A review of epidemiologic studies of nonnarcotic analgesics and chronic renal disease.

The relationship of long-term and heavy exposure of nonnarcotic analgesics to the risk of chronic renal disease (CRD) has been the object of intensive clinical, pharmacologic, toxicologic, and epidemiologic research for 4 decades. The clinical evidence of an increased risk has been suggestive but inconclusive. The experimental evidence in animal models has been inconsistent, and in any case it cannot be generalized to humans. The epidemiologic evidence has been unsatisfactory for the most part: most of the early studies had severe methodologic limitations; moreover, they related mainly to phenacetin-containing drugs and did not have useful information on other analgesics. Since 1980, 9 analytical epidemiologic studies have attempted to confirm that a causal relationship exists between phenacetin or other analgesics and CRD. In the aggregate, despite methodologic flaws, this work suggests that excessive use of phenacetin-containing analgesics probably causes renal papillary necrosis and interstitial nephritis. In contrast, there is no convincing epidemiologic evidence that nonphenacetin-containing analgesics (including acetaminophen, aspirin, and mixtures of these two compounds) or that nonsteroidal antiinflammatory drugs cause CRD. Moreover, the nature of dose-response relationships, the types of renal disease possibly caused by analgesics, and the cofactors that might be related both to analgesic use and to the development of CRD in humans are still uncertain, and the pathologic mechanisms of analgesic-induced CRD in humans remain unclear. It may take many years before all the outstanding issues are settled. Until they are, as a matter of good clinical judgment it would be prudent to consider all analgesics as potentially nephrotoxic and, as much as possible, to avoid excessive, protracted use.