# Review no. 2: a beginner’s guide for calculating eGFR slope using linear mixed-effects model in R—step-by-step methods and code examples

## TL;DR

This paper provides a beginner-friendly guide on calculating eGFR slope using linear mixed-effects models in R, with step-by-step methods and code examples.

## Contribution

The paper introduces a practical, hands-on approach to using linear mixed-effects models for eGFR slope calculation in R.

## Key findings

- Linear mixed-effects models are more efficient than individual linear regression models for calculating eGFR slope.
- The paper demonstrates how to calculate individual eGFR slopes and compare mean slopes between groups using R.
- The method accounts for both population-level and subject-specific variations in eGFR data.

## Abstract

Estimated glomerular filtration rate (eGFR) slope, which quantifies the annual change in kidney function using longitudinal measurements of eGFR, has emerged as a critical surrogate endpoint for chronic kidney disease progression in clinical trials. Accordingly, there is an increasing number of observational studies calculating the eGFR slope as an outcome or exposure of interest. To calculate the eGFR slope, linear mixed-effects models are considered statistically more efficient and therefore recommended over linear regression models constructed for each individual. Linear mixed-effects models can account for both fixed effects at the population level and random effects, which allow for subject-specific intercepts and slopes—e.g. individual-specific variation from the population mean. We introduce the practical implications of linear mixed-effects models using R programming: calculating an individual’s eGFR slope and comparing mean eGFR slopes between groups. This article is based on the hands-on seminar at the annual meeting of the 68th Japanese Society of Nephrology in 2025.

The online version contains supplementary material available at 10.1007/s10157-026-02835-8.