Persistent organic pollutants in Norwegian men from 1979 to 2007: intraindividual changes, age-period-cohort effects, and model predictions

Abstract

Background: Longitudinal monitoring studies of persistent organic pollutants (POPs) in human populations are important to better understand changes with time and age, and for future predictions.

Objectives: We sought to describe serum POP time trends on an individual level, investigate age-period-cohort effects, and compare predicted polychlorinated biphenyl (PCB) concentrations to measured values.

Methods: Serum was sampled in 1979, 1986, 1994, 2001, and 2007 from a cohort of 53 men in Northern Norway and analyzed for 41 POPs. Time period, age, and birth cohort effects were assessed by graphical analyses and mixed-effect models. We derived the predicted concentrations of four PCBs for each sampling year using the CoZMoMAN model.

Results: The median decreases in summed serum POP concentrations (lipid-adjusted) in 1986, 1994, 2001, and 2007 relative to 1979 were -22%, -52%, -54%, and -68%, respectively. We observed substantial declines in all POP groups with the exception of chlordanes. Time period (reflected by sampling year) was the strongest descriptor of changes in PCB-153 concentrations. Predicted PCB-153 concentrations were consistent with measured concentrations in the study population.

Conclusions: Our results suggest substantial intraindividual declines in serum concentrations of legacy POPs from 1979 to 2007 in men from Northern Norway. These changes are consistent with reduced environmental exposure during these 30 years and highlight the relation between historic emissions and POP concentrations measured in humans. Observed data and interpretations are supported by estimates from the CoZMoMAN emission-based model. A longitudinal decrease in concentrations with age was evident for all birth cohorts. Overall, our findings support the relevance of age-period-cohort effects to human biomonitoring of environmental contaminants.

Figure 1

Concentrations (ng/g lipid, log_e scale) of selected POPs analyzed in repeated serum samples of men (n = 51, 51, 45, 48, and 52 in 1979, 1986, 1994, 2001, and 2007, respectively) from Northern Norway. p,p´-DDE, 1,1-dichloro-2,2-bis(p-chlorophenyl)ethylene. Parlar 50 represents toxaphenes, and trans-nonachlor the chlordanes. Boxes represent the 25th–75th percentiles, horizontal lines represent the median, whiskers indicate 1.5 times the length of the interquartile range above and below the 75th and 25th percentiles, respectively, and outliers are represented as data points. *p < 0.05, and **p < 0.001 for comparisons between pairs of consecutive sampling years.

Figure 2

Individual trend lines for PCB-153 serum concentrations (ng/g lipid) measured in 1979, 1986, 1994, 2001, and 2007 in 53 men from Northern Norway, according to birth year quartile. (A) 1925–1934, (B) 1934–1936, (C) 1936–1941, and (D) 1941–1950.

Figure 3

Relative contributions of individual POPs that accounted for > 1% of summed POPs in 1979 in serum from men (n = 51, 51, 45, 48, and 52 in 1979, 1986, 1994, 2001, and 2007, respectively) in Northern Norway.

Figure 4

APC plots showing observed (A–C) and predicted serum PCB-153 concentrations (D–F, using the CoZMoMAN model assuming higher fish consumption in earlier birth cohorts). A and D show longitudinal variation among birth cohorts according to sampling period, B and E show longitudinal variation among birth cohorts according to age quartile (Q), and C and F cross-sectional variation among sampling periods according to age quartile. Data points indicate ng/g lipid-adjusted concentrations (median concentrations are displayed in A–C). Differences between birth cohorts in A were significant in 1986 and 2007 (Kruskal–Wallis rank sum test, p < 0.05).