EXPERT REACTION: Early-stage research raises question over pesticide residues and IVF success

Infertile couples seeking to conceive using IVF can have significant barriers to successful pregnancy, and environmental factors may play a role in IVF failure. This paper suggests that one such environmental effect is pesticide residue intake on fruit and vegetables.

There are a number of limitations to this study: the number of participants is relatively small, dietary intake was determined by questionnaire (which could be subject to recall bias), other medication intake which could affect pregnancy was not surveyed, and pesticide intake was not confirmed by any measurements (either of pesticides on the food samples or measurement of pesticide residues in urine).
 
Possibly the most remarkable finding is that women who ate the smallest amount of 'high pesticide' fruit and vegetables had a total pregnancy loss that was much smaller than that of women who ate the smallest amount of 'low pesticide' fruit and vegetables (around 15 per cent compared to 40 per cent). This low level of pregnancy loss is remarkable in itself, but the implication that eating small amounts of 'high pesticide residue' fruit and vegetables is protective is not plausible. If this is a statistical aberration of the low numbers of participants, and the true figure is higher, the entire argument goes away.
 
In summary, while we should carefully study the potential effects of environmental contaminants of pregnancy, the current study does not provide plausible evidence that pesticide residues are a source of IVF failure.

For a woman undergoing infertility treatment, eating fruit and vegetables on which low levels of pesticides remain, may or may not reduce her chances of getting pregnant and experiencing a live birth. No matter what the authors and the editors say, a conclusion cannot be reached on the basis of the present study.

Proving cause and effect by correlating outcomes (be they pregnancy, birth, sperm counts, disease and metabolic disorders, mortality or whatever) with exposure to pesticides, industrial chemicals, air pollutants (to name a few of the usual suspects), is notoriously difficult. Nonetheless, the results of studies such as this catch our attention (and that of the editors of journals who publish these studies!) and can cause alarm. They shouldn't.

Researchers using epidemiological methods need to find a cohort of subjects who are exposed to the suspected cause, match them with a cohort of similar composition without such exposure (not done in this case), specify closely what outcomes are being studied, and consider how they could have come about by some other factor that they have not taken into account. A large group is needed - much larger than the 325 who participated in the present study. With small groups, there is always a danger of finding a 'false positive' result that comes about because of statistical variation. Try tossing a penny twenty times and you are unlikely to get 10 heads and 10 tails - that's statistical variation.

In the present study, the participants were already known to have impaired fertility - why else would they be undergoing fertility treatment?  While that doesn't rule out the possibility that this makes them unduly sensitive to environmental effects such as exposure to pesticides residues, they are hardly representative of the community at large. And the impacts on the high exposure group, 18 per cent lower probability of clinical pregnancy and 26 per cent lower probability of live birth, are modest. No factors of ten involved, just relatively small changes. For the low-exposure group, the impacts vanish.

Pesticide residues on food are tightly regulated in America and Australia but it's a complex area and there is room for doubt, albeit a very small doubt. The pesticides are considered by regulators on a case by case basis. There are no guidelines for the sort of mixtures to which people might be exposed by eating a range of fruits and vegetables that have been treated with a range of different pesticides. That multiple exposure is always a worry but there are no data to help us assess whether it's a serious worry or not. In the present case, the investigators did not even know which pesticides were involved, and - apart from knowing the maximum residue limits - they did not know how much of each or in total might have been ingested. 

Added to that uncertainty is the well-known flaw in recall studies, that the people don't remember accurately what they did during the last year. And what other exposures there might have been apart from fruit and vegetables. Fly and bug spray? Home gardening including lawn care?

So, is there a problem? Probably not, and the benefits of eating fresh fruit and vegetable would outweigh any impacts of the minute quantities of pesticide remaining on them.