Expected Estimation Errors in Studies of the Cortisol Awakening Response: A Simulation

Area of Science:

• Biomedical research
• Psychoneuroendocrinology
• Biostatistics

Background:

• Biomarker studies are susceptible to misestimation due to small sample sizes, small effect sizes, and measurement noise.
• Inaccurate biomarker estimates can negatively impact critical design decisions in research.

Purpose of the Study:

• To investigate the probability of misestimation in biomarker studies using real-world designs.
• To assess the impact of sample size and measurement noise on the accuracy of effect estimates.

Main Methods:

• A simulation study was conducted using 212 study designs from a meta-analysis on psychosocial correlates of cortisol awakening response.
• 100,000 simulated datasets were generated for each design to calculate error probabilities.

Main Results:

• Small sample sizes (n < 100) and noisy measurements significantly increased the probability of errors.
• The average probability of an effect being in the wrong direction was approximately 20%, reaching up to 40% in some designs.
• Misestimation probabilities averaged around 40%, with some designs exhibiting up to 80% error, even for statistically significant effects.

Conclusions:

• Current study designs in biomarker research often yield inaccurate estimates.
• Recommendations for improving study design are provided to enhance the accuracy of biomarker estimates.