Based on a publicly posted research preprint co-authored by Dr Arash Ardalan.
The research question
Do associations with body measurements, parental obesity and physical activity vary across the body fat distribution?
Study context
The preprint describes a cross-sectional sample of 2,873 students aged 14–20 from school-based sampling in two cities in southern Iran. Body measurements, body fat percentage, parental obesity and physical activity were recorded.
Statistical approach
- Use quantile regression to model selected conditional percentiles of body fat percentage.
- Examine anthropometric measures and parental-obesity and physical-activity variables in the reported models.
- Describe percentile-specific associations separately for girls and boys.
What the analysis provided
The manuscript reported associations across multiple percentiles, showing how a distributional analysis can reveal patterns that a single average-based summary may overlook.
The cross-sectional design and self-reported family and activity information limit causal interpretation. This case summarises the posted preprint version.
Practical lessons for your project
- Choose meaningful percentiles
- Select quantiles that answer the research question and are supported by the available data.
- Interpret the model precisely
- Explain results as changes in a conditional outcome percentile for given predictor values.
- Report uncertainty
- Present estimates with appropriate uncertainty and explain the limits of the sampling and measurements.
Original preprint
Arash Ardalan; Seyyed Taghi Heydari; Marzieh Alamolhoda
Associations Between Anthropometric Measures and Body Fat Percentage in Iranian Adolescents: a Quantile Regression Analysis
Research Square · Preprint, version 1 · 2021