This study links local marine species' consumption with blood levels of TDCs and THs, and demographic profile of participants (age, sex and community). We found that residents from the study communities consumed local cod more than any other species, and that participants that were older (> 50 years), male and/or residents of Burin consumed more cod than other demographic groups. There were some obvious trends in TDC concentrations by age, sex and community; older participants (> 50 years) had greater plasma concentrations of PBB-153, PCBs and p,p’-DDE, and males had higher concentrations of all TDCs than females. Participants from Burin in southern NL generally had greater concentrations of PBB-153 and PBDEs, while participants from NWV in northeastern NL generally had greater concentration of PCBs and p,p’-DDE. Frequency of local cod consumption was also found to be positively associated with several PCB congeners, p,p’-DDE and ∑14TDCs. Seasonal effects of seafood consumption on plasma TDC concentrations were also observed; several TDCs were positively associated with higher consumption of local seafood in fall, winter and spring. We did not find any relationship between any TDCs and THs, however multivariate regression analysis revealed that local cod consumption frequency was significantly associated with several PCB congeners, ∑8PCBs, and p,p’-DDE.
For THs, we found that TSH levels were significantly higher in NWV (2.20 ± 0.19 mIU/L) than in Burin (1.99 ± 0.13 mIU/L); however, both results were still below the upper limit for the lab’s normal reference range (4.94 mIU/L). An upper limit of normal (euthyroid) concentration of 2.5 mIU/L has been suggested as a guideline for diagnosing subclinical hypothyroidism, as 95% of humans with no known thyroid conditions will have TSH values below this level; individuals with values above this limit are more likely to have underlying causes artificially elevating their TSH levels (Wartofsky and Dickey, 2005). In NWV, approximately 31% (n = 8) of participants had TSH values that were above the subclinical hypothyroidism limit cutoff of 2.5 mIU/L, while the proportion in Burin was lower at approximately 15% (n = 7), excluding participants with known hypothyroidism. Therefore, it seems that being above the subclinical hypothyroidism cutoff point was more common in NWV than in Burin, even if the average TSH concentrations above the subclinical thyroid threshold were not different. This may be explained by the higher proportion (78%) of participants > 50 years of age in the NWV group compared with the proportion that was > 50 years of age (69%) in the Burin group. The skewed age distribution of our participants in this study may likely have impacted the thyroid hormones results, as age has been shown to be an important factor in TSH levels, and age is associated with an increased proportion of individuals with serum TSH concentrations above the 2.5 mIU/L euthyroid limit (Surks and Hollowell, 2007). Our finding of higher FT3 in males than in females aligns with other studies (Strich et al., 2017).
PBDE concentrations were higher in younger participants (< 50 years) while PBB-153, PCBs, and p,p’-DDE were generally higher in older participants (> 50 years), regardless of community. This may reflect the sources of these chemicals and the length of time they have been in production. PCBs have been in production and use since the 1930s, and DDT (the parent compound of p,p’-DDE) has been widely used as an insecticide since 1939. PCB ad DDT were both restricted and/or banned by the Stockholm Convention in 2004 as part of the “Dirty Dozen” initial chemicals that were entered under the treaty. In Canada, DDT was phased out in the 1970s and completely banned by 1990; PCBs were banned from import, manufacture, and sale in 1977. However, due to their bioaccumulative nature and long half-lives, they have persisted in the environment for decades and tend to be found in higher concentrations in older individuals. It is possible that older participants had had a long time to accumulate the legacy chemicals (PCBs and DDT), which peaked in use for much of their life span, but were phased out before some of the youngest participants were born. Age has also been shown to be a predictor for PCB and p,p’-DDE concentration in other studies (Laden et al., 1999). On the other hand, PBBs (in production and use from the 1970s onwards) do not follow this trend, being “newer” flame retardants. However, other studies have observed a positive association between PBB-153 and age, corroborating with our findings of higher PBB-153 in older participants (Lim et al. 2008). PBDEs were manufactured from the1960s onwards, and reached peak production and usage in the 1990s. They were banned under the Stockholm Convention in 2009, long after PCBs and DDT, but many PBDEs are still present in various older consumer products (i.e., automobiles, furniture, carpet, and electronic goods) to this day. PBDEs have been found to be inversely related to age, with younger populations having higher concentrations of these chemicals, likely due to a high degree of TDC exposure that occurs in infants and younger children from breastmilk and hand to mouth activity (Garí and Grimalt, 2013). These results may indicate an age-related cohort effect, particularly for PBDEs, which were higher in younger participants who would have been infants or young children around the peak of PBDE production and use (1980-2000s). Since these chemicals have now been banned, this age group (20–50 years of age) would have had a unique exposure, which would be different from older (> 50 years) or younger (< 20 years) age groups. This possible age-cohort effect of plasma TDC concentrations requires a larger-scale study to explore this effect fully.
Our study found that males had higher plasma concentrations of TDCs than females, and higher consumption of seafood among males (than females) could be the reason. Research has found higher seafood consumption by males and a higher estimated daily intake of DDT and PBDEs via seafood consumption (Guo et al., 2010). Men have also been shown to have a higher intake of PCBs/kg body weight than women from consuming contaminated fish in Norway (Knutsen et al., 2008).
We found that relatively newer TDCs, such as PBBs and PBDEs, were significantly higher in Burin participants. In comparison, legacy TDCs such as PCBs and p,p’-DDE concentrations were higher in NWV participants. PBDEs and PBB were banned from production and use much more recently than PCBs and DDT. Therefore, many products containing PBBs and PBDEs may be still found in regular consumer use or are transitioning into urban effluents and landfills. Abbasi et al. (2014) found that the largest PBDEs flow from consumer products to their waste phase (deposition of EDC-containing products into landfills, recycling centers, wastewater treatment plants) was between 2005 and 2008. PBDEs and PBBs have been documented up and down the SLR in sediments (Pelletier and Rondeau, 2013), fish (Houde et al., 2014), birds (Gentes et al., 2012), and marine mammals (Simond et al., 2017). Therefore, it is plausible that higher concentrations of these TDCs in participants from Burin (south coast of NL) results from contamination from upstream in the SLR making its way into the marine food chain, where they then could be ingested by humans through local seafood consumption.
PCBs and DDT have been banned from manufacture and use for longer than PBB and PBDEs and thus there are fewer of these chemicals transitioning from consumer products into the waste phase. They have a global distribution, with PCBs being documented in seafood species such as cod in different parts of the North Atlantic Ocean (Karl et al., 2016). PCBs and DDT/p,p’-DDE have also been found in Arctic ice (Gregor et al., 1995) and high trophic level Arctic organisms such as the polar bear (Letcher et al. 1995), indicating the widespread presence of these chemicals, both geographically and by trophic level. It appears that climate change may be causing TDCs locked in snow/ice to be released upon melting; these contaminants could then travel south via the Labrador Current into marine ecosystems. Residents of NWV (northeastern NL) may be exposed to these TDCs by consuming marine species from these contaminated ecosystems, whereas residents of Burin (southern NL) are mostly exposed to contamination from the SLR.
It is difficult to discern why we did not see any association between TDCs and THs in the rural NL population while other studies have found significant associations. The major limitation of the study was small sample size (n = 80). However, many smaller-scale research projects have successfully explored associations between TDCs and THs. For example, Byrne et al. (2018), who had a sample size of n = 85, explored PBDEs and thyroid hormones in a remote Indigenous population in coastal Alaska, where they found PBDEs in this isolated population and an association with thyroid hormones (TSH, FT3, and T3). Other examples of successful pilot/exploratory studies with small sample sizes include English et al. (2017), who looked at PBDEs in feces from Australian children (n = 61 but only n = 46 were testable), Li et al. (2017), who conducted a pilot study of PCBs in the breast milk of mothers from a recycling site (n = 46), Lin et al. (2011) who looked at PBDEs and thyroid hormones in cord blood (n = 54), Zota et al., (2011) who measured PBDEs and thyroid function in pregnant women (n = 25), and Zhang et al. (2010) who investigated PCBs and PBDEs and thyroid hormone homeostasis in e-waste recycling workers (n = 50). A recent meta-analysis of correlations between PBDEs and thyroid hormones (Zhao et al., 2015) found that PBDE effects on THs (TSH and total T4) varied and were dependent on exposure and serum concentrations. Lower PBDE concentrations (< 30 ng/g lipid) were negatively correlated with THs and higher PBDE concentrations (> 100 ng/g lipid) were positively correlated with THs, while mid-range (30–100 ng/g lipid) PBDE levels did not correlate with a change in THs. This non-monotonic dose-response relationship between PBDEs and THs is what makes determining the health effects so difficult. Some of our PBDEs had ranges that would put them in this 30–100 ng/g lipid “no effect” range, which may be contributing to the lack of significant relationships between TDCs and THs in our study population.
Additionally, there were low TDC levels in our plasma samples compared with a past sampling of the general Canadian population. Before this study was conducted, the prevalence of TDCs in the NL population was unknown. However, the Canadian Health Measures Survey (CHMS) collects physical samples (i.e. blood) for biomonitoring of contaminant exposure in the Canadian population. Data on TDCs in blood samples from CHMS collection cycle 1 (2007–2009; Haines et al., 2017) were compared with 75th percentile and 95% confidence intervals [CI] from our study (Table 10). There is a 9-year gap between the most recent blood contaminant exposure data from the CHMS (2009) and the current study (2018), and TDC concentrations in the Canadian population may have changed during this time. Additionally, CHMS data are from a much wider age range of participants (3–79 years) than the current study (19–75 + years), which may affect the comparison between the two studies as higher PBDE and PCB serum concentrations have been reported in toddlers and children than in adults (Toms et al., 2009; Marek et al., 2013). Cycle 1 CHMS blood samples were analyzed recently (2016 to 2017) and therefore reflect current instrumentation/analytic capabilities. There is a clear trend in the present study of finding much lower plasma PCB, PBDE, PBB, and p,p’-DDE concentrations than the Canada-wide CHMS from the previous decade.
Table 10
Comparison of TDC concentrations (75th percentile [95% CI]) between the present study and the Canadian Health Measures Survey (CHMS 2007–2009) from Haines et al. (2017).
EDC congener | NL concentrations P75 (95% CI) | CHMS concentrations P75 (95% CI) |
PBB-153 | 0.73 (0.35–0.67) | <MDL |
PBDE-47 | 5.49 (3.81–6.30) | 120 (100–140) |
PBDE-99 | 1.43 (0.99–1.51) | 21 (< MDL-26) |
PBDE-100 | 1.39 (0.71–2.36) | 21 (< MDL-27) |
PBDE-153 | 12.28 (7.35–14.05) | 33 (27–39) |
PCB-99 | 1.43 (0.99–1.51) | <MDL |
PCB-105 | 0.84 (0.50–0.81) | <MDL |
PCB-118 | 4.56 (2.77–4.69) | 50 (41–58) |
PCB-128 | 0.08 (0.02–0.11) | <MDL |
PCB-138 | 14.73 (7.63–11.32) | 110 (97–130) |
PCB-153 | 31.63 (17.09–25.99) | 210 (180–240) |
PCB-156 | 4.40 (2.29–3.45) | 32 (29–36) |
PCB-170 | 10.00 (5.39–8.21) | 59 (52–65) |
PCB-180 | 34.46 (17.55–27.56) | 190 (170–220) |
PCB-183 | 2.36 (1.25–1.86) | 19 (15–22) |
p,p’-DDE | 117.30 (69.55-106.65) | 1700 (1300–2100) |
MDL = Method detection limit, concentrations in ng/g lipid |
In conclusion, the residents of Newfoundland showed exposure to environmental TDCs via local seafood consumption. A larger population based study is required for total exposure analysis and adverse outcomes.