Plasma concentrations of lipophilic persistent organic pollutants and glucose homeostasis in youth populations

Abstract

Background: Exposure to lipophilic persistent organic pollutants (POPs) is ubiquitous. POPs are metabolic disrupting chemicals and are potentially diabetogenic.

Methods: Using a multi-cohort study including overweight adolescents from the Study of Latino Adolescents at Risk (SOLAR, N = 301, 2001-2012) and young adults from the Southern California Children's Health Study (CHS, N = 135, 2014-2018), we examined associations of POPs and risk factors for type 2 diabetes. SOLAR participants underwent annual visits for a median of 2.2 years and CHS participants performed a single visit, during which a 2-h oral glucose tolerance test was performed. Linear mixed models were used to examine associations between plasma concentrations of POPs [4,4'-dichlorodiphenyldichloroethylene (4,4'-DDE), hexachlorobenzene (HCB), PCBs-153, 138, 118, 180 and PBDEs-154, 153, 100, 85, 47] and changes in glucose homeostasis across age and pubertal stage.

Results: In SOLAR, exposure to HCB, PCB-118, and PBDE-153 was associated with dysregulated glucose metabolism. For example, each two-fold increase in HCB was associated with approximately 2 mg/dL higher glucose concentrations at 30 min (p = 0.001), 45 min (p = 0.0006), and 60 min (p = 0.03) post glucose challenge. Compared to individuals with low levels of PCB-118, individuals with high levels exhibited a 4.7 mg/dL (p = 0.02) higher glucose concentration at 15 min and a 3.6 mg/dL (p = 0.01) higher glucose concentration at 30 min. The effects observed with exposure to organochlorine compounds were independent of pubertal stages. PBDE-153 was associated with the development of dysregulated glucose metabolism beginning in late puberty. At Tanner stage 4, exposure to PBDE-153 was associated with a 12.7 mg/dL higher 60-min glucose concentration (p = 0.009) and a 16.1 mg*dl^-1*hr^-1 higher glucose AUC (p = 0.01). These associations persisted at Tanner 5. In CHS, PBDE-153 and total PBDE were associated with similar increases in glucose concentrations.

Conclusion: Our results suggest that childhood exposure to lipophilic POPs is associated with dysregulated glucose metabolism.

Keywords: Flame retardants; Obesogens; Organochlorines; Pesticides; Polybrominated compounds; Type II diabetes.

Figure 1.

Correlation of lipophilic compounds in the SOLAR and CHS cohorts. Correlations between two continuous variables were calculated using the Pearson correlation coefficient; correlations between two binary variables were calculated with polychoric correlation coefficients; and correlations between a continuous and binomial variable were calculated with polyserial correlation coefficients. * indicates chemicals detected in less than 75% of samples and which were analyzed as binary (detect/non-detect); ** indicates correlations which were statistically significant based upon a p<0.05.

Figure 2.

Point estimates and 95% confidence intervals for associations of (A) Hexachlorobenzene (HCB); (B) polybrominated diphenyl ether (PBDE)-153, and (C) Total PBDEs with risk factors for type 2 diabetes in SOLAR and CHS. Associations of HCB with glucose concentrations during the Oral Glucose Tolerance Test were not time dependent, so panel A represents the change in glucose concentration and glucose AUC per doubling of HCB across all developmental stages for SOLAR. Panels B and C represents the difference in outcome for (B) detected vs. non-detected levels of PBDE-153 and (C) the 80^th vs. 20^th percentile of total PBDE exposure across developmental stages. * indicates chemicals detected in less than 75% of samples and which were analyzed as binary (detect/non-detect). Notes: OGTT Time point 0 corresponds to fasting glucose; Glu: Glucose; AUC: Area under the curve; ISI Matsuda: Matsuda Insulin Sensitivity Index.

Figure 3.

Associations of a mixture of eleven lipophilic POPs with glucose responses to an oral glucose tolerance test in adolescents in late puberty (SOLAR) and young adults (CHS). Effect estimates (ψ) and 95% confidence intervals indicate the change in glucose levels (A) or glucose area under the curve (B) when increasing all lipophilic POPs levels from low to high. For POPs detected in >75% of participants, low vs. high levels were determined by the median; for POPs levels detected in <75% of participants, low vs. high levels were defined as detected versus non-detected. Glu: Glucose; AUC: Area under the curve.