Improvement in estimates of GFR by using fat-free mass as compared to body weight in Indians: pilot study

Background

Estimated glomerular filtration rate (eGFR) may be calculated by different equations with variable accuracy. The accuracy of creatinine-based eGFR equations may vary across ethnic groups. These are also influenced by differences in body composition. Many populations have higher adiposity for a given body mass index, this disparity is very striking in Indians and has led to the description of a ‘thin-fat’ Indian.

We undertook a pilot study to compare eGFR estimated from clinical equations using fat-free mass instead of total body weight, in healthy Indians.

Methods and results

We studied 26 healthy Indian volunteers (11 men, 15 women), aged 49 (34,55) years [Median (Min, Max)]. We recorded vitals, anthropometric and biochemical measurements. eGFR values were estimated by Cockcroft-Gault (CG-BSA), Modification of Diet in Renal Disease (MDRD), and Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formulae. eGFR values based on the CG-BSA equation were re-calculated after replacing coefficient of total body weight with fat-free mass and muscle mass (obtained by BIA) and lean mass (obtained by DEXA scan).

We used Tc 99m DTPA renogram for GFR measurement as the gold standard.

Median eGFR using conventional formulae overestimated the GFR when compared to mGFR (81 ml/min/1.73m ² ); CG-BSA equation (99 mL/min/1.73 m ² ); EPI-CKD equation (97 mL/min/1.73 m ² ) and MDRD equation (89 mL/min/1.73 m ² ).

Of the CG-BSA formulae, eGFR using total Body weight is the farthest from gold standard. And estimated GFR using fat free mass had closest median (67 mL/min/1.73 m ² ) to the gold standard (79 mL/min/1.73 m ² )

Bland-Altman analysis showed the Mean difference of Fat free mass-based CG-BSA formula from the gold standard was the lowest of all four CG formulae (−12.009, CI -19.266 to -4.752); showing that it had the best agreement with the gold standard of the four CG-BSA formulae. Also, it showed that the mean difference of MDRD formula from gold standard is the lowest (9.191, CI 1.008 to 17.375) followed by Cockcroft Gault formula using Fat Free mass (−12.009, CI -19.266 to -4.752).

Mean difference of EPI formula (15.151, CI 8.182 to 22.120) and Cockcroft Gault formula using Muscle mass (−15.809, CI -22.756 to -8.861) follow, with very similar Limits of Agreement (LOA).

Conclusion

This pilot study showed that existing conventional eGFR equations CG-BSA, MDRD, and CKD-EPI overestimate eGFR in healthy Indian subjects. The conventional CG-BSA had the least agreement with measured GFR by the gold standard. However, the replacement of coefficient of total body weight by FFM, MM and LM in the CG-BSA formula improved the estimates of eGFR in healthy Indian volunteers.

Significance Statement

Estimation of glomerular filtration rate (eGFR) is an important practice in clinical medicine and there are various equations available to obtain it. eGFR varies with serum creatinine and body composition which is different for the Indian population as compared to the African and Caucasian population. There has been no validation of the commonly used equations in the Indian population. Usage of coefficients based on body composition such as fat-free mass, lean mass or muscle mass to calculate the eGFR may be explored in the Indian population in view of the presence of lower muscle mass and higher adiposity in them, as per the concept of “thin-fat Indian”