Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Folate in pregnancy and breast cancer

Study [1]
Results
Comment

Much media attention has been given to a possible link between folate use in pregnancy and breast cancer developing years later. This emphasises the problems of deciding whether there is a problem, and how much bigger the problem is with one intervention than another.

Study [1]

This is a follow up of a randomised trial conducted in the late 1960s where women were randomised to placebo (1971), 0.2 mg folate (488), or 5 mg folate (488) daily. Patient records were held by a registry, and the 210 deaths out of 2928 women randomised by 2002 analysed. All cause mortality, cardiovascular mortality, and all cancer and breast cancer mortality were separately analysed, both unadjusted and adjusted for maternal age, smoking, height, weight, and other variables.

Results

There were no statistically significant differences for any outcome between women who had taken 0.2 mg folate and those who had taken placebo.

There was one statistically significant result for women who had taken 5 mg folate. All cause cancer deaths were more common, with 24 of 488 (4.9%) with 5 mg folate, and 69 of 1971 (3.5%) with placebo. The adjusted hazard ratio was 1.7 (1.06 to 2.7), but the unadjusted ratio had a confidence interval that included 1.

Results for all cause cancer deaths and breast cancer deaths are shown in Figure 1, together with the number of events. For breast cancer deaths, neither adjusted nor unadjusted hazard ratio had a confidence interval that included 1.



Figure 1: Cancer and breast cancer deaths over 35 years, percent and number of cases





Comment

The authors of this study have linked folate intake in pregnancy and breast cancer mortality over the succeeding 35 years. The title of the paper specifically makes the link: "Taking folate in pregnancy and risk of maternal breast cancer".

Yet there was no significant association between breast cancer mortality and folate, even with a 5 mg of folate, a dose about 10 times higher than that taken in pregnancy. Out of over 20 tests of significance, only one was statistically significant, about what we might expect by chance alone using 95% confidence. And the number of breast cancer events was only 14 with folate, and only six with 0.2 mg and eight with 5 mg (Figure 1).

So where is the news? What is the message? Is there any message from a study too small to detect anything, and in which nothing was detected? Bandolier has long been a supporter of a journal of negative results (maybe there is one), and perhaps that would have been the right journal for this paper.

Reference:

  1. D Charles et al. Taking folate in pregnancy and risk of maternal breast cancer. BMJ 2004 329:1375-1376.

previous or next story