Urinary Porphyrin Excretion in Neurotypical and Autistic Children

James S. Woods; Sarah E. Armel; Denise I. Fulton; Jason Allen; Kristine Wessels; P. Lynne Simmonds; Doreen Granpeesheh; Elizabeth Mumper; J. Jeffrey Bradstreet; Diana Echeverria; Nicholas J. Heyer; James P.K. Rooney

Disclosures

Environ Health Perspect. 2010;118(10):1450-1457. 

In This Article

Results

The Study Population

Table 1 describes the demographic distributions for the study population by diagnosis category. Among all children enrolled, 278 had urinary measures of porphyrins and Hg. Among these subjects, 117 were determined to have NT development, 100 met the criteria for AU, and 27 were determined to have PDD-NOS. An additional 34 had other neurodevelopmental diagnoses that included Rett syndrome (n = 1), Asperger's syndrome (n = 4), attention deficit hypersensitivity disorder (ADD/ADHD) (n = 12), sensory integration disorder (n = 5), and language and speech delay (n = 3). Fifty-five subjects had undergone chelation therapy, including 3 NT, 36 AU, 8 PDD-NOS, and 8 other. Only the 197 children who had not been chelated and who had diagnoses of NT, AU, or PDD-NOS were included in further analyses, depicted in Table 2. Only five AU and four PDD-NOS cases were girls, consistent with the much lower frequency of autism and related disorders among females. Therefore, although female subjects are included in descriptive analyses of porphyrin levels (Table 3), they were not included in the logistic regression analyses. Thus, logistic regression analyses (Table 4) were conducted among the group of 133 male children that included 59 AU, 15 PDD-NOS, and 59 NT.

As noted in Table 2, age distributions by diagnosis were similar among male subjects. In addition, most covariates were not statistically different between diagnostic groups. In particular, mean urinary past Hg levels, whether unadjusted (micrograms per liter) or adjusted for creatinine (micrograms per gram), were not significantly different between groups. This was also true for the potential sources of Hg exposure, including the mean number of amalgam fillings (both currently in the child or in the mother over the course of pregnancy) and the mean sum of vaccines administered to the child in total, or before 2002.

Urinary Porphyrin Concentrations in Children

The distribution of six urinary porphyrins is presented in Table 3 and Figures 1 and 2. Table 3 presents the mean (± SD) creatinine-adjusted porphyrin concentrations by sex for all nonchelated NT and AU subjects. Values were stratified by 2-year age groups between 2 and 12 years of age. Figures 1 and 2 show the variation in these porphyrin levels by age for males only.

Figure 1.

Distributions of urinary porphyrins by age (mean and 95% CI). (A) Table describes number of subjects by age group. (B) Bar graphs represent mean and 95% CIs of individual creatinine-adjusted urinary porphyrins (nmol/g) by age group for nonchelated NT (n = 57) and AU (n = 59) boys, age 2–12 years. Graphs depict substantial excess and variable excretion of most porphyrins among AU compared with NT. Porphyrins were evaluated as described under "Materials and Methods."
*NT significantly (p ≤ 0.05) different from same-age AU.

Figure 2.

Association between urinary porphyrins and age. Scatterplots and simple linear regression fit lines of natural logs of individual creatinine-adjusted urinary porphyrins by age group for nonchelated NT (n = 57) and AU (n = 59) boys, age 2–12 years. Graphs clearly depict decreasing porphyrin concentrations with age among NT and disruption of that effect among AU.

Distributions of urinary porphyrins by age (mean and 95% CI). (A) Table describes number of subjects by age group. (B) Bar graphs represent mean and 95% CIs of individual creatinine-adjusted urinary porphyrins (nmol/g) by age group for nonchelated NT (n = 57) and AU (n = 59) boys, age 2–12 years. Graphs depict substantial excess and variable excretion of most porphyrins among AU compared with NT. Porphyrins were evaluated as described under "Materials and Methods." *NT significantly (p ≤ 0.05) different from same-age AU.

Association between urinary porphyrins and age. Scatterplots and simple linear regression fit lines of natural logs of individual creatinine-adjusted urinary porphyrins by age group for nonchelated NT (n = 57) and AU (n = 59) boys, age 2–12 years. Graphs clearly depict decreasing porphyrin concentrations with age among NT and disruption of that effect among AU.

Average concentrations of most porphyrins were elevated in two NT male subjects < 2 years of age compared with older NT males (Table 3). No AU children < 2 years of age were included in the study. Among males in the 2- to 12-year age groups, the mean concentrations of hexacarboxyl- (p < 0.01), pentacarboxyl- (p < 0.001), and copro- (p < 0.009) porphyrins were significantly higher among AU compared with NT groups based on ANOVA F-test values, whereas the heptacarboxyl porphyrin was more of borderline significance (p < 0.06). Uro- and precoproporphyrins did not differ significantly between AU and NT groups.

We observed substantial variation in creatinine-adjusted urinary porphyrin levels among AU males as well as decreasing concentrations of heptacarboxyl-, hexacarboxyl-, pentacarboxyl-, and coproporphyrins with increasing age among NT males (Figure 1). Scatterplots with simple linear regression fit lines show inverse associations between age and porphyrins among NT males, while also demonstrating that this pattern is disrupted among those with AU (Figure 2).

Urinary Porphyrins and Risk of Autism

Logistic regression models of age-adjusted associations between porphyrin levels and AU, AU plus PDD-NOS, and PDD-NOS alone (all vs. NT) in males indicated significant associations of hexacarboxyl-, pentacarboxyl-, and coproporphyrins with AU (Table 4). A one-unit increase in the natural log of the creatinine-adjusted value for coproporphyrin is associated with a 2-fold risk for AU (OR = 2.03; 95% CI, 1.15–3.57). Similar associations were observed with pentacarboxyl porphyrin (OR = 2.36; 95% CI, 1.3–4.07) and with hexacarboxyl-, pentacarboxyl-, and coproporphyrins combined (OR = 2.38; 95% CI, 1.42–3.97). In contrast, porphyrin levels did not differ between PDD-NOS and NT males in this study; consequently, combining PDD-NOS and AU subjects weakened associations. Thus, this analysis seems to indicate that AU is a distinct entity from PDD-NOS in terms of being associated with altered porphyrin excretion. Alternatively, too few PDD-NOS subjects were available to this exploratory study to demonstrate an association with PDD-NOS as a less markedly affected portion of the autistic spectrum. Further studies involving greater numbers of subjects with PDD-NOS as well as other recognized disorders of the autistic spectrum are required to determine the extent to which the strength of the association varies with the degree of ASD. Urinary Hg and other Hg-related measures were not significantly associated with AU based on logistic models with and without adjustment for porphyrins (data not shown).

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.

processing....