Quintana, PJE, RJ Delfino, S Korrick, A Ziogas, FW Kutz, EL Jones, F Laden, and E Garshick. 2004. Adipose Tissue Levels of Organochlorine Pesticides and Polychlorinated Biphenyls and Risk of Non-Hodgkin’s Lymphoma. Environmental Health Perspectives 112: 854-861.

Quintana et al. found strong associations between the risk of non-Hodgkin's Lymphoma (NHL) and levels of several organochlorine compounds measured in body fat, particularly heptachlor epoxide and dieldrin. People with the highest body burdens of heptachlor epoxide and dieldrin had three times the risk of NHL, compared to people with the lowest levels of these persistent pesticide residues in their body fat.

Subjects were excluded from the study if they had been poisoned by pesticides, or were not considered to be representative of the general population for other reasons.

Quintana et al. identified 203 people in the EPA dataset as diagnosed with NHL. Pesticide exposure levels were missing for 28 of those, leaving 175 cases for this study. These cases were then matched to 481 controls who did not have a cancer diagnosis and who had been diagnosed with either a heart attack (myocardial infarction) or an accidental injury or death. Controls were matched to cases by gender, race, age and geographic region

For certain compounds, Quintana et al. divided up the samples into 4 groups, or quartiles, based on contaminant residue levels, with the same number of subjects in each quartile. The first quartile contained the 25% of samples from controls with the lowest concentrations in fatty tissue were in the first quartile, and the 25% with the highest concentrations were in the fourth quartile. Their statistical analysis looked for trends in odds-ratios across quartiles.

The researchers noted that the levels of some compounds in fatty tissue tended to correlate with others, both in cases and controls. They therefore carried out two separate statistical analyses, first each contaminant by itself, and then contaminants in pairs. The latter was undertaken to determine if some of the patterns seen when looking at pesticides one-by-one were the result of correlations among the pesticides.

What did they find?

Six of the nine organochlorines included in the final analysis showed a positive association between higher contamination levels and risk of NHL: p,p'-DDT , p,p'-DDE, ß-BHC, dieldrin, oxychlordane, and heptachlor epoxide. These six all showed a significant trend of increased risk with higher exposure levels. Three did not: trans-nonachlor, hexachlorobenzene and PCBs.

[Significance levels for the trend lines can be seen by hovering over the contaminant.]

The strongest associations were seen for dieldrin and heptachlor epoxide (graph to right). Cases with the highest levels of heptachlor epoxide in their fat (those in the highest quartile) were more than three times as likely to have NHL compared to those with the lowest levels of heptachlor epoxide in their fat (those in the lowest quartile). For dieldrin there was a 2.7 fold increase in risk.

For several compounds not only were overall trends significant but individual quartiles had 95% confidence limits with a lower limit above 1 (indicated in the graph above by the asterisks).

The associations for dieldrin and heptachlor epoxide were also seen in the analyses when they were paired with other contaminants. Trends for the other contaminants, however, disappeared in runs with paired pesticides. One interpretation is that the correlations among the pesticides that were evident were making it appear that some pesticides were associated with increased risk, until those correlations were factored into consideration.

Why is this important?

This study offers some of the strongest epidemiological evidence to date linking persistent organic pollutants to NHL. Its strengths derive from its sample size and the direct measurements of contamination levels in fatty tissues. Most previous studies have asked participants about their exposure, or estimated exposure based on occupation, rather than measuring it directly.

As Quintana et al. acknowledge, however, their work also has several limitations, the most important of which may be the fact that chemical residue analysis was conducted over 2 decades ago, when limits of detection were much higher. In addition, they did not distinguish between several types of NHL. If associations between organochlorines and NHL vary among NHL types, then the approach used here may underestimate the strength of the associations.

Other studies have noted associations between PCBs and NHL, for example research published by Rothman et al. in 1997. Quintana et al. suggest their current study may have been biased against finding any association between NHL and PCBs, because data for controls were collected slightly earlier than data for cases during a period in which PCB levels were declining. Some values for PCB levels were also missing and this may have further biased the results.

Because of regulatory steps taken by many countries, heptachlor epoxide and dieldrin contamination levels are decreasing, not just in the environment but also in human tissues. Consistent with those declines and the findings reported here by Quintana et al., NHL incidence rates have begun to drop, at least in the US and northern Europe. Implementation of the Stockholm Convention on Persistent Organic Pollutants should ensure that body burdens of these contaminants continue to fall. If the epidemiological links here are indeed causal, then declines in NHL rates should continue.