BackgroundCompared with general obesity, abdominal obesity has been shown to be more closely correlated to chronic diabetic complications such as cardiovascular disease and diabetic retinopathy. The association between obesity and diabetic kidney disease is unclear. We did two independent studies, one cross-sectional study and one 5 year prospective study, to investigate the association of general and abdominal obesity with diabetic kidney disease. MethodsIn the cross-sectional study in a hospital in Chongqing, China, body composition was assessed with dual-energy X-ray absorptiometry in 1016 patients with type 2 diabetes. Variables measured to define general obesity were BMI, total body fat percentage, and fat-mass index (ratio of fat mass to height squared). Abdominal obesity variables were waist circumference, waist-to-height ratio, and visceral adipose tissue. Diabetic kidney disease was defined as chronic kidney disease stage 3–5 (estimated glomerular filtration rate [eGFR] <60 mL/min per 1·73 m2). In the prospective study, 279 patients with type 2 diabetes without diabetic kidney disease at baseline were followed up for 5 years after initial assessment in a community centre. BMI, waist circumference, waist-to-height ratio, and waist-to-hip ratio were used as indicators of obesity type in the prospective study. Obesity-related variables were split into low, median, and high tertiles and patients were stratified by values accordingly for both studies. Ethical approval was granted by the Ethics Committee of the First Affiliated Hospital of Chongqing Medical University. Informed consent was provided by all participants. Multivariate logistic analysis was done to determine whether parameters of obesity could be risk factors for diabetic kidney disease and adjustments made for confounding factors. The odds ratios (ORs) of general obesity parameters and abdominal obesity parameters were calculated, with reference to the lowest tertile of those parameters; OR was 1. FindingsIn the cross-sectional study, 470 patients had stage 1 chronic kidney disease, 374 patients had stage 2 disease, and 172 patients had stages 3–5 disease. Participants with higher BMI (24·67 kg/m2 [SD 3·33] in those with stage 1 chronic kidney disease vs 24·78 kg/m2 [3·14] in those with stage 2 chronic kidney disease vs 25·40 kg/m2 [3·18] in those with stages 3–5 chronic kidney disease; p=0·040), total body fat percentage (29·21% [6·17] vs 29·57% [6·16] vs 30·74% [6·18]; p=0·024), fat-mass index (7·31 kg/m2 [2·21] vs 7·43 kg/m2 [2·21] vs 7·89 kg/m2 [2·14]; p=0·014), waist circumference (92·71 cm [9·31] vs 93·98 cm [9·08] vs 95·52 cm [8·88]; p=0·047), waist-to-height ratio (0·57 [IQR 0·54–0·61] vs 0·58 [0·55–0·62] vs 0·61 [0·56–0·65]; p<0·0001), and visceral adipose tissue (117·59 cm2 [43·34] vs 121·67 cm2 [43·10] vs 133·31 cm2 [45·01]; p=0·0022) were more likely to have a lower eGFR, compared with those with lower values. Logistic regression analyses showed that variables of general obesity were associated with a risk of diabetic kidney disease (in the median vs high tertiles, for BMI ptrend=0·0371; for total body fat percentage ptrend=0·044; for fat-mass index ptrend=0·0053). However, there was no correlation when visceral adipose tissue was adjusted (in the median vs high tertiles, for BMI ptrend=0·18; for total body fat percentage ptrend=0·66; for fat-mass index ptrend=0·26). In the prospective study, 41 patients developed diabetic kidney disease after 6 years follow-up. No association was noted between BMI and risk of diabetic kidney disease or in the crude or adjusted models (in the crude model 1·11 [0·49–2·50] in the median tertile vs 1·10 [0·48–2·47] in the high tertile, p=0·81; in the waist-to-height ratio-adjusted model 0·43 [0·26–1·18] vs 0·76 [0·32–1·80], p=0·24; in the multivariable model 0·44 [0·13–5·54] vs 1·33 [0·25–6·97], p=0·40). Abdominal obesity variables were associated with risk of diabetic kidney disease after adjustment for BMI (for waist circumference 1·39 [1·10–8·03] in the median tertile vs 2·59 [1·45–9·31] in the high tertile, p=0·022; for waist-to-hip ratio 2·63 [0·71–5·79] vs 5·16 [1·38–9·31], p=0·015; for waist-to-height ratio 3·35 [0·86–7·08] vs 5·96 [1·51–13·06], p=0·03). The associations between the risk of diabetic kidney disease and waist circumference, waist-to-hip ratio, waist-to-height ratio were the same after adjustment for other confounders. InterpretationAbdominal obesity is more closely associated with the risk of diabetic kidney disease than general obesity in China. FundingNational Key Clinical Specialties Construction Program of China, the National Natural Science Foundation of China and The Fundamental Science and Advanced Technology Research of Chongqing.