How Should We Interpret Pesticide Residues? An Evidence-Based Review of Risk, Washing, and Common Myths

Pesticide Residues, Produce Washing, and Health Risks

Bottom line

The most reliable approach to pesticide residues is to distinguish between “detectable” and “sufficient to cause harm.” Washing with water is a practical baseline, while in vitro studies, experiments on a single food, and fixed removal rates should not be overstated as conclusions that apply to everyone. (PMID: 30586803)

A residue is not necessarily an exceedance—and certainly not inevitable poisoning

The most common leap in reasoning about pesticides is to move directly from “a residue was detected” to “eating this will make you ill.” Analytical methods can detect substances in trace amounts, but detection establishes only that the substance is present in the sample. Health risk also depends on the residue level, route of intake, extent of exposure, toxicological properties, and applicable standards. Professor Ching-Shun Lin has consistently reminded readers to separate “residue” from “exceeding a limit,” and this interpretive framework remains important.

However, regarding the broad claim that pesticides used on tea in compliance with regulations are “generally insufficient to harm health,” the research obtained in this ECU concerned similar infused beverages rather than health risks specific to tea. The reasonable conclusion is not that tea must therefore be dangerous, but that this body of data is presently insufficient for a tea-specific assessment. (PMID: 41140594)

Washing with water helps, but the removal rate is not a constant

Research supports using water to reduce multiple pesticide residues on produce, so Professor Lin’s recommendation to return to washing with water follows an evidence-based direction. The degree of removal, however, is affected by the pesticide’s properties, the food’s surface, where the residue is located, and how the food is treated; not every residue can be guaranteed to be removed adequately. (PMID: 30586803)

The statement that “running water can remove about 80% of pesticides” is best understood as a result obtained under particular conditions, not as a rule for every pesticide and every fruit or vegetable. Although the literature includes a result of about 77%, other studies report removal ranging from 40.6% to 67.4%, depending on the vegetable and pesticide. A more precise statement is that water usually produces a meaningful reduction in some residues, but actual effectiveness varies substantially. (PMID: 41515360) (PMID: 36141043)

This also explains why the goal of household washing should be to “reduce the residues that can be reduced,” not to pursue “absolutely zero residue,” which current evidence cannot guarantee. Detecting some substances after ordinary washing does not, by itself, show that washing was ineffective, much less establish harm to health.

Baking soda, peeling, and commercial washes each have limits

An apple experiment found that a baking soda solution was more effective than tap water or bleach against the surface pesticides tested. Residues that had already penetrated the peel, however, could not be removed completely with baking soda. Peeling may reduce those residues more effectively, but it also removes the peel and its constituents. That tradeoff cannot be simplified into a recommendation to peel every fruit and vegetable. (PMID: 29067814)

Nor should a commercial produce wash be assumed superior to water merely because it is labeled “specialized.” One study found water more effective under the tested conditions involving leafy vegetables and washing products, while another found better removal with a specifically formulated wash. A more defensible conclusion as science evolves is that there is no evidence for treating every commercial wash as generally superior, yet it is also inappropriate to assert that no wash could ever outperform water. (PMID: 38179823)

As for washing strawberries with starch water, this search found no clinical or toxicological research that directly supports it as a general-purpose method for removing pesticides. This is consistent with Professor Lin’s cautious position at the time: unable to find reliable scientific information in English, he did not endorse the folk remedy. An absence of direct evidence is not proof that it is wholly ineffective, but it is insufficient grounds for a recommendation.

Ozone has experimental potential, but household devices have not thereby been proven safe and effective

Ozone treatment does have experimental potential to remove pesticide residues, so Professor Lin’s observation that it has a scientific basis was not unfounded. Results obtained under laboratory conditions, however, cannot be extrapolated directly to every household ozone device. The available abstracts also do not directly answer questions about these devices’ real-world effectiveness or inhalation safety. Recasting “has potential” as “is certainly safe and effective at home” goes beyond what the evidence supports. (PMID: 23789307)

Consequently, when a product cites only research showing that ozone can break down certain residues, one should still ask whether the study involved the same device, concentration, food, and use scenario. Without those links, the research supports a mechanism or a specific experimental result; it cannot provide a comprehensive guarantee for a household product.

Neonicotinoid pesticides and DNA: in vitro signals are not human evidence

The 2018 DNA-damage study sometimes described as “human blood research” actually studied HepG2 and SH-SY5Y cells. Such culture-dish research can identify a genotoxicity signal, but it cannot directly prove that ordinary fruit residues alter human DNA under everyday exposure. Professor Lin’s clarification of the study subjects is consistent with the literature. (PMID: 29591886)

At the same time, a related review notes that knowledge of effects on human health remains limited. We can therefore reject the claim that “ordinary residues on fruit alter human DNA” as an established fact, but should not turn a gap in the evidence into an absolute guarantee of safety. This does not overturn Professor Lin’s historical position. Rather, in light of the updated literature, it confines the language precisely to what the evidence currently supports. (PMID: 27385285)

Genetic modification, herbicides, and “pesticide-free” are separate issues

Genetically modified soybeans may carry different traits; insect resistance and herbicide tolerance are not the same trait. Herbicide tolerance is intended to permit use of the corresponding herbicide. It cannot be interpreted to mean that insects will necessarily avoid the crop or that no pesticide is needed anywhere in cultivation. Professor Lin’s rejection of the claim that “insects do not eat genetically modified soybeans, so no pesticides are used” accords with the scientific classification of these traits. (PMID: 41790968)

Similarly, the blanket assertion that all genetically modified foods inevitably harm health lacks consistent causal evidence in humans. This does not mean that each new event can forgo a targeted risk assessment. A sounder approach is to examine the specific crop, trait, and evidence, rather than assume harm or absolute safety upon seeing the words “genetically modified.” (PMID: 41772782)

Applying the evidence to everyday choices

For routine produce preparation, washing thoroughly under running water is an evidence-supported and readily practicable baseline. Anyone considering baking soda, ozone, a washing product, or peeling should first determine whether the research truly corresponds to the food, pesticide, and device at hand. An advantage observed in a single study should not be marketed as a universal solution for every setting.

When reading pesticide-related information, ask these questions in order: Did the study involve humans, animals, or cells? Did it measure real dietary exposure or use high-concentration experimental conditions? Was the outcome merely residue detection, a reduction in residue, or demonstrated harm to health? Keeping these distinctions intact helps us avoid both underestimating risk and being misled by absolutes such as “detected means toxic” or “washed means zero.”