Accuracy of retinal changes in predicting microalbuminuria among elderly hypertensive patients: a cross-sectional study from a teaching hospital in South India

Abstract

ObjectiveMicroalbuminuria is an important marker of end organ damage in hypertensive patients, but is often not tested for, especially
in a resource-poor setting. We studied the accuracy of retinal changes in predicting microalbuminuria among a cohort of geriatric
hypertensive patients.

MethodsOne hundred and eighty elderly hypertensive patients aged more than 65years were assessed for their demographic characteristics,
smoking status, duration of hypertension, current severity of hypertension, body mass index, left ventricular hypertrophy
by electrocardiogram (ECG), and high sensitivity C-reactive protein (HsCRP). Optic fundi were assessed for retinopathy after
pupillary dilatation, and were photographed. Microalbuminuria (albumin–creatinine ratio) was measured as an average of two
nonconsecutive overnight spot urine samples. Patients with pre-diabetes, diabetes, metabolic syndrome, treatment with angiotensin
converting enzyme inhibitors and/or angiotensin receptor blockers, overt nephropathy or proteinuria, and active infection
were excluded.

ResultsMean age was 74±6.56years. One-third were obese and 18.9% had left ventricular hypertrophy. Prevalence of microalbuminuria
was 39.4% and prevalence of retinopathy was 40%. Microalbuminuria showed a strong association with retinopathy (P<0.0001). Logistic regression identified association of microalbuminuria with duration of hypertension (P=0.001) and elevated high sensitivity C-reactive protein (P=0.021). Retinopathy was associated with duration of hypertension (P=0.001) and smoking (P<0.0001). Tests of accuracy for retinopathy as a predictor of microalbuminuria showed a sensitivity of 72% and specificity
of 81%.

ConclusionPrevalence of microalbuminuria and retinopathy were quite high in our cohort of elderly hypertensive patients. Retinal changes
of any grade probably have moderate accuracy in predicting microalbuminuria and hence we can initiate work-up for target organ
damage, especially in a resource-poor setting.