A Novel Depression Risk Prediction Model Using NHANES Data With Mendelian Randomization Validation

Area of Science:

• Biochemistry
• Epidemiology
• Medical Informatics

Background:

• Depression poses a significant public health challenge, yet effective screening tools using routine clinical indicators are scarce.
• Developing accessible depression screening methods is crucial for early intervention and improved patient outcomes.

Purpose of the Study:

• To develop and validate a practical depression risk prediction model using readily available biochemical markers.
• To facilitate widespread early depression screening and timely intervention in general clinical settings.

Main Methods:

• Utilized data from the National Health and Nutrition Examination Survey (NHANES) for model development and validation.
• Employed the Mendelian randomization (MR) approach to investigate causal relationships between biochemical markers and depression.
• Developed and compared two prediction models using LASSO and multivariate logistic regression, selecting the more parsimonious Model 2 (14 variables).

Main Results:

• Model 2 demonstrated comparable predictive performance to a more complex model across multiple statistical metrics.
• Mendelian randomization analysis confirmed bidirectional relationships between specific biomarkers and depression.
• Elevated body mass index was associated with increased depression risk (OR: 1.061). Depression correlated with higher alkaline phosphatase (ALP) and lower blood urea nitrogen (BUN) and total bilirubin (TB) levels.

Conclusions:

• The validated Model 2 offers a pragmatic and accurate tool for large-scale depression screening in clinical settings.
• This model's simplicity and predictive power support timely intervention and therapeutic strategies for depression.