Naphthalene Biomarkers and Relationship with Hemoglobin and Hematocrit in White, Black, and Hispanic Adults: Results from the 2003–2004 National Health and Nutrition Examination Survey

Abstract

Naphthalene is an important contaminant in indoor and outdoor air. Acute overexposure can have toxic effects, resulting in hemolysis. There have been no studies evaluating the impact of environmental exposure on red blood cell indices. We examined 1- and 2-hydroxynaphthalene urinary metabolites (NAP1 and NAP2) in non-Hispanic White, non-Hispanic Black, and Mexican-American adults in the USA and their relationship with hemoglobin (Hb) and hematocrit (HCT). Using the 2003–2004 National Health and Nutrition Examination Survey data, weighted generalized linear regression analyses were used to examine the association between Hb (in grams per deciliter) and HCT (in percent) with NAP1 and NAP2 (per 100,000 ng/L). Beta coefficients ± SE are reported. NAP1 and NAP2 were highest in non-Hispanic Blacks, followed by non-Hispanic Whites, and lowest in Mexican-American adults. There was a positive association between NAP1 and Hb (0.39 ± 0.11, p = 0.0034) and HCT (1.14 ± 0.28, p = 0.0009) after adjusting for age, gender, race, education, and smoking. Stratified analysis by smoking showed similar results with the association being stronger for smokers (Hb 0.63 ± 0.23, p = 0.02; HCT 1.43 ± 0.79, p = 0.09) than nonsmokers (Hb 0.34 ± 0.14, p = 0.03; HCT 1.08 ± 0.42, p = 0.02). The association was also stronger for non-Hispanic blacks (Hb 0.54 ± 0.20, p = 0.02; HCT 1.43 ± 0.55, p = 0.02) than for non-Hispanic whites (Hb 0.37 ± 0.18, p = 0.06; HCT 1.20 ± 0.51, p = 0.03) and was not significant for Mexican-Americans (Hb 0.30 ± 1.7, p = 0.10; HCT 0.99 ± 0.52, p = 0.08). NAP2 was not significantly associated with Hb or HCT. The observed disparity in NAP1 and NAP2 levels by race/ethnicity is consistent with published literature. The origin of these differences in exposure is unclear but may reflect differences in environmental exposure as well as genetic susceptibility. The positive association between NAP1 with HCT and Hb is an unexpected finding. Further research is needed to understand the possible biological mechanisms or other explanations for this association.

Introduction

Naphthalene is an important contaminant in indoor and outdoor air [1]. It is a fairly volatile compound with a vapor pressure of 0.087 mmHg at 25 °C. As a bicyclic aromatic compound, it is among the most volatile of the polycyclic aromatic hydrocarbons (PAHs). It is among the most commonly identified contaminants at National Priority List (Superfund) sites in the USA [2].

Naphthalene has been the subject of many studies of indoor and outdoor air quality. Several important environmental sources have been identified for naphthalene including biomass burning, the combustion of fossil fuels, chemical manufacturing (including surfactants and phthalic anhydride), tobacco use, and exposure to environmental tobacco smoke. Naphthalene is among the most commonly detected PAHs at high concentrations in association with incense smoke [3, 4]. Cooking, especially deep-frying, can be a significant source of naphthalene indoors [5, 6].

Naphthalene is also a common active ingredient in mothball and moth repellent products [1, 7]. Unintentional exposure to naphthalene in moth repellents is a common reason for calls to Poison Control Centers, with over 1,443 human exposure incidents reported to the National Poison Data System in 2009 [8]. Some studies have reported that the presence of mothballs can be a significant contributor to indoor levels of naphthalene [9], and this observation was recently confirmed in a study of indoor air in Chinese residences [10]. Inhalant abuse of consumer products containing naphthalene has also recently been recognized as case reports in the scientific literature [11, 12].

The acute health effects from overexposure to naphthalene have been most commonly reported in association with accidental and intentional exposures to mothball and other products containing this active ingredient. Naphthalene undergoes oxidative hepatic metabolism, yielding alpha-naphthol and other metabolites that produce oxidative stress [7]. This results in oxidation and denaturation of hemoglobin, the production of Heinz bodies, and increased erythrocyte susceptibility to hemolysis. Methemoglobinemia is also a recognized clinical complication from acute overexposure to naphthalene. Individuals with glucose-6-phosphate dehydrogenase deficiency have long been recognized as being at increased risk of severe hemolytic anemia in association with exposure to naphthalene [13, 14].

The ability to detect urinary monohydroxylated naphthols has led to the development of biomarkers to study occupational exposure to naphthalene [15–17]. More recently, the development of high-throughput analytical methods has enabled these biomarkers to be applied in non-occupational and population-based studies [18–20]. Tobacco use and exposure to environmental tobacco smoke have been found to have a positive correlation with urinary naphthalene metabolites, including 1- and 2-hydroxynaphthalene (NAP1 and NAP2, respectively) [20, 21]. One epidemiologic study from 1999 to 2002 reported factors including age, race/ethnicity, and education attainment as being significant predictors of urinary excretion of PAH metabolites [21].

The purpose of this investigation was to examine the association between naphthalene exposure and hemolysis in the USA. Based upon the acute toxicology of naphthalene and its ability to cause hemolysis, we hypothesized that higher levels of environmental exposure, measured as 1- and 2-hydroxynaphthalene, would be associated with lower levels of hematocrit and hemoglobin. A secondary aim of our investigation was to assess whether levels of monohydroxylated metabolites of naphthalene are significantly different among individuals of different ethnicities, as has been previously reported.

Materials and Methods

Design and Participants

Data from the National Health and Nutrition Examination Survey (NHANES) 2003–2004 were used for the analyses. NHANES uses a representative sample of non-institutionalized US civilians, selected by a complex, multistage probability design. Details of the survey design, data collection methods, and data files for NHANES are available from the National Center for Health Statistics (NCHS) website [22]. Briefly, participants were interviewed in their homes and subsequently examined in mobile examination centers (MEC) across 15 US geographic locations. The NHANES study was approved by the NCHS ethics review board with informed consent obtained from all participants before data collection. The analytic sample for this study consists of adults (18+ years of age) without treated anemia and with complete data on naphthalene, hemoglobin, and hematocrit (n = 2,450).

Measurement of Naphthalene Biomarkers

Blood concentrations were determined on a specimen obtained by venipuncture during the visit to the MEC. The methods used to derive both hemoglobin (Hb, in grams per deciliter) and hematocrit (HCT, in percent) measurements in NHANES are based on the Beckman Coulter method of counting and sizing in combination with an automatic diluting and mixing device for sample processing and a single beam photometer for hemoglobinometry.

Urine specimens were collected from participants in standard urine collection cups during the MEC visit. Both 1- and 2-hydroxynaphthalene were measured in nanograms per liter. The analytical procedure for the measurement of monohydroxylated metabolites of PAHs involved enzymatic hydrolysis of urine, solid-phase extraction, and analysis using capillary gas chromatography combined with high-resolution mass spectrometry [23]. Isotope dilution with ¹³C-labeled standards was used for quantification.

Covariates

Self-reported race and ethnicity were used to classify participants as non-Hispanic White, non-Hispanic Black, or Mexican-American (i.e., persons of Mexican origin living in the USA). Age was defined as the age in years at the time of the household interview. Education was based on number of years the participant attended and completed school, and categorized as less than high school, high school, and more than high school. Smoking history was assessed during the interview and classified into current, former, or never smokers. Iron deficiency was defined as: total iron binding capacity of <240 μg/dL, or iron of <60 μg/dL, or transferrin saturation <20 %. Secondhand smoke was based on the presence of any household member who smoked inside the participant’s home.

Data Analysis

All statistical analyses were performed using SAS (version 9.2, SAS Institute; Cary, NC, USA), SUDAAN (version 10.0.0, RTI International; Research Triangle Park, NC, USA). Sample weights, provided by NCHS, were used to correct for differential selection probabilities and to adjust for non-coverage and non-response. Univariate and multivariate linear models were used to determine associations between hemoglobin and hematocrit with the naphthalene biomarkers of interest using the Taylor linearization method. A two-tailed p value of .05 was used as the threshold for statistical significance.

Results

Table 1 shows the mean levels of NAP1 and NAP2 by demographic characteristic of the population. Significant differences in NAP1 and NAP2 urinary metabolite excretion were observed by race. The highest NAP1 and NAP2 concentrations were observed in non-Hispanic Blacks, followed by non-Hispanic Whites, and lowest in Hispanics. Non-Hispanic Blacks had significantly higher excretion levels of NAP1 metabolite compared to non-Hispanic Whites (p = 0.019) and Hispanics (p = 0.003). Non-Hispanic Blacks also had significantly higher excretion levels of NAP2 metabolite compared to non-Hispanic Whites (p = 0.045) and Hispanics (p = 0.002) after adjusting for age, gender, education, and smoking.

Table 1.

Urinary NAP1 and NAP2 concentrations, weighted means (SE), by characteristic of NHANES 2003–2004 adults

Variable	N (%)	1-hydroxynaphthalene NAP1 (ng/L)	2-hydroxynaphthalene NAP2 (ng/L)
Age		p value (0.2236)	p value (0.0016)
18–39 years	663 (40.36 %)	7,741.10 (723.47)	8,484.77 (707.42)
40–59 years	442 (28.05 %)	9,830.82 (1239.10)	7,271.07 (637.28)
60+ years	498 (31.60 %)	9,813.29 (1114.22)	5,287.00 (540.95)
Sex		p value (0.9931)	p value (0.0363)
Male	1,229 (50.16 %)	8,059.83 (489.37)	7,189.06 (521.20)
Female	1,221 (49.84 %)	8,065.00 (853.36)	6,319.02 (547.87)
Smoker		p value (<0.0001)	p value (<0.0001)
Current	346 (24.56 %)	15,301.23 (1371.6)	15,718.46 (1105.02)
Past	385 (27.32 %)	7,608.01 (873.46)	3,822.89 (250.86)
Never	678 (48.12 %)	6,272.85 (900.75)	4,536.25 (434.54)
Iron deficiency		p value (0.2265)	p value (0.5235)
Yes	350 (47.36 %)	9,000.80 (760.72)	7,449.56 (632.39)
No	389 (53.64 %)	7,212.92 (1287.07)	6,757.83 (984.86)
Race		p value (0.0011)	p value (0.0055)
White	1,011 (41.27 %)	8,153.09 (606.81)	6,738.26 (561.85)
Black	640 (26.12 %)	10,288.75 (1087.91)	9,684.32 (899.56)
Hispanic	698 (28.49 %)	5,295.09 (535.34)	5,186.59 (437.91)
Other	101 (4.12 %)	8,741.37 (3177.03)	4,385.68 (780.95)
Education		p value (0.2746)	p value (0.0004)
<High school	1,352 (55.21 %)	7,224.45 (650.76)	6,415.56 (644.34)
High school	399 (16.29 %)	9,476.03 (1239.12)	8,830.09 (733.84)
>High school	698 (28.50 %)	7,976.50 (1001.21)	5,973.47 (441.24)
Secondhand smoke		p value (<0.0001)	p value (<0.0001)
Yes	514 (21.13 %)	12,887.12 (812.35)	12,850.59 (786.94)
No	1,919 (78.87 %)	6,671.24 (559.92)	4,960.15 (274.16)

Wald t test

The results confirmed tobacco abuse and exposure to secondhand smoke at home as being associated with higher levels of naphthalene metabolite excretion. Current smokers had significantly higher levels of NAP1 and NAP2 urinary metabolites, compared to nonsmokers. Subjects reporting exposure to secondhand smoke at home had significantly higher NAP1 and NAP2 excretion in comparison to those who did not report ETS exposure.

After adjusting for age, gender, race, education, and smoking, a positive association was observed between NAP1 and Hb (0.39 ± 0.11, p = 0.0034) as well as HCT (1.14 ± 0.28, p = 0.0009). A stratified analysis by smoking status showed a stronger association for smokers (Hb 0.63 ± 0.23, p = 0.02; HCT 1.43 ± 0.79, p = 0.09) than nonsmokers (Hb 0.34 ± 0.14, p = 0.03; HCT 1.08 ± 0.42, p = 0.02). NAP2 was not significantly associated with Hb or HCT. The results of these multivariate analyses appear in Table 2.

Table 2.

Association between NAP1 and NAP2, and Hb and HCT, beta coefficients ± SE

	Overall, model 1a	Overall, model 2b	Smokersa	Nonsmokersa
Hb (g/dL)
NAP1, per 100,000 ng/L	0.53 ± 0.12 (p = 0.0004)	0.39 ± 0.11 (p = 0.0034)	0.63 ± 0.23 (p = 0.0152)	0.34 ± 0.14 (p = 0.0311)
NAP2, per 100,000 ng/L	1.25 ± 0.38 (p = 0.0047)	0.53 ± 0.37 (p = 0.17)	0.68 ± 0.68 (p = 0.33)	0.29 ± 0.42 (p = 0.5009)
HCT (%)
NAP1, per 100,000 ng/L	1.50 ± 0.30 (p = 0.0001)	1.14 ± 0.28 (p = 0.0009)	1.43 ± 0.79 p = (0.091)	1.08 ± 0.42 (p = 0.0227)
NAP2, per 100,000 ng/L	3.42 ± 1.07 (p = 0.0059)	1.49 ± 1.18 (p = 0.229)	1.73 ± 2.25 (p = 0.4535)	0.94 ± 1.24 (p = 0.4605)

^aAdjusted for age, gender, race, and education

^bAdjusted for age, gender, race, education, and smoking

Table 3 summarizes the results of multivariate analyses of the association between NAP1, NAP2, Hb, and HCT by race. The results showed that the significant association between NAP1 and hematological endpoints was stronger for non-Hispanic Blacks (Hb 0.54 ± 0.20, p = 0.02; HCT 1.43 ± 0.55, p = 0.02) than for non-Hispanic Whites (Hb 0.37 ± 0.18, p = 0.06; HCT 1.20 ± 0.51, p = 0.03). No significant association was observed between naphthalene metabolites and Hb or HCT in Hispanics.

Table 3.

Association between NAP1 and NAP2, and Hb and HCT by race ethnicity, beta coefficients ± SE

	White	Black	Hispanics
Hb (g/dL)
NAP1, per 100,000 ng/L	0.38 ± 0.18 (p = 0.0598)	0.54 ± 0.20 (p = 0.0156)	0.30 ± 1.7 (p = 0.1005)
NAP2, per 100,000 ng/L	0.82 ± 0.50 (p = 0.1199)	−0.15 ± 0.34 (p = 0.6654)	1.36 ± 0.86 (p = 0.1362)
HCT (%)
NAP1, per 100,000 ng/L	1.20 ± 0.51 (p = 0.0339)	1.43 ± 0.55 (p = 0.0210)	0.99 ± 0.52 (p = 0.0807)
NAP2, per 100,000 ng/L	2.19 ± 1.60 (p = 0.1920)	−0.20 ± 1.01 (p = 0.8480)	3.36 ± 2.72 (p = 0.2386)

^aAdjusted for age, gender, education, and smoking

Discussion

The results of our analyses are consistent with previous findings of other studies showing higher levels of urinary naphthalene metabolites among current smokers and individuals exposed to ETS. Significant differences in excretion of NAP1 and NAP2 metabolites were also observed in different ethnicities in our investigation. The higher level of NAP1 and NAP2 excretion in non-Hispanic Black subjects is an interesting finding that warrants further investigation.

Aside from the assessment of smoking status and ETS exposure, NHANES does not routinely include survey questions about potential sources of exposure to naphthalene. One previously published study of hemolysis among children with gluclose-6-phosphate dehydrogenase deficiency found higher rates of naphthalene mothball usage among African Americans [24]. An interesting aspect of that study was the report that mothballs were being used in a manner inconsistent with the product labeling, for its fresh scent and not as a moth repellent. The generalizability of this observation to explain our observed differences in biomarkers of exposure between race–ethnicity is limited. Future investigations should consider assessment of naphthalene mothball use as well as other exposure factors, including exposure to wood smoke and other products of combustion, that increase levels of naphthalene in the environment.

In addition to the level of environmental exposure, another factor that could affect the excretion of NAP1 and NAP2, is differences in metabolic pathways for naphthalene. One study of healthy Korean males found a significant association between NAP2 and indicators of oxidative stress, only among individuals with a null-type genetic polymorphism of glutathione S-transferase M1 [25]. Other studies on pooled human liver microsomes have suggested that the activity of certain cytochrome P450 isoforms (including CYP1A2 and CYP 2D6) may significantly affect the bioactivation of naphthalene [26]. This suggests that genetic variability in naphthalene metabolism might impact excretion of NAP1 and NAP2 among individuals of different ethnicities.

The hypothesis that higher levels of environmental exposure to naphthalene would be associated with negative impacts on Hb and HCT was not confirmed. In contrast, the opposite effect was observed. Increasing levels of urinary NAP1 was positively associated with both Hb and HCT, among smokers and nonsmokers. A biological mechanism for this previously unreported association is not known and points to the need for confirmation of our findings in other studies.

It is possible that this association may have been observed indirectly, as a result of the fact that as a product of combustion, environmental exposure to naphthalene is usually highly correlated with measurements of carbon monoxide [27, 28]. Previous studies of smokers have reported that their increased exposure to carbon monoxide leads to the development of compensatory increases in red blood cell indices including HCT and Hb [29]. The NHANES 2003–2004 data do not include measurements of carboxyhemoglobin in the study population. Future studies that concurrently measure biomarkers of naphthalene and carboxyhemoglobin may provide additional insight towards the associations that were observed in this study.

While NAP1 and NAP2 are widely utilized as biomarkers of exposure to naphthalene [18–20], there are limitations that warrant additional consideration. In addition to being a biomarker of exposure to naphthalene, NAP1 is also a metabolite of the pesticide aldicarb [30]. This may limit the specificity of the observed associations with Hb and HCT.

Another limitation relates to whether NAP1 and NAP2 are the most accurate biomarkers of human exposure to naphthalene. Research on adults with occupational exposure to PAHs has reported that dihydroxy urinary metabolites of naphthalene were highly correlated with protein adducts of naphthoquinones, which are the toxic products of naphthalene metabolism [31]. A more recent study has extended these findings among the general population and occupationally exposed persons [32]. In this study, 1,2-dihydroxynaphthalene in urine was found to be the most sensitive biomarker of internal dose for naphthalene and was detected at much higher concentrations than NAP1 and NAP2. The dihydroxy urinary metabolites of naphthalene are difficult to isolate and quantify from an analytical perspective [32], but in the future, they may prove to be a more accurate biomarker of exposure to naphthalene in biomonitoring studies. In addition, some investigators are researching the potential use of unmetabolized naphthalene to assess environmental exposure [33]. At the current time, most population-based biomonitoring studies continue to utilize monohydroxylated metabolites of naphthalene as indicators of internal dose [19, 34].

Strengths of our study include measures of both naphthalene exposure and hematological endpoints on a representative sample of the civilian, non-institutionalized US population. Detailed data were available on important covariates, including smoking and exposure to secondhand smoke. Our study was based upon cross-sectional data, which limits the ability to assess cause–effect relationships, but the results present interesting conclusions for future epidemiological studies. The nature of our research identifies some of the ongoing challenges in the interpretation of biomonitoring studies, which include the need for more specific data relating to exposure and genetic susceptibility, as well as a need for further validation of biomarkers of exposure to naphthalene.

Conclusions

Biomarkers of environmental exposure to naphthalene, measured as NAP1 and NAP2 metabolites in the 2003–2004 NHANES study sample, differ based upon race–ethnicity. The specific exposure factors that explain these differences are unclear and warrant further investigation. A positive association between NAP1 excretion and hemoglobin and hematocrit was observed. Further research is needed to better understand this association. The development and validation of newer biomarkers of exposure to naphthalene will present opportunities to better understand internal dose and health implications in the general population.