Our Stolen Futurea book by Theo Colborn, Dianne Dumanoski, and John Peterson Myers
 
 

 

 

Posted 20 November 2006. Research published July 2006.

Lee, D-H, I-K Lee, K Song, M Steffes, W Toscano, BA Baker, and DR Jacobs. 2006. A Strong Dose-Response Relation Between Serum Concentrations of Persistent Organic Pollutants and Diabetes. Results from the National Health and Examination Survey 1999–2002. Diabetes Care 29:1638-1644.


Context
What did they do?
What did they find?
What does it mean?

Latest news about diabetes

More news about
 

According to the US Centers for Disease Control, from 1980 through 2004, the number of Americans with diabetes more than doubled (from 5.8 million to 14.7 million). The chances that an American child will become diabetic are now 1 in 3. The odds for a Latinos in the US and indigenous peoples around the Pacific are even worse, 1 in 2.

Prevailing wisdom blames Western lifestyles and diet. An alternative explanation-- contaminants interfering with glucose and insulin metabolism-- has begun to gain traction, based on studies in the lab with cells and mice, and on epidemiological research with people. These explanations are not mutually exclusive: both could be at work at the same time.

New research by Lee et al., summarized here, adds substantial weight to the hypothesis that contaminants are involved. They find a strong dose response relationship between type II diabetes risk and body burden of 6 persistent organic pollutants (POPs). Five of the 6 have highly significant associations when examined singly. The association is especially strong between diabetes risk and an estimate of the summed exposure to all 6 POPs studied simultaneously.

 

What did they do? Lee et al. analyzed data obtained by the Centers for Disease Control in the National Health and Examination Survey (NHANES; 1999–2002). This periodic survey assesses the health of the American public. The sampling protocol is carefully designed to obtain representative data. In this analysis, Lee et al. assessed the statistical relationship between risk of type II diabetes and 6 persistent organic pollutants: One PCB (hexachlorobiphenyl), two
dioxins (heptadioxin and OCDdioxin) two pesticides (oxychlordane and trans-nonachlor) and a pesticide metabolite (DDE, a metabolite of DDT).

They selected these contaminants because they were detectable in over 80% of participants. Total sample size in the study was 2,016. The diagnosis of diabetes was confirmed by medical interview.

Each organochlorine was assessed individually: Individuals with contamination level beneath the limit of detection for a given contaminant were used as the reference group ('control') for calculating an odds-ratio. The remaining individuals, all with detectable levels, were divided into 5 groups based on percentile exposure: up to 25th; up to 50th, up to 75th, up to 90th, and above 90th percentile.

To examine the association for the combination of POPs, each person studied was assigned a score of 0 to 5 for each contaminant based on which category of exposure they were in (reference, 25th, 50th...).. The sum of the scores (minimum 0, maximum 30) was then used as an index of total POPs exposure. People were separated into groups based on the sum of the scores (from reference to 25th, 50th, etc.) and then an odds ratio estimating the relative risk of type II diabetes was calculated for each group.

What did they find?

In general, older people had higher levels of individual contaminants than younger. Men tended to have lower concentrations. For all but one contaminant (PCB153), Hispanics tended to have higher levels as did poorer people.

Among the 2,016 people in the study, 217 had type II diabetes.

Five of the 6 POPs demonstrated a strong trend of increasing risk of diabetes with increasing body burden of POPs.

Table of probabilities

Because all people had detectable amounds of DDE, Lee et al. used the 2nd exposure category as reference group for this contaminant. Red line is an odds-ratio of 1.

 
POPs and diabetes risk

Levels of trans-Nonachlor showed the most striking individual relationship with diabetes risk, with the odds ratio for the highest exposure group rising to 11.8 (95% confidence limits ran from 4.4 to 31.3).

Overall, the lower boundary of thirteen of 30 calculated confidence limits for all contaminants was greater than 1. For people in the two highest exposure groups, 9 of 13 of the lower estimate of the 95% confidence intervals were greater than 1.

When they analyzed the index of simultaneous exposure to all 6 POPs, they first observed that no one in the survey had undetectable levels of every contaminant. This lead them to use the 2nd lowest exposure group, (up to the 25th percentile), as the reference for calculating the odds ratio for higher exposure groups.

Summed POPs and diabetes risk  

Compared to people in the lowest exposure category (1), people in the highest were almost 38 times more likely to have diabetes (graph to left). All odds ratios calculated for categories 2 through 5 were significant.

The trend of increasing risk with increasing exposure was also highly significant (p < 0.001)

Red line is set at OR=1.0. Vertical black lines show 95% confidence limits. Upper limit shown numerically.

Because only 2 people in the lowest exposure group had diabetes, the estimates of the confidence limits were quite broad. For example, for group 5 the limits ranged from 7.8 to 182. In part this was due to the fact that only 2 out of 463 people in the the lowest group had diabetes (compared to 63 out of 246 in group 5). Lee et al. therefore provided a separate estimate of the odds ratios using group 2 as the reference group. This led to estimates of 1.1, 2.7 and 2.7 for groups 3, 4 and 5, respectively. Confidence limits for groups 4 and 5 did not include OR=1.

Lee et al. report that there was "no association between obesity and diabetes among subjects with nondetectable levels of POPs."

What does it mean? Numerous experimental studies have proven links in animals and cells between contaminants, including persistent organic pollutants, and changes in insulin and glucose metabolism associated with diabetes. Some prior epidemiological research has found associations, for example, between diabetes and dioxin.

The odds ratios found by Lee et al. are nonetheless strikingly high. They caution that epidemiological studies like theirs can't be used to prove causation, but "we think that the relation between POPs and diabetes observed in this study may be causal for several reasons." They point out that there are high correlations among the levels of exposure to POPs in this study and that the actual causal factor may be another POP that was not measured. They also highlight the unexpected finding that in people in the study with nondetectable levels of POPs there was no relationship between obesity and diabetes.

The good news is that POPs levels in people have begun to decline, at least for some contaminants. Indeed, Swedish epidemiologists have suggested that a recent decline in POPs in that country may be the cause of recent drop there in cases of non-Hodgkin's lymphoma.

Data on trends in POPs use and contamination would suggest that people with the heaviest burdens will have been born during the decade that followed steps taken, beginning in the mid-1970s, to reduce environmental exposures to POPs. This is the cohort that is now in their reproductive years, raising questions about how their own exposures may be affecting insulin metabolism in their offspring. Steps taken in under the auspices of the United Nations Stockholm Convention on Persistent Organic Pollutants should lead to lower levels in the future.

 
   
   

 

 

 

OSF Home
 About this website
Newest
Book Basics
  Synopsis & excerpts
  The bottom line
  Key points
  The big challenge
  Chemicals implicated
  The controversy
  Recommendations
New Science
  Broad trends
  Basic mechanisms
  Brain & behavior
  Disease resistance
  Human impacts
  Low dose effects
  Mixtures and synergy
  Ubiquity of exposure
  Natural vs. synthetic
  New exposures
  Reproduction
  Wildlife impacts
Recent Important    Results
Consensus
News/Opinion
Myths vs. Reality
Useful Links
Important Events
Important Books
Other Sources
Other Languages
About the Authors
 
Talk to us: email