservative about this evidence. However it does direct future research, which occasionally finds something new and worthwhile. The best studies are predefined, and just look at one important thing (e.g. randomised controlled trials). When these find something significant, not only is this real, but also causative.
Unless future prospective studies prove birthweight is related to ambient temperature, pregnant women need not worry.
Full comments from Dr Christina Pagel:
Firstly, the paper concerns a group of women who were having children in the 1950s in Aberdeen, raising the following issues:
- Environmental conditions, dietary advice and so forth were completely different back then to now. For instance, smoking is not taken into account (because that information is not available), yet at that time many pregnant women smoked, something which is known to significantly affect birthweight. Women nowadays eat very differently and live differently, and are generally healthier. It is also possible that back then people were restricted to more seasonal diets than is normal now. All of these factors could affect birthweight. It is not really valid to compare the women then to women now when such things have not been taken into account.
- The temperatures in this study, even in the height of summer, are about 10-15C. This is much colder than is normal in most of England in the summer and certainly does not qualify in any way as ‘heatwave’ temperatures. The range of temperatures discussed in this paper, therefore, cannot have much to say one way or another on the effect of ambient temperature on birthweight. A more interesting study would look at birthweights in traditionally hot places near the equator that have a similar standard of living to the UK (eg Hawaii/Bermuda) if we really wanted to know about ‘extreme’ heat and its effect on pregnant women. To be fair to the authors, they do not claim that their study should be used in the way in which the Royal College of Obstetricians and Gynaecologists have used it.
- How accurately could the weight of babies in the 1950s be measured? The authors give birthweights to the nearest hundredth of a gram - were these weights really so accurately recorded? The error in the measurements may be what accounts for the effect the authors are reporting.
Secondly, as the authors make clear, the effects of social class and birth order have a much higher impact on the eventual birthweight of the baby - about 10 times larger than say a 10 degree temperature difference! This makes the presence or absence of a temperature effect immaterial. From a statistical point of view, the authors have gone looking for an effect and tested lots of different things. They found something, but if you look hard enough you’ll always find something. They have not corrected for this ‘multiple testing’ effect in presenting the significance of their results.
The advice given in the Royal College of Obstetricians and Gynaecologists press release is applicable in nearly all its points to anyone (stay covered, drink water, don’t go out between 11am and 3pm etc), and it is known that pregnant women can be prone to high blood pressure and so should be careful of that. So in that sense, the advice is not ‘bad’ but it is somewhat unnecessary. I certainly don’t think that women should be worrying about the effect of such temperatures on their babies’ health!
Author: Sense About Science
Document type: For The Record
Published: 6 September 2007