Drinking coffee can double the risk of miscarriage reported The Guardian and many other news sources yesterday. “Pregnant women who consumed two or more mugs of coffee a day were twice as likely to miscarry than those who abstained from caffeine,” The Guardian said. The media coverage suggested that pregnant women may wish to reduce or stop drinking drinks containing caffeine, including coffee and tea.
The newspaper story is based on a study in 1,063 pregnant women in San Francisco. The study found that women who drank more than 200mg or more of caffeine a day – the amount contained in two or more regular cups of coffee or five 12oz (330 ml) cans of caffeinated drink – doubled their risk of miscarriage compared with women who drank no caffeine. However, this study has some limitations, including difficulty in making sure that the results are not affected by other factors that are known to increase the risk of a miscarriage.
Overall, the advice that pregnant women should avoid drinking too many drinks containing high levels of caffeine while they are pregnant seems sensible. Currently in the UK, the Food Standards Agency recommends that women limit their caffeine intake during pregnancy to 300mg a day.
Where did the story come from?
Dr De-Kun Li and colleagues from the Division of Research of Kaiser Permanente (a US not-for-profit integrated health care organisation) carried out this research. The study was partly funded by the California Public Health Foundation. It was published in the peer-reviewed: American Journal of Obstetrics and Gynecology.
What kind of scientific study was this?
This was a prospective cohort study looking at caffeine consumption during pregnancy and risk of miscarriage in women. All women from the San Francisco and south San Francisco areas who were members of the Kaiser Permanente Medical Care Program (KPMCP) and had a positive pregnancy test between October 1996 and October 1998 were asked to participate in the study. Of the 2,729 women researchers asked, 1,063 agreed (39%) and they completed the enrolment interview before their 15th week of pregnancy.
The enrolment interview was carried out in person, and the women were asked about what caffeine-containing beverages they had drunk since their last period. This included coffee and tea (either caffeinated or decaffeinated), caffeinated fizzy drinks and hot chocolate. They were asked how often they drank these drinks (either in a day or a week), how much they drank, at what point in the day they drank them, and if they had changed their consumption of caffeine since becoming pregnant. Some women had already miscarried by the time of their enrolment interview, and they were only asked about caffeine consumption up to the end of their pregnancy.
The researchers used the women’s answers to calculate their average daily caffeine intake. For every 150ml of liquid, caffeinated coffee was estimated to have about 100mg of caffeine, decaffeinated coffee 2mg of caffeine, 39mg for caffeinated tea, 15mg for a caffeinated fizzy drink, and 2mg for hot chocolate. The interviewers also asked the women other questions about themselves and other factors that might affect risk of miscarriage including their age, race, income, marital status, smoking, alcohol consumption, Jacuzzi use, exposure to magnetic fields in pregnancy, whether they had experienced nausea and vomiting associated with their pregnancy, and whether they had had a previous miscarriage.
The researchers followed women up to the 20th week of their pregnancy, to find out whether they had a miscarriage. They did this by searching the KPMCP in-patient and out-patient databases, looking at the women’s medical records, and by contacting the women themselves if records could not confirm their pregnancy status. The researchers then looked at whether women’s risk of miscarriage before 20 weeks varied with their caffeine intake (categorised as no caffeine a day, less than 200mg a day, or 200 mg or more a day). They also took into account the other factors that might affect miscarriage risk.
What were the results of the study?
Most women drank beverages containing up to 200mg of caffeine a day during their pregnancy. Women who drank more caffeine were more likely to have certain risk factors for miscarriage, including being older than 35, not having experienced pregnancy-related vomiting, having drunk alcohol since their last period, having used a Jacuzzi during pregnancy, and having had a previous miscarriage.
By 20 weeks, 172 of the 1,063 women (16%) had miscarried. Consuming 200mg or more of caffeine a day more than doubled the risk of miscarriage compared with drinking no caffeine, from 12 to 25 per cent. Although drinking lower amounts, up to 200mg of caffeine a day, did increase the risk of miscarriage compared with drinking no caffeine, this increase did not reach statistical significance.
What interpretations did the researchers draw from these results?
The researchers concluded that high levels of caffeine consumption in pregnancy did increase the risk of miscarriage.
What does the NHS Knowledge Service make of this study?
Although this study does have some strong points, such as its relatively large size and the fact that it attempted to follow women prospectively, it does have some limitations:
- Only about four in 10 pregnant women who were asked, agreed to participate in the study. This may mean that the women studied were not representative of the population of the San Francisco and south San Francisco areas as a whole. The study only looked at women in a relatively small geographical area, and this may not be representative of women from different countries or of different ethnic backgrounds (for example, there was a low proportion of black women in this study – only about seven per cent).
- Although the study tried to collect its data prospectively, some women had had a miscarriage before they were interviewed about their caffeine consumption. The fact that these women had experienced a miscarriage may have affected their reporting of their caffeine intake (either making them over- or underestimate), particularly if they thought it might have played a role in their miscarriage. It is unclear what proportion of women had experienced a miscarriage before their interview. However, the authors reported that they conducted their analyses separately according to whether women had experienced miscarriage before the interview, and found similar results.
- The measure of the amount of caffeine drunk is also likely to have some level of error. Women were also asked to remember how much caffeine they had drunk since their last period. This may have been difficult to remember accurately, particularly in terms of exact volumes. Amounts of caffeine in drinks such as coffee will also vary with the type of coffee used and the method of preparation.
- One of the main difficulties with interpreting this type of study is that because groups are not randomised, they are not likely to be balanced for factors that might affect outcome. This was the case in this study, where the authors report that women who drank the greatest amounts of caffeine were also more likely to have risk factors for miscarriage, including being older, having had miscarriages before, and having consumed alcohol. Although the researchers did try to account for these factors in their analyses, it is difficult to ensure that the adjustments they made would eliminate their effect. Other, unknown, factors may also be imbalanced between the groups and could be affecting the outcome.
It seems like the old saying “everything in moderation” applies here. If women are worried about their caffeine intake, it might be a good idea to moderate it for the duration of their pregnancy.
Sir Muir Gray adds...
Another bit of information that women who are pregnant should take on board, but it is always useful to see the findings of other research studies.
Analysis by Bazian
Edited by NHS Website
Links to the headlines
The Guardian, 21 January 2008
The Daily Telegraph, 21 January 2008
The Times, 21 January 2008
Daily Mail, 21 January 2008
Channel 4 News, 21 January 2008
BBC News, 21 January 2008