I’ve often suspected that the benefit of eating organic food is over-rated
So I reviewed some of the USDA data which is what most sources use to asses non-organic food risks/hazards. I was mostly looking at Appendix B of this report.
The USDA assesses food/pesticide pairs (pairs) and shows the number of samples that were detected to have a known pesticide/fungicide/insecticide (pesticide), the range of values of that pesticide on the pair, and the number of samples that went over the EPA tolerance level (EPA-TL) for that pair.
I was concerned mainly with the percentage of samples in each pair that exceeded the EPA tolerance level.
There were around 50 fruits and vegetables and 500 pesticides in their report.
After analyzing the data I found that when eating one of these foods at random you have a 3.8% chance of eating a sample that exceeds the EPA-TL for one or more pesticides.
Of the 185 violations that occurred, basil accounted for 149 of them. Since Basil is a clear outlier it’s interesting to see what happens when we remove it from the data.
If you were to avoid non-organic basil your chance of eating a sample that exceeds the EPA-TL for one or more pesticides drops to 0.9%.
After mulling this over it seems to me your motivation to eat organic shouldn’t be motivated only by the estimated percentage of pesticides you will eat, but also how hazardous exceeding the EPA-TL really is. I have no idea what the answer to that question is.
In addition to the question above it would also be useful if the USDA data showed the average difference between the violating samples pesticide levels and the EPA-TL rather than just a range of pesticide levels per pair.
Basil in particular had many pairs that often exceeded the EPA-TL by more than 10x and 3 pairs that exceeded the EPA-TL by more than 100x.
After this short project I feel as though my hesitancy towards purchasing organic food for health purposes is about right, with the exception of organic basil.
Appendix H of the USDA report shows pairs with over 5% detection rate. In this appendix it shows the mean of detection levels for each pair but does NOT separate out the mean of detection levels for each pair that violated the EPA-TL.
Really, the best heuristic you could use for safety is Appendix H. If the mean detection rate does not exceed the EPA-TL and assuming a reasonably low standard deviation you could just eat the foods which have a mean detection rate below the EPA-TL. Which is all of the pairs listed there.
My lil’ analysis has the following flaws:
- Completely trusting EPA-TL
- No data for average strength of EPA-TL violations
- No data for exactly how hazardous each pesticide is at varying doses