Ordinal dose-response modeling approach for the phthalate syndrome

Area of Science:

• Reproductive toxicology
• Developmental toxicology
• Environmental health

Background:

• Phthalate syndrome (PS) encompasses male reproductive developmental effects from gestational exposure to toxic phthalates.
• Current statistical methods often treat PS as a single endpoint, limiting detailed analysis of varied severity.

Purpose of the Study:

• To enhance dose-response modeling for PS by incorporating endpoint severity levels.
• To develop a more nuanced approach for assessing risks associated with developmental toxicants.

Main Methods:

• Ordinal dose-response modeling applied to diisobutyl phthalate (DIBP) exposure data in male rats.
• Categorization of PS endpoints into ordinal levels based on fertility impact.
• Benchmark dose estimation for each severity level using bootstrap procedures and sensitivity analysis.

Main Results:

• Ordinal modeling yielded benchmark dose estimates with lower variability and closer values compared to traditional dichotomous methods.
• Sensitivity analysis validated the reliability of the bootstrap results for ordinal modeling.
• The log-logistic model, applied ordinally, demonstrated improved estimation accuracy.

Conclusions:

• Ordinal dose-response modeling effectively addresses severity variations in PS endpoints.
• This method is adaptable for future inclusion of more severity levels and cumulative risk assessments.
• The approach offers a more refined tool for evaluating phthalate exposure risks.