Objective  To analyze the correlation between kidney size and renal function in community patients with chronic kidney diseases (CKD), and to provide a reference for prognosis evaluation of CKD patients.

  Methods  With random cluster sampling, 9 257 permanent residents aged 20 – 75 years were recruited in a town of Songjiang district, Shanghai for a survey including questionnaire survey, physical examination and laboratory test from June 2016 to November 2017. From 8 207 residents completing the survey, 914 CDK patients without unilateral nephrectomy were identified. Then, the CDK patients′ kidney size were measured with B-mode ultrasonography and estimated glomerular filtration rate (eGFR) was adopted to assess the patients′ renal function. Pearson correlation analysis was used to analyze the relationship between kidney size and eGFR, and multivariate logistic regression and areas under the receiver operating characteristics curve (AUC) were used to compare the efficiency of different models for renal function decline prediction.

  Results  Of all the patients, 706 (77.2%) had normal renal function and 208 (22.8%) had decreased renal function. Compared to those in the patients with normal renal function, significantly shorter length and width of both kidneys were measured in both male and female patients with decreased renal function (all P < 0.001). Pearson correlation analysis revealed significantly positive correlations of left renal length (r = 0.278) and width (r = 0.268), right renal length (r = 0.263) and width (r = 0.134) with eGFR (all P < 0.001) and the correlation between left kidney size and eGFR was the strongest after adjusting for body surface area (r = 0.411, P < 0.001). Multivariate logistic regression models for kidney size-related renal function decline prediction were constructed with corrected kidney size and other diagnostic indicators. Compared with other models, the model 3 with the independent variables of corrected kidney size, gender, age and hyperuricemic state is of strongest efficiency in predicting renal function decline, with an AUC of 0.836 (95% confidence interval: 0.810 – 0.859).

  Conclusion  There is a significant correlation between kidney size and renal function in community CKD patients. Kidney size combined with other related factors are important indicators for predicting the decline of renal function and could be employed to evaluate the prognosis of CKD patients.