Abnormal permeability to proteins and glomerular lesions: a meta-analysis of experimental and human studies.

Whether abnormal protein traffic through the glomerular capillary is one of the possible causes of glomerular injury has been a matter of considerable controversy. Experimental and clinical evidence indicate an association between the two, and several explanations of potential pathophysiologic pathways are available that may explain protein-dependent forms of renal injury. However, none of the mechanisms suggested thus far have been definitely proven. A meta-analytic review was used to further explore the relationship between urinary protein excretion and glomerular injury in experimental and human proteinuric nephropathies. In experimental models a definite positive association (P < 0.0001) was found between urinary protein excretion and glomerulosclerosis. The overall effect size d for this relationship was 2.36 (95% confidence interval, 2.13 to 2.60), corresponding to a correlation coefficient of r = 0.76. Similarly, the correlation between albuminuria and glomerular sclerotic lesions was highly significant (overall d = 2.88; 95% confidence interval, 2.33 to 3.42; P < 0.0001; correlation coefficient, r = 0.82). In humans the meta-analytic correlation yielded analogous, although less striking, results (overall d = 0.58; 95% confidence interval, 0.43 to 0.73; P < 0.0001; correlation coefficient, r = 0.28). In conclusion, meta-analysis of data from studies examining urinary protein excretion revealed a significant positive correlation with glomerular sclerosis in both experimental models and human diseases. This study supports the role of abnormal protein traffic through the glomerular capillary as one of the possible causes of renal injury. However, the strong relationship found does not necessary establish a cause-and-effect relationship.