Objectives: To assess the relation between renal function and delirium and to assess and compare the relation between cerebral white matter lesion (WML) and renal function as estimated by three formulas for the estimated glomerular filtration rate (eGFR) in older adult hospitalized veterans with and without delirium.
Methods: Commonly used formulas to assess renal function-the four-variable Modification of Diet in Renal Disease (MDRD), the six-variable MDRD, and the Cockcroft-Gault eGFR equations-were used to assess renal function in 100 older adult hospitalized veterans with delirium (delirium group) and 100 hospitalized veterans without delirium (nondelirium group) that were age, sex, and race matched. WML location and volumes were assessed using brain computed tomography imaging for each of the 200 veterans in the study. One radiologist, blinded to the diagnoses of the veterans, examined head computed tomography scans for WML in the cortex, subcortex (frontal, temporal, parietal, occipital lobes), basal ganglia (globus pallidus, caudate, putamen), and internal capsule. WML were graded as not present, <1 cm, 1 to 2 cm, or >2 cm. Exploratory χ2 analyses were used to determine the association between the stage of chronic kidney disease and WML. Simple logistic regression analyses were then used to estimate the strength of association between the stages of kidney disease and WML for particular regions of the brain.
Results: The mean age of delirium group and nondelirium group veterans was 66 years. χ2 tests revealed no reliable relation between stages of renal disease and delirium. χ2 exploratory analyses of WML in brain regions by renal disease stages demonstrated significant differences in associations among the MDRD-4, MDRD-6, and Cockcroft-Gault formulas for measuring eGFR. The MDRD-4 formula was least associated with the presence or absence of WML. The Cockcroft-Gault estimation of eGFR was most associated with the presence or absence of WML. Simple logistic regressions showed notable increases in the association between stages of renal failure and WMLs in specific areas of the brain, with the MDRD-4 being the least associative with the fewest specific areas and the Cockcroft-Gault formula being the most associative with the most specific areas.
Conclusions: The association between stages 2 through 5 of chronic kidney disease and WLM support the role of kidney function as a potential risk factor for WML in older adult military veterans. The Cockcroft-Gault formula is an important renal index of suspected WML and renal stages 2 through 5, superior to the MDRD-6 and MDRD-4, respectively, in association with WML in older adult military veterans.