[Development of a statistical metabolic syndrome simulator]

Area of Science:

• Cardiovascular disease research
• Public health policy
• Biostatistics

Background:

• Metabolic Syndrome (Mets) is a cluster of risk factors including visceral obesity, hyperglycemia, hypertension, and dyslipidemia.
• Current diagnostic criteria and definitions for Mets vary internationally, leading to inconsistent medical evaluations.
• There is a need for unified tools to assess Mets prevalence across different populations and criteria.

Purpose of the Study:

• To develop a Metabolic Syndrome (Mets) simulator.
• To calculate Mets prevalence based on various diagnostic criteria using group checkup data.
• To provide a tool for informing health policies related to lifestyle-related diseases.

Main Methods:

• Utilized actual checkup data to analyze characteristics of diagnostic criterion components (waist circumference, BMI, triglycerides, HDL, blood pressure, blood sugar, HbA1c).
• Applied normal transformation methods and calculated variance-covariance matrices for transformed variables.
• Generated multivariate normal random values to simulate Mets prevalence across five criteria (NCEP-ATPIII, AHA/NHLBI, IDF, Japan, MHLW).

Main Results:

• A Mets simulator was successfully constructed and validated using checkup data from Tokyo.
• Performance and sensitivity analyses were conducted, including examination of cases with missing diagnostic components.
• The simulator estimated prevalence, mean/variance values, and missing data, revealing similarities between Japanese and IDF criteria.

Conclusions:

• A simulator capable of assessing Mets prevalence across five diagnostic criteria was developed.
• The simulator allows for the examination of prevalence changes in response to variations in diagnostic criterion components.
• This tool offers valuable insights for public health policies aimed at preventing lifestyle-related diseases.