Key Points

Low kidney volume was a risk factor of incident CKD.A nonlinear relationship existed whereby individuals in the bottom tenth percentile of kidney volume exhibited exaggerated risk of CKD and albuminuria.Kidney volume could improve the classification of kidney disease risk.

Background

Low total kidney volume (TKV) is a risk factor of CKD. However, evaluations of nonlinear relationships, incident events, causal inference, and prognostic utility beyond traditional biomarkers are lacking.

Methods

TKV, height-adjusted TKV, and body surface area–adjusted TKV of 34,595 White British ancestry participants were derived from the UK Biobank. Association with incident CKD, AKI, and cardiovascular events were assessed with Cox proportional hazard models. Prognostic thresholds for CKD risk stratification were identified using a modified Mazumdar method with bootstrap resampling. Two-sample Mendelian randomization was performed to assess the bidirectional association of genetically predicted TKV with kidney and cardiovascular traits.

Results

Adjusted for eGFR and albuminuria, a lower TKV of 10 ml was associated with a 6% higher risk of incident CKD (hazard ratio, 1.06; 95% confidence interval [CI], 1.03 to 1.08; P = 5.8×10−6) in contrast to no association with incident AKI (hazard ratio, 1.00; 95% CI, 0.98 to 1.02; P = 0.66). Comparison of nested models demonstrated improved accuracy over the Chronic Kidney Disease Prognosis Consortium Incident CKD Risk Score with the addition of body surface area–adjusted TKV or prognostic thresholds at 119 (tenth percentile) and 145 ml/m2 (50th percentile). In Mendelian randomization, a lower genetically predicted TKV by 10 ml was associated with 10% higher CKD risk (odds ratio, 1.10; 95% CI, 1.06 to 1.14; P = 1.3×10−7). Reciprocally, an elevated risk of genetically predicted CKD by two-fold was associated with a lower TKV by 7.88 ml (95% CI, −9.81 to −5.95; P = 1.2×10−15). There were no significant observational or Mendelian randomization associations of TKV with cardiovascular complications.

Conclusions

Kidney volume was associated with incident CKD independent of traditional risk factors, including baseline eGFR and albuminuria. Mendelian randomization demonstrated a bidirectional relationship between kidney volume and CKD.