Understanding Heterogeneity in Clinical Cohorts Using Normative Models: Beyond Case-Control Studies

Area of Science:

• Neuroscience
• Biomedical Science
• Psychiatry

Background:

• Traditional case-control studies assume homogenous clinical groups, which is often biologically inaccurate.
• Existing frameworks like the Research Domain Criteria (RDoC) aim to address this by linking symptom dimensions to biological domains but lack methods for cohort stratification.
• The case-control approach limits inference on diagnostic label validity due to inherent heterogeneity.

Purpose of the Study:

• To introduce normative modeling as a method for parsing clinical cohort heterogeneity at the individual subject level.
• To demonstrate how normative modeling can map variations within a cohort and relate them to specific symptoms.
• To provide a framework for understanding disease as deviations from normal functioning.

Main Methods:

• Developed and applied normative modeling to analyze individual differences in a large healthy cohort (N=491).
• Mapped the relationship between trait impulsivity and reward-related brain activity.
• Identified outliers and quantified their deviation from the norm (outlier magnitude).

Main Results:

• Identified participants as outliers within the normative distribution of reward-related brain activity.
• Demonstrated that outlier magnitude correlates with hyperactivity symptoms in attention-deficit/hyperactivity disorder (ADHD), but not inattention.
• Showed individualized patterns of abnormality linked to specific ADHD symptoms.

Conclusions:

• Normative modeling offers a framework to study disorders at the individual level without dichotomizing cohorts.
• Disease can be conceptualized as an extreme of the normal range or as a deviation from normal functioning.
• This approach enables inferences on how behavioral variables and diagnostic labels map onto biological data.