Phthalate Exposure and Allergy in the US Population
Our most consistent finding was for MBzP and allergic symptoms in adults. MBzP was positively associated with current asthma, current wheeze, current hay fever, and current rhinitis as well as, but nonsignificantly, with allergic sensitization. There was some suggestion that the association between MBzP and allergic symptoms was driven by allergic sensitization, but the evidence was not strong. MBzP is the primary metabolite of BBzP, an HMW phthalate used in plastics and other materials in the home. In a population-based case–control study conducted from 1997 to 2000, Jaakkola et al. (2006) reported that plastic wall materials in the home were associated with incident asthma among 1,453 Finnish adults. There is increasing evidence that MBzP may be associated with allergic outcomes. Prenatal urinary MBzP levels were associated with the development of eczema by 5 years of age in 407 children in a birth cohort study (Just et al. 2012b) and with increased airway inflammation as measured by exhaled nitric oxide (Just et al. 2012a). In a cross-sectional study of 101 Taiwanese children 3–9 years of age, BBzP concentration in house dust was associated with allergic symptoms and asthma (Hsu et al. 2011). Although some phthalates, particularly di-n-butyl phthalate, are included in pharmaceuticals (Hernández-DÃaz et al. 2009), BBzP is not approved for pharmaceutical use and, thus, MBzP in urine is unlikely to be a consequence of the use of allergy or asthma medications.
Much of the mechanistic work to date on phthalates and allergy has focused on DEHP and its ability to modulate responses to allergens (Jaakkola and Knight 2008; Kimber and Dearman 2010). In murine models and in human lung epithelial cells, DEHP, but not BBzP, has been shown to have an adjuvant effect on immune response to allergens (Guo et al. 2012; Koike et al. 2009; Larsen et al. 2007; Nishioka et al. 2012). In a small human study, 16 adults with sensitivity to house dust mites and 16 without sensitivity were exposed to airborne dust containing low or high levels of DEHP. Those exposed to high levels had an attenuated immune response, whereas those exposed at low levels had mucosal inflammation and nonsensitized individuals had no response (Deutschle et al. 2008). In murine models, BBzP enhanced anti-ovalbumin responses at high doses, but not at the lower doses potentially more consistent with human exposures (Dearman et al. 2009). Few studies have evaluated whether BBzP, or its metabolite MBzP, have independent effects on immune responses at levels relevant to human exposure. One study reported that topical administration of BBzP did not stimulate an immune response in mice (Butala et al. 2004). Our results for DEHP and allergic sensitization in adults are consistent with the mechanistic data; however, we have no information on allergen exposure.
LMW phthalates were not positively associated with allergic symptoms or sensitization, except for MEP and allergic sensitization among Mexican Americans. All other race/ethnicity groups had inverse associations between sIgE and MEP. In addition, MEP was also inversely associated with hay fever and allergic sensitization in adults. In a previous NHANES sample (1999–2000), MEP levels among Mexican Americans were not different from non-Hispanic whites and were lower than for non-Hispanic blacks (Silva et al. 2004), suggesting exposure level did not explain the observed difference. Interestingly, in a study of Dominican (67%) and African-American (33%) children 5–9 years of age, Just et al. (2012a) reported that children with higher urinary levels of MEP had higher fractional exhaled nitric oxide. There was limited evidence for differences between Dominican and African-American children for MEP exposure and allergic sensitization in children, but the sample sizes were small. In a Japanese cross-sectional study of 134 residents of 41 dwellings conducted in 2006–2007, DEP was inversely associated with respiratory and allergic symptoms in both children and adults, consistent with our results (Kanazawa et al. 2010). It is possible that both age and race/ethnicity may influence allergic response to MEP, but currently the data are too limited to explore this extensively. MEP is the primary metabolite of DEP, a phthalate primarily used in fragrances (Api 2001). Some fragrances can be 25–50% DEP by volume (Agency for Toxic Substances and Disease Registry 1995). The inverse association with MEP and sensitization among adults could suggest fragrance avoidance by allergen-sensitized individuals. Among Mexican-American adults, we saw a positive association with MEP. Other investigators have noted that Mexican women who used multiple cosmetic and fragrance products had higher levels of MEP than those who did not, consistent with DEP exposure through the use of fragrances (Romero-Franco et al. 2011). The differential findings for Mexican Americans warrant further characterization of their exposures in the future.
Biological markers of phthalate metabolites are often used to assess exposure because of the complexity of evaluating all sources of exposure. Although the use of biological markers is common, these phthalate measures are limited with respect to the time period they represent. The biological half-lives of these metabolites are < 1 day, and studies have shown that temporal variability in these measures limits their usefulness in estimating the associations with long-term exposure (Baird et al. 2010; Hauser et al. 2004; Hoppin et al. 2002). Given the short biological half-lives of phthalates and the relatively nonvarying state of allergen-specific IgE in serum, our findings for allergic sensitization and HMW phthalates should be considered cautiously. These findings might reflect a preference for plastics among those with allergic sensitization because these surfaces are easier to clean and less likely to be reservoirs for dusts and other allergic triggers. In contrast, our findings for allergic symptoms are less challenged by the use of biological markers because symptoms were experienced within the past year, although daily variation in exposure may influence these findings as well.
Our limited findings for children were unexpected given the previous literature. HMW phthalates, particularly DEHP, were associated with current rhinitis, but none of the associations were statistically significant. In a Swedish study, dust levels of BBzP were associated with rhinitis and eczema in children 3–8 years of age, whereas dust levels of DEHP were associated with asthma; there was also some evidence of an association with rhinitis and DEHP in the highest quartile (Bornehag et al. 2004). Recently, Just et al. (2012b) reported that prenatal exposure to MBzP was associated with maternally reported eczema by 5 years of age and that current exposures were associated with higher fractional exhaled nitric oxide (Just et al. 2012a). The HMW phthalate metabolites, MCOP and MCNP [mono(carboxynonyl) phthalate], were associated with current asthma in a cross-sectional study of 623 Norwegian 10-year-old children (Bertelsen et al. 2013); these two metabolites were not associated with respiratory symptoms in children or adults in our sample. In our analysis, there was no evidence of an association of any phthalates with current itchy rash in either children or adults (data not shown). The lack of consistency with other studies may be related to the age of the children included in our analysis because no children < 6 years of age had data on urinary phthalates. Another possible explanation is that the relevant exposures for children are not those that are currently occurring but rather those that occurred earlier, such as prenatally. In addition, our study included a large representative sample of children in the United States, whereas previous studies used smaller, more demographically similar groups. It may be that the diversity in our sample limited our ability to observe effects restricted to subsets of the population.
Phthalate metabolites, although chemically similar, are not highly correlated except for those derived from the same parent compound. To minimize confounding by correlated phthalates, we combined all metabolites from the same chemical (i.e., DEHP) into one summary variable. Given the multiple phthalates and ways to consider allergy and allergic sensitization, we conducted many statistical analyses. Rather than correcting our results for the number of comparisons or limiting our presentation to those chemicals for which we had a priori evidence, we have chosen to present our results uncorrected for multiple comparisons and present the full results in order to provide a more complete picture of the complexity of this research question.
Our previous work has shown poor concordance between self-reported respiratory symptoms and allergic sensitization (Hoppin et al. 2011). To evaluate if our results for respiratory symptoms were driven by allergic sensitization, we ran additional models for those who were both allergen sensitized and had the allergic symptom. These results suggested some interplay of allergic sensitization and symptoms in response to phthalates, but the evidence was inconclusive.
Discussion
Our most consistent finding was for MBzP and allergic symptoms in adults. MBzP was positively associated with current asthma, current wheeze, current hay fever, and current rhinitis as well as, but nonsignificantly, with allergic sensitization. There was some suggestion that the association between MBzP and allergic symptoms was driven by allergic sensitization, but the evidence was not strong. MBzP is the primary metabolite of BBzP, an HMW phthalate used in plastics and other materials in the home. In a population-based case–control study conducted from 1997 to 2000, Jaakkola et al. (2006) reported that plastic wall materials in the home were associated with incident asthma among 1,453 Finnish adults. There is increasing evidence that MBzP may be associated with allergic outcomes. Prenatal urinary MBzP levels were associated with the development of eczema by 5 years of age in 407 children in a birth cohort study (Just et al. 2012b) and with increased airway inflammation as measured by exhaled nitric oxide (Just et al. 2012a). In a cross-sectional study of 101 Taiwanese children 3–9 years of age, BBzP concentration in house dust was associated with allergic symptoms and asthma (Hsu et al. 2011). Although some phthalates, particularly di-n-butyl phthalate, are included in pharmaceuticals (Hernández-DÃaz et al. 2009), BBzP is not approved for pharmaceutical use and, thus, MBzP in urine is unlikely to be a consequence of the use of allergy or asthma medications.
Much of the mechanistic work to date on phthalates and allergy has focused on DEHP and its ability to modulate responses to allergens (Jaakkola and Knight 2008; Kimber and Dearman 2010). In murine models and in human lung epithelial cells, DEHP, but not BBzP, has been shown to have an adjuvant effect on immune response to allergens (Guo et al. 2012; Koike et al. 2009; Larsen et al. 2007; Nishioka et al. 2012). In a small human study, 16 adults with sensitivity to house dust mites and 16 without sensitivity were exposed to airborne dust containing low or high levels of DEHP. Those exposed to high levels had an attenuated immune response, whereas those exposed at low levels had mucosal inflammation and nonsensitized individuals had no response (Deutschle et al. 2008). In murine models, BBzP enhanced anti-ovalbumin responses at high doses, but not at the lower doses potentially more consistent with human exposures (Dearman et al. 2009). Few studies have evaluated whether BBzP, or its metabolite MBzP, have independent effects on immune responses at levels relevant to human exposure. One study reported that topical administration of BBzP did not stimulate an immune response in mice (Butala et al. 2004). Our results for DEHP and allergic sensitization in adults are consistent with the mechanistic data; however, we have no information on allergen exposure.
LMW phthalates were not positively associated with allergic symptoms or sensitization, except for MEP and allergic sensitization among Mexican Americans. All other race/ethnicity groups had inverse associations between sIgE and MEP. In addition, MEP was also inversely associated with hay fever and allergic sensitization in adults. In a previous NHANES sample (1999–2000), MEP levels among Mexican Americans were not different from non-Hispanic whites and were lower than for non-Hispanic blacks (Silva et al. 2004), suggesting exposure level did not explain the observed difference. Interestingly, in a study of Dominican (67%) and African-American (33%) children 5–9 years of age, Just et al. (2012a) reported that children with higher urinary levels of MEP had higher fractional exhaled nitric oxide. There was limited evidence for differences between Dominican and African-American children for MEP exposure and allergic sensitization in children, but the sample sizes were small. In a Japanese cross-sectional study of 134 residents of 41 dwellings conducted in 2006–2007, DEP was inversely associated with respiratory and allergic symptoms in both children and adults, consistent with our results (Kanazawa et al. 2010). It is possible that both age and race/ethnicity may influence allergic response to MEP, but currently the data are too limited to explore this extensively. MEP is the primary metabolite of DEP, a phthalate primarily used in fragrances (Api 2001). Some fragrances can be 25–50% DEP by volume (Agency for Toxic Substances and Disease Registry 1995). The inverse association with MEP and sensitization among adults could suggest fragrance avoidance by allergen-sensitized individuals. Among Mexican-American adults, we saw a positive association with MEP. Other investigators have noted that Mexican women who used multiple cosmetic and fragrance products had higher levels of MEP than those who did not, consistent with DEP exposure through the use of fragrances (Romero-Franco et al. 2011). The differential findings for Mexican Americans warrant further characterization of their exposures in the future.
Biological markers of phthalate metabolites are often used to assess exposure because of the complexity of evaluating all sources of exposure. Although the use of biological markers is common, these phthalate measures are limited with respect to the time period they represent. The biological half-lives of these metabolites are < 1 day, and studies have shown that temporal variability in these measures limits their usefulness in estimating the associations with long-term exposure (Baird et al. 2010; Hauser et al. 2004; Hoppin et al. 2002). Given the short biological half-lives of phthalates and the relatively nonvarying state of allergen-specific IgE in serum, our findings for allergic sensitization and HMW phthalates should be considered cautiously. These findings might reflect a preference for plastics among those with allergic sensitization because these surfaces are easier to clean and less likely to be reservoirs for dusts and other allergic triggers. In contrast, our findings for allergic symptoms are less challenged by the use of biological markers because symptoms were experienced within the past year, although daily variation in exposure may influence these findings as well.
Our limited findings for children were unexpected given the previous literature. HMW phthalates, particularly DEHP, were associated with current rhinitis, but none of the associations were statistically significant. In a Swedish study, dust levels of BBzP were associated with rhinitis and eczema in children 3–8 years of age, whereas dust levels of DEHP were associated with asthma; there was also some evidence of an association with rhinitis and DEHP in the highest quartile (Bornehag et al. 2004). Recently, Just et al. (2012b) reported that prenatal exposure to MBzP was associated with maternally reported eczema by 5 years of age and that current exposures were associated with higher fractional exhaled nitric oxide (Just et al. 2012a). The HMW phthalate metabolites, MCOP and MCNP [mono(carboxynonyl) phthalate], were associated with current asthma in a cross-sectional study of 623 Norwegian 10-year-old children (Bertelsen et al. 2013); these two metabolites were not associated with respiratory symptoms in children or adults in our sample. In our analysis, there was no evidence of an association of any phthalates with current itchy rash in either children or adults (data not shown). The lack of consistency with other studies may be related to the age of the children included in our analysis because no children < 6 years of age had data on urinary phthalates. Another possible explanation is that the relevant exposures for children are not those that are currently occurring but rather those that occurred earlier, such as prenatally. In addition, our study included a large representative sample of children in the United States, whereas previous studies used smaller, more demographically similar groups. It may be that the diversity in our sample limited our ability to observe effects restricted to subsets of the population.
Phthalate metabolites, although chemically similar, are not highly correlated except for those derived from the same parent compound. To minimize confounding by correlated phthalates, we combined all metabolites from the same chemical (i.e., DEHP) into one summary variable. Given the multiple phthalates and ways to consider allergy and allergic sensitization, we conducted many statistical analyses. Rather than correcting our results for the number of comparisons or limiting our presentation to those chemicals for which we had a priori evidence, we have chosen to present our results uncorrected for multiple comparisons and present the full results in order to provide a more complete picture of the complexity of this research question.
Our previous work has shown poor concordance between self-reported respiratory symptoms and allergic sensitization (Hoppin et al. 2011). To evaluate if our results for respiratory symptoms were driven by allergic sensitization, we ran additional models for those who were both allergen sensitized and had the allergic symptom. These results suggested some interplay of allergic sensitization and symptoms in response to phthalates, but the evidence was inconclusive.