Vitamin D status in relation to inflammatory risk and albuminuria associated with polycyclic aromatic hydrocarbon exposure in the US population

Abstract

Exposure to polycyclic aromatic hydrocarbons (PAHs) has been associated with both systematic inflammation and renal dysfunction. Reports have suggested that anti-inflammatory properties of vitamin D may provide protection against renal injury. This cross-sectional study tested the hypothesis that serum 25-hydroxyvitamin D [25(OH)D] moderates the inflammation and albuminuria associated with PAH exposure. Data were obtained from 5,982 subjects aged 20–79 years in the National Health and Nutrition Examination Survey (2001–2010). PAH exposure was estimated by urinary PAH metabolites. Inflammation was defined as serum C-reactive protein (CRP) > 3 mg/L and albuminuria as urinary albumin-to-creatinine ratio > 30 mg/g. The results found that greater PAH exposure was linked with inflammation and albuminuria. Individuals with PAH exposure also tended to have lower 25(OH)D and lower vitamin D was associated with both elevated CRP (Odds ratio [OR] = 1.28, 95% confidence interval [CI] = 1.07–1.54) and urinary albumin (1.35, 95%CI = 1.03–1.77) for any given PAH exposure. Those with lower serum 25(OH)D-to-urinary PAH ratios were likewise at a greater risk of elevated CRP and albuminuria. The findings support prior suggestions that exposure to PAHs is associated with inflammation and albuminuria but suggests further that the risk is higher when vitamin D is lower. Thus, nutritional status becomes an important variable in PAH risk assessment.

Background

Polycyclic aromatic hydrocarbons (PAH) are a class of organic chemicals ubiquitously present in the environment. PAH enters the environment primarily as a result of incomplete combustion of wood, oil, coal, and cigarette tobacco.¹ Although there are more than 100 types of PAH, the US population is commonly exposed to a select few (naphthalene, fluorine, phenanthrene and pyrene) whose urinary metabolites are found in >99% of samples collected.²

Exposure to assorted PAH chemicals has been associated with an array of adverse effects including asthma, compromised cardio-renal function, decreased glomerular filtration rate, diabetes and hypertension.^3–6 Although the mechanisms underlying the apparent toxicity remain unclear, evidence indicates that inflammation may provide a common link.^4,7 A recent review also made a similar observation that PAH exposure can trigger oxidative stress or activate inflammatory cytokines (e.g., Interleukin 1 beta, IL-1β), leading to endothelial dysfunction and vessel thickening.⁸ Several reports further suggest that a family of transcription factors such as the aryl hydrocarbon receptor (AHR) are also involved in the inflammatory process.^9,10 On the positive side, vitamin D may reduce the cardiovascular health risk arising from oxidative stress and inflammation.^11,12 Kendrick et al (2009) reported that individuals with vitamin D deficiencies were at greater risk of cardiovascular disease including angina, myocardial infarction or stroke.¹³ More broadly, there appears to be an inverse relationship between serum vitamin D and all-cause mortality.^14–16

A recent general theory proposes that poor health outcomes may evolve from the cumulative and potentially synergistic effects of multiple stressors.¹⁷ Because of the widespread exposure to PAH and their subsequent depot-like storage in adipose, PAH is likely a prime candidate stressor. The continuous renal PAH elimination also makes the glomerular filtration apparatus a focal target for PAH toxicity. Glomerular damage or glomerular inflammation could easily contribute to increases in urinary albumin. Nutrition may play a key role in the moderation of environmental chemical toxicity making nutritional deficiencies a potentially complementary stressor.¹⁸ In the absence of a current report, the following study was designed to test the hypothesis that lower serum 25(OH)D is associated with an increased risk of systemic inflammation and albuminuria associated exposure to PAH in the US population.

Materials and methods

Study design and population

To ensure sufficient sample size and data consistency for producing reliable statistics, the data set used combined 5 cycles of the National Health and Nutrition Examination Survey (NHANES, 2001–2002, 2003–2004, 2005–2006, 2007–2008, and 2009–2010 following the NHANES analytic and data reporting guidelines;^19,20 please also see Statistical Analysis section for details). NHANES were conducted by the National Center for Health Statistics (NCHS) at the Center for Disease Control and Prevention (CDC), to assess the health and nutritional status of a statistically representative sample of the civilian, non-institutionalized US population with a multistage, stratified sampling design. The protocol was approved by the NCHS Institutional Review Board, and all subjects provided written informed consent. This observational research was conducted in accordance with the STROBE guideline (Strengthening the Reporting of Observational Studies in Epidemiology).²¹

This analysis is limited to non-Hispanic Whites, non-Hispanic Blacks, and Mexican Americans, 6,337 participants aged 20–79 years old with valid measurements of urinary PAH metabolites and serum 25(OH)D. To minimize the potential bias due to excessive influence of outliers or missing data, the current analysis further excluded those with serum 25(OH)D greater than 150 nmol/L (n = 17), body mass indices (BMI) > 65 or < 14.5 (n = 67), missing data for other covariates (e.g., serum C-reactive protein, n = 12) or who were pregnant (n = 259). After adjusting for sample weights, the remaining 5,982 subjects in the analyses represent an estimated U.S. non-institutional population of 797,235,494 persons aged 20–79 years old.

Urinary analyses: PAH metabolites, creatinine, and albuminuria

Urine samples were obtained from participants during the physical examination, shipped on dry ice, and stored frozen until analysis.²² The eight urinary PAH metabolites identified in >99% of the samples (1-napthol, 2-napthol, 2-hydroxyfluorene, 3-hydroxyfluorene, 1-hydroxyphenanthrene, 2-hydroxyphenanthrene, 3-hydroxyphenanthrene, and 1-hydroxypyrene) were included in the current analysis. PAH exposure was estimated from the integrated sum of urinary PAH metabolites normalized to urinary creatinine to correct for urine volume dilution.

Urine creatinine was measured by the modified kinetic Jaffé method in 2003–2006 and by an enzymatic (creatinase) method in 2007–2010. Urinary albumin was determined using fluorescein immunoassay and a Sequoia-Turner digital fluorometer (Model 450). A urinary albumin-creatinine ratio (milligrams per gram, mg/g) was then calculated for each subject.

Blood analyses: 25-Hydroxyvitamin D [25(OH)D] and C-reactive protein (CRP)

Collection and measurement of serum 25-Hydroxyvitamin D [25(OH)D] were performed by CDC following the NHANES Laboratory/Medical Technologists Procedures Manual. In brief, 25-hydroxylated forms of both dietary (D₂) and endogenous (D₃) 25(OH)D were determined initially by RIA (Diasorin, Stillwater, MN) in 2001–2006 and then later in 2007–2010 by a combination of HPLC and mass spectrometry as described previously.²³ The CDC then used regression equations to normalize the values across the two methodologies.²⁴ The total 25(OH)D employed in the current analyses is the sum of the two 25 hydroxylated precursors.

High sensitivity serum C-reactive protein was measured using latex-enhanced nephelometry with the Dade Behring Nephelometer II Analyzer System (Dade Behring Diagnostics, Inc., Somerville, New Jersey) with a limit of detection near 0.02 mg/dL. Approximately 3% of values were below LOD and were assigned fill values calculated as LODs divided by the square root of 2. Individuals with serum a CRP > 3 mg/L were considered at high risk of inflammation.²⁵ Blood sample collection, measurement procedures, quality control and assurance were performed following the NHANES Laboratory/Medical Technologists Procedures Manual and are reported elsewhere.²²

Determination of demographic variables and other covariates

Demographic information on age, sex, race/ethnicity, cigarette smoking, and alcohol use was self-reported and body mass index was calculated from measured height and weight.

Diabetes mellitus was assigned based on a self-reported diagnosis of diabetes, the use of diabetic medications (insulin or oral agents), a nonfasting plasma glucose ≥ 200 mg/dL, or a fasting plasma glucose ≥ 126 mg/dL. Hypertension was defined as an average systolic blood pressure ≥ 140 mm Hg, a diastolic blood pressure ≥ 90 mm Hg, a physician’s diagnosis or the use of anti-hypertensive medications.²⁶

Statistical analyses

Demographic characteristics were first compared between vitamin D status using the Pearson chi-square test for categorical variables (e.g., race/ethnicity) and t-tests for continuous variables (e.g., measures of urinary PAH metabolites). 25(OH)D status was partitioned as adequate, ≥ 50 nmol/L, or inadequate, < 50 nmol/L.²⁷ The Spearman rank-order correlation was also used to examine the relationships of urinary PAH metabolites and serum 25(OH)D to CRP versus urinary albumin-to-creatinine ratio (UACR) on the continuous scale. Logarithmic transformations were performed to normalize the continuous data whenever necessary.

Unadjusted and adjusted odds ratios (OR) with 95% confidence interval (95% CI) generated from logistic regression models were then used to examine the roles of vitamin D status and PAH exposure in relation to elevated CRP and albuminuria. PAH exposure, assessed by creatinine-adjusted total urinary PAH metabolites, was analyzed as a categorical variable based on its weighted distribution (<33th percentile, 33–66th percentile, >66th percentile) and as a continuous variable. Possible confounding factors, such as age, gender, race/ethnicity, body mass index, cigarette smoking, alcohol consumption, diabetes, and hypertension, were added to adjust the association of interest in the multivariate-adjusted logistic regression models. The interaction of urinary PAH metabolites (creatinine-adjusted) and Vitamin D was examined by including the interaction term in the analysis. The combined influence of vitamin D and PAH exposure as a continuous variable was also estimated by examining the relation of the vitamin D-PAH exposure ratio (log-transformed) to elevated CRP and albuminuria with covariate adjustments.

NHANES provided sampling weights to account for the complex survey design in order to ensure that calculated estimates are representative of the U.S. civilian non-institutionalized population. In the current analysis, sample weights for NHANES 2001–2010 analyses were computed by combining the sample weights for each individual survey cycle (2001–2002, 2003–2004, 2005–2006, 2007–2008, and 2009–2010) following the NCHS analytical method guideline.²⁸ All statistical analyses were adjusted with the sample weights using SUDAAN 10.01 (Research Triangle Institute, 2009) to produce unbiased estimates with statistical significance set at 0.05.

Results

The demographic characteristics of study participants by vitamin D status are given in Table 1. The geometric mean and 5th-to-95th percentile of the current study population for serum Vitamin D were 66.1 and 30.7–107.3 nmol/L, respectively (data not shown). Overall, except for cigarette smoking, the demographic characteristics of those assigned to the inadequate 25(OH)D group were different. 25(OH)D was generally lower among participants who were younger, female, minorities (Hispanic Blacks or Mexican Americans), hypertensive, diabetic, used less alcohol and had higher BMIs (p < 0.05).

Table 1.

Demographic characteristics of study subjects^a.

		Vitamin D status
Characteristics	Overall	Inadequateb serum 25(OH)D	Adequate serum 25(OH)D
Sample size, n	5982	1,934	4,048
	Percent in category (SE)
Age (years)*
60–79	20.8 (0.68)	18.7 (1.17)	21.5 (0.79)
40–59	41.8 (0.85)	41.1 (1.58)	42.1 (1.07)
20–39	37.4 (0.86)	40.2 (1.50)	36.4 (1.02)
Sex*
Females	50.4 (0.73)	54.8 (1.47)	49.0 (0.86)
Males	49.6 (0.73)	45.2 (1.47)	51.0 (0.86)
Race/ethnicity*
Mexican Americans	8.94 (0.91)	14.3 (1.68)	7.26 (0.73)
Non-Hispanic Blacks	11.7 (0.84)	33.4 (2.01)	4.97 (0.48)
Non-Hispanic Whites	79.3 (1.32)	52.3 (2.47)	87.8 (0.94)
Body mass index (kg/m2)*
> 30	35.3 (0.80)	49.2 (1.29)	31.0 (0.89)
25–30	33.2 (0.74)	29.5 (1.26)	34.3 (0.85)
<25	31.5 (0.79)	21.3 (0.98)	34.7 (0.94)
Ever smoking
Yes	49.9 (1.02)	49.4 (1.26)	50.1 (1.28)
No	50.1 (1.02)	50.6 (1.26)	49.9 (1.28)
Daily alcohol use*
≥ 1 drink/day	15.7 (0.59)	12.7 (0.99)	16.6 (0.67)
< 1 drink/day	84.3 (0.59)	87.3 (0.99)	83.4 (0.67)
Hypertension*
Yes	34.7 (0.81)	39.0 (1.45)	33.4 (0.92)
No	65.3 (0.81)	61.0 (1.45)	66.6 (0.92)
Diabetes*
Yes	9.1 (0.49)	11.3 (0.88)	8.41 (0.57)
No	90.9 (0.49)	88.7 (0.88)	91.6 (0.57)
Serum C-reactive protein*
> 3 mg/L	34.2 (0.78)	44.6 (1.41)	31.0 (0.94)
≤ 3 mg/L	65.8 (0.78)	55.4 (1.41)	69.0 (0.94)
Urinary albumin-to-creatinine ratio (UACR)*
> 30 mg/g	8.40 (0.47)	11.6 (0.87)	7.40 (0.53)
≤ 30 mg/g	91.6 (0.47)	88.4 (0.87)	92.6 (0.53)

^aDemographic variables (e.g., race/ethnicity) are presented as percent (%) in category (standard error) and were compared between vitamin D status. P-values were calculated with chi-square test.

^bThe cutoff for serum 25(OH)D adequacy was ≥ 50 nmol/L.²⁷

^*Significant difference (p < 0.05) in demographic characteristics between subjects with serum 25(OH)D above vs. below 50 nmol/L (cutoff).

Abbreviations: 25(OH)D; 25-Hydroxyvitamin D; 95% CI, 95% confidence interval; ANOVA, analysis of variance; CRP, C-reactive protein; PAH, polycyclic aromatic hydrocarbon.

Inadequate 25(OH)D was also associated with an increased likelihood of elevated serum C-reactive protein and urinary albumin-to-creatinine ratio (UACR). Only 7.40% of the subjects with adequate Vit D developed albuminuria, while the proportion of albuminuria was significantly higher, 11.6%, in the inadequate group. Similar findings were observed when the relationships of serum 25(OH)D with CRP and the urinary albumin-to-creatinine ratio (UACR) were examined on a continuous scale using Spearman correlation test (Figure 1). Subjects with lower 25(OH)D values were more likely to have elevated serum CRP and urinary albumin. Both individual PAH, and likewise total PAH metabolites, in urine were significantly higher in the inadequate 25(OH)D group (Appendix A).

Figure 1.

Scatterplot matrix showing Spearman correlation among urinary PAH metabolites, serum 25(OH)D, CRP, and urinary albumin-to-creatinine ratio (UACR) (all four variables were log-transformed and, except for the association (r = −0.02, p = 0.18) between urinary PAH and serum 25(OH)D, all correlations (marked with *) are statistically significant, p < 0.05.

Consistent with the findings observed in Figure 1, urinary PAH metabolites were significantly associated with both elevated CRP and albuminuria regardless of Vitamin D status in the crude (unadjusted) logistic regression models (Appendix B). For instance, for those with lower 25(OH)D, the risk of elevated CRP was 55% greater in those with the highest (>75^th percentile) PAH exposure (odds ratio [OR] = 1.55, 95% confidence interval [95% CI] = 1.15–2.10) versus those with the lowest PAH exposure (<25^th percentile). The increased risk was intermediate, 24%, in the middle (25^th −75^th percentile) exposure group (OR= 1.14, 95% CI = 0.96–1.60). When 25(OH)D was labeled adequate, the risk of elevated CRP was similarly, though less dramatically, associated with both high (OR= 1.37) and intermediate (OR= 1.23) exposure to PAH. There were no statistically significant interactions found between urinary total PAH metabolites (both categorical and continuous) and vitamin D status for elevated CRP or albuminuria.

Of note, both PAH exposure and vitamin D status remained significantly associated with elevated CRP and albuminuria even after controlling for multiple covariates in multivariate-adjusted logistic regression models (Table 2). Comparable findings were observed when using log-transformed total PAH metabolites concentrations (data not shown). A dose-response pattern was evident from the trend in increasing adjusted odds ratios as urinary PAH metabolites increased for both elevated CRP and albuminuria. On the other hand, lower vitamin D status is associated with elevated CRP (Odds ratio [OR] = 1.28, 95% confidence interval [CI]: 1.07–1.54) and albumin (OR = 1.35, 95% CI = 1.03–1.78) for any given measure of urinary PAH metabolites. The interaction between urinary total PAH metabolites and vitamin D status remain insignificant in the multivariate models (data not shown). Of covariates examined in multivariate-adjusted analyses, body mass index, diabetes and hypertension were three major risk factors (for elevated CRP and albuminuria), followed by sex (for elevated CRP) and age and race/ethnicity (for albuminuria).

Table 2.

Multivariate-adjusted odds ratio for inflammatory and albuminuria risk associated with urinary levels of PAH metabolites.

	Elevated CRP a		Albuminuria a
Characteristic	Odds ratio (95% CI)	P-value	Odds ratio (95% CI)	P-value
Urinary PAH metabolites (ng/ g creatinine) b		<0.001		0.01
High (> 75th percentile)	1.55 (1.24–1.95)		1.57 (1.10–2.24)
Medium (25th-75th percentile)	1.14 (0.94–1.38)		1.03 (0.75–1.40)
Low (< 25th percentile)	1.00 (reference)		1.00 (reference)
Serum 25(OH)D (mg/dL)		0.01		0.03
Inadequate	1.28 (1.07–1.54)		1.35 (1.03–1.77)
Adequate c	1.00 (reference)		1.00 (reference)
Age (years)		0.92		0.01
60–79	1.04 (0.85–1.27)		1.64 (1.18–2.28)
40–59	1.01 (0.84–1.20)		1.13 (0.85–1.50)
20–39	1.00 (reference)		1.00 (reference)
Sex (female)	2.14 (1.84–2.49)	<0.001	1.12 (0.90–1.38)	0.31
Race/ethnicity		0.64		<0.001
Mexican Americans	1.08 (0.91–1.29)		1.71 (1.30–2.24)
Non-Hispanic Blacks	1.06 (0.88–1.28)		1.27 (0.95–1.69)
Non-Hispanic Whites	1.00 (reference)		1.00 (reference)
Body mass index (kg/m2)		<0.001		0.02
≥ 30	6.87 (5.56–8.48)		0.93 (0.69–1.25)
25–29.9	2.31 (1.87–2.85)		0.66 (0.49–0.90)
<25	1.00 (reference)		1.00 (reference)
Ever cigarette smoking (yes)	1.14 (0.96–1.35)	0.13	1.10 (0.89–1.38)	0.37
Daily alcohol use (yes)	0.96 (0.78–1.19)	0.69	0.92 (0.67–1.26)	0.61
Hypertension (yes)	1.17 (1.01–1.37)	0.04	1.84 (1.47–2.31)	<0.001
Diabetes (yes)	1.28 (1.03–1.59)	0.03	3.96 (3.17–4.94)	<0.001

^aElevated c-reactive protein (CRP) was defined by serum CRP > 3 mg/L and albuminuria was defined as urinary albumin-to-creatinine ratio (UACR) > 30 mg/g.

^bThe cutoff for serum 25(OH)D adequacy was ≥ 50 nmol/L.²⁷.

^cUrinary PAH metabolites categories: high (> 75^th percentile, 179.5 ng/g creatinine), medium (25–75^th percentile, 179.5–37.1 ng/g creatinine), and low (<25^th percentile, 37.1 ng/g creatinine).

Abbreviations: 25(OH)D; 25-Hydroxyvitamin D; 95% CI, 95% confidence interval; CRP, C-reactive protein; OR, odds ratio; PAH, polycyclic aromatic hydrocarbon.

In the secondary analysis, comparable results were also observed when both serum 25(OH)D and urinary PAH metabolites were analyzed together as a continuous ratio for inflammation (Figure 2A) and albuminuria (Figure 3A), respectively. There was an apparent inverse dose-response relationship between the 25(OH)D/PAH ratio (log-transformed) with covariate adjustment, which tightened up (narrower confidence interval) particularly when extreme outliers (n = 7, observations with 3 standard deviations from the mean,) were excluded from the analysis (Figures 2B and 3B).

Figure 2.

Odds ratio for the association between the ratio of serum 25(OH)D-to-total PAH metabolites in urine (log-transformed) and inflammation (adjusting for age, sex, body mass index, race/ethnicity, smoking status, alcohol consumption, diabetes, and hypertension) for all observations (A) with and (B) without outliers observations (n = 7). Dotted curves: 95% confidence interval curves; rug plot on the x-axis describing the distribution of the serum 25(OH)D-to-urinary PAH ratio (log-transformed).

Figure 3.

Odds ratio for the association between the ratio of serum 25(OH)D-to-total PAH metabolites in urine (log-transformed) and albuminuria (adjusting for age, sex, body mass index, race/ethnicity, smoking status, alcohol consumption, diabetes, and hypertension) for all observations (A) with and (B) without outliers & influential observations (n = 7). Dotted curves: 95% confidence interval curves; rug plot on the x-axis describing the distribution of the serum 25(OH)D-to-urinary PAH ratio (log-transformed).

Discussion

The current population-based study shows that urinary PAH metabolites were associated with increased risk of inflammation and albuminuria in a dose-response manner. The associations of interest are independent of (adjusted for) other conventional risks factors (e.g., age, smoking, diabetes, and hypertension). The analyses also support the suggestion that low vitamin D status exacerbates the renal and inflammatory health risks associated with any given PAH exposure (Table 2). Similar results were observed when the effects of PAH and vitamin D were evaluated as the ratio of serum 25(OH)D-to-urinary PAH metabolites. In this later case, the greater the relative influence of vitamin D in the ratio, the lower the associated risk of elevated CRP and albuminuria.

The observations referred above support anti-inflammatory properties of Vitamin D^29,30 as the oxidative stress and resulting inflammatory responses play a key role in the development of albuminuria. ^31,32 For instance, lower serum vitamin D has been linked to increased risk of inflammation and endothelial dysfunction.³³ This observation is consistent with the current finding of an inverse relationship between 25(OH)D and CRP. Available 25(OH)D₂ and D₃ are substrates for renal 1-hydroxylase activity which produces the most active form of the vitamin (calcitriol). Declining renal function can impair 1-hydroxylase activity and could contribute to lower concentrations of effective vitamin D. The resulting decline in intestinal absorption has been associated with other systemic pathologies.¹¹

Collectively, the findings support previous reports that the toxicity of PAH involves chronic inflammation, a leading risk factor in the development of cardiovascular disease, asthma, cancer, and renal disease.³⁴ In addition to CRP, other inflammatory markers observed following PAH exposure in coke oven workers include: elevated reactive oxygen species (ROS), lipid peroxides, immunoglobulin A and TNF-alpha.³⁵ Interestingly, a recent paper reported that Vitamin D deficiency can increase asthma risk in children.³⁶ Mechanistically, vitamin D may facilitate the metabolic clearance of PAH by stimulating the expression of key biotransformation and anti-oxidant genes.³⁷ The metabolic detoxification of PAH can however be problematic in that some intermediates are also potentially toxic and carcinogenic.³⁸ The funneling of these water soluble byproducts to the kidney for excretion may explain in part the observed relationship between elevated PAH and compromised renal function as evident in albuminuria. PAHs may also reciprocate by increasing the degradation of vitamin D. The tobacco related PAH, benzo[a]pyrene increases CYP24A1 activity, which in turn degrades calcitriol.⁹

The interplay between vitamin D and environmental toxins may also have a socioeconomic component. Minority communities across the U.S. often experience greater pollution and suffer from poor nutrition.¹⁷ For instance, people of color and lower-income groups tend to bear a disproportionately higher burden from air pollution,^39,40 while also suffering from vitamin D insufficiency (Table 1). This combination may explain in part why both elevated CRP and albuminuria are more prevalent among minority communities, although only the latter reached statistical significance (Table 2). Even among nutritionally healthy African Americans, skin pigmentation can reduce vitamin D synthesis and prevent the achievement optimal serum 25(OH)D regardless of the season.⁴¹

There are some limitations to the current study. Most notably, the current study was cross-sectional, analyzing samples collected at single time point to assess aspects of chronic exposure, long term vitamin D status and health outcomes. As in any cross-sectional study, one can only demonstrate associations of interest, but not causation. Since all of the measures are dynamic and vary for instance with sun exposure, diet and metabolic status,⁴² the use of single determinations may not capture possible temporal variations. Despite this concern, more frequent sampling suggests that vitamin D values are relatively stable. Prospective longitudinal studies would seem warranted to address this question in more detail. Another potential limitation is the existence of possible residual confounding. Despite adjusting for a number of leading risk factors for inflammation and albuminuria (e.g., smoking and hypertension), the possibility of residual confounding cannot be ruled out. In the end however, the study adds strong confirmatory support for synergistic effects of serum 25(OH)D and urinary PAH metabolites on inflammation and renal dysfunctions (albuminuria) and suggests they be included with more traditional tools for evaluation of vitamin D status and PAH exposure, respectively.^23,43

On the other hand, the data were collected from a large nationally representative study sample following standardized protocols with rigorous quality control and assurance procedures. This allowed the use of multivariate-adjusted statistical analyses to examine the hypothesis that inflammation is a common denominator in the relationship between toxic volatiles, evolving renal dysfunction and vitamin D. Although comparable findings from the secondary analysis were obtained when data were examined (e.g., Vitamin D in serum) on the continuous scale, a detailed identification of optimal cutoff values for key variables including Vitamin D status could further improve this approach.

In conclusion, the current analysis is the first population-based cohort study to assess the role of both PAH exposure and vitamin D status in relation to inflammatory risk and renal status. Inadequate vitamin D status may increase the susceptibility to disease associated with PAH exposure. These findings also suggest the future importance of incorporating both intrinsic physiological (metabolic status) and extrinsic (nutritional and socioeconomic status) when making assessments of population-level chemical toxicities. Understanding the specific mechanisms by which vitamin D may participate in this process also becomes a new target that will require further analyses.

Selected Abbreviations and Acronyms:

CRP

C-reactive protein

NHANES

The National Health and Nutrition Examination Survey

OR

Odds ratio

PAHs

polycyclic aromatic hydrocarbons

25(OH)D

25-Hydroxyvitamin D

95% CI

95% confidence interval

Appendix A. The distributions of urinary PAH metabolites (median and interquartile range, ng/L).^a

			Vitamin D status
Urinary measurements (ng/L)	n	Overall	Inadequate b serum 25(OH)D	Adequate serum 25(OH)D
Naphthalene metabolites
1-Hydroxynaphthalene (1-naphthol)*	5892	2218 (913–7875)	2932 (1165–10970)	2044 (868–7052)
2-Hydroxynaphthalene (2-naphthol)*	5931	3388 (1419–8720)	4744 (1862–2140)	3030 (1303–7635)
Fluorene metabolites
2-Hydroxyfluorene*	5919	2845 (132–756)	384 (166–1130)	259 (124–657)
3-Hydroxyfluorene*	5894	101 (46.9–349)	141.05 (56.9–545)	93.8 (43.9–298)
Phenanthrene metabolites
1-Hydroxyphenanthrene*	5937	154 (81.9–278)	170 (90.7–323)	150 (79.4–266)
2-Hydroxyphenanthrene*	5892	67.2 (34.8–129)	83.7 (42.0–161)	63.3 (32.0–119)
3-Hydroxyphenanthrene*	5888	97.2 (48.6–195)	116 (57.8–242)	93.0 (45.9–184)
Pyrene metabolites
1-Hydroxypyrene*	5911	88.3 (39.8–191)	111 (49.6–246)	82.6 (37.2–174)
Σ PAH metabolites w/o creatinine adjustment (ng/L)*	5982	7555 (3358–9262)	9969 (4198–7687)	6860 (3128–6859)
Σ PAH metabolites w/ creatinine adjustment (ng/g creatinine)*	5982	67.8 (37.1–179)	77.4 (38.2–230)	65.5 (36.5–161)
Urinary creatinine (mg/dL)	5982	115 (64.2–172)	132 (78.9–194)	110 (58.1–165)

^aThe distribution of urinary PAH metabolite was presented as median (interquartile range) and was compared between vitamin D status. P-values were calculated with log transformed data using t-test.

^bThe cutoff for serum 25(OH)D adequacy was ≥ 50 nmol/L.²⁷

^*Significant difference (p < 0.05) between subjects with serum 25(OH)D above vs. below serum 25(OH)D of 50 nmol/L (the cutoff).

Abbreviations: 25(OH)D; 25-Hydroxyvitamin D; 95% CI, 95% confidence interval; ANOVA, analysis of variance; CRP, C-reactive protein; PAH, polycyclic aromatic hydrocarbon.

Appendix B. Crude odds ratio (95%CI) for inflammatory risk and albuminuria associated with urinary levels of urinary PAH metabolites (ng/g creatinine) by vitamin D status

	Elevated CRP a		Albuminuria a
PAH metabolites (ng/g creatinine)	Inadequate b serum 25(OH)D	Adequate serum 25(OH)D	Inadequate serum 25(OH)D	Adequate serum 25(OH)D
High (> 75th percentile) c	1.55 (1.15–2.10)	1.37 (1.08–1.73)	1.71 (1.06–2.75)d	1.70 (1.12–2.59)
Medium (25th-75th percentile)	1.24 (0.96–1.60)	1.23 (0.99–1.53)	1.44 (0.92–2.25)d	1.05 (0.75–1.46)
Low (< 25th percentile)	Reference	reference	Reference	reference
Continuous c	1.15 (1.05–1.26)	1.10 (1.03–1.17)	1.16 (1.02–1.33)	1.22 (1.09–1.39)

^aElevated c-reactive protein (CRP) was defined by serum CRP > 3 mg/L and albuminuria was defined as urinary albumin-to-creatinine ratio (UACR) > 30 mg/g.

^bThe cutoff for serum 25(OH)D adequacy was ≥ 50 nmol/L.²⁷

^cUrinary PAH metabolites were assessed as categorical variable: high (> 75^th percentile, 179.5 ng/g creatinine), medium (25–75^th percentile, 179.5–37.1 ng/g creatinine), and low (< 25^th percentile, 37.1 ng/g creatinine) as well as continuous variable (logarithm transformed, ng/g creatinine).

^dExcept the marginal association (p = 0.07) between albuminuria and categorical PAH metabolites for people with low vitamin D status (denoted by “^d”), PAH metabolites was statistically (p < 0.05) associated with elevated CRP as well as albuminuria.

Abbreviations: 25(OH)D; 25-Hydroxyvitamin D; 95% CI, 95% confidence interval; CRP, C-reactive protein; OR, odds ratio; PAH, polycyclic aromatic hydrocarbon.