|
|
| |||||||||||||||||||||||||||||||||||||||||||||||
Background Several epidemiologic studies have demonstrated an association between heavy consumption of nonnarcotic analgesics and the occurrence of chronic renal failure, but it is unclear which is the cause and which is the effect.
Methods In a nationwide, population-based, case–control study of early-stage chronic renal failure in Sweden, face-to-face interviews were conducted with 926 patients with newly diagnosed renal failure and 998 control subjects, of whom 918 and 980, respectively, had complete data. We used logistic-regression models to estimate the relative risks of disease-specific types of chronic renal failure associated with the use of various analgesics.
Results Aspirin and acetaminophen were used regularly by 37 percent and 25 percent, respectively, of the patients with renal failure and by 19 percent and 12 percent, respectively, of the controls. Regular use of either drug in the absence of the other was associated with an increase by a factor of 2.5 in the risk of chronic renal failure from any cause. The relative risks rose with increasing cumulative lifetime doses, rose more consistently with acetaminophen use than with aspirin use, and were increased for most disease-specific types of chronic renal failure. When we disregarded the recent use of analgesics, which could have occurred in response to antecedents of renal disease, the associations were only slightly attenuated.
Conclusions Our results are consistent with the existence of exacerbating effects of acetaminophen and aspirin on chronic renal failure. However, we cannot rule out the possibility of bias due to the triggering of analgesic consumption by predisposing conditions.
Previous case–control studies evaluating analgesic use in relation to chronic renal failure have had methodologic shortcomings,9,10,11,12 including a failure to identify patients early enough in the course of their disease to ensure that the disease itself had not led to a change in the use of analgesics; a failure to specify diagnostic criteria; a failure to adjust for the use of other analgesics; incompleteness of data on exposure; and the use of proxy respondents. To investigate whether acetaminophen and aspirin affect the development of chronic renal failure, we conducted an analysis as part of a larger population-based case–control study in Sweden in which we tried to avoid such shortcomings.
Methods
Study Subjects
The Swedish Population Register13 provided a well-defined data base of all 5.3 million people born in Sweden, 18 to 74 years of age, who were living in the country during the period from May 20, 1996, through May 31, 1998 (the ascertainment period).
Eligible patients were men whose serum creatinine level exceeded 3.4 mg per deciliter (300 µmol per liter) for the first time or women whose serum creatinine level exceeded 2.8 mg per deciliter (250 µmol per liter) for the first time. To help us identify patients, medical laboratories provided monthly lists of serum creatinine measurements. Physicians who treated patients with renal disease determined patients' eligibility for the study by reviewing the medical records of patients with elevated serum creatinine levels. Patients with chronic renal failure whose cause was prerenal (e.g., severe heart failure) or postrenal (i.e., obstruction of the urinary tract) and patients who had received kidney transplants were excluded. A second creatinine measurement, three months after the first, was obtained when the chronic nature of the renal failure was uncertain. To allow for day-to-day variation, the thresholds for eligibility with this second measurement were lower (2.8 mg per deciliter for men and 2.3 mg per deciliter [200 µmol per liter] for women); patients with lower values were excluded from the study. The diagnosis of underlying disease was based on the results of routine clinical evaluation.
The controls were randomly selected throughout the ascertainment period from the Swedish Population Register13 and were frequency-matched to the patients with renal failure according to age (in 10-year age groups) and sex. The ethics committees of the participating centers and the Swedish Data Inspection Board approved the study protocol. All study subjects provided oral informed consent before being enrolled.
Collection of Data
Subjects received a mailed, self-administered questionnaire and later underwent a face-to-face, computer-assisted interview. The interviewers, from Statistics Sweden (a government agency), were unaware of the study hypotheses and were trained to interview subjects in a standardized manner. The interviewers also ensured that the answers to the mailed questionnaire were complete. Interviews with patients with renal failure lasted an average of 80 minutes, and interviews with control subjects lasted an average of 70 minutes.
Lifetime Exposure to Analgesics
Each subject reviewed a booklet that included color pictures of the packaging of all analgesics containing acetaminophen or phenacetin as well as most other frequently sold nonnarcotic analgesics (78 major brands of the 174 that were on the Swedish market between 1960 and 1996). Subjects reported their lifetime consumption of the brands of drugs that appeared in the booklet as well as their lifetime consumption of aspirin and any other analgesics. Information about the consumption of brand-name drugs was converted to amounts of generic drug ingredients.
Regular use of an analgesic was defined as use at least twice a week for two months. All subjects were asked to report their age during each period of regular use, the duration of regular use, and the dose used during that period. Subjects whose cumulative lifetime dose exceeded 20 tablets of an analgesic but who did not use it regularly were classified as sporadic users. Nonusers were defined as those who reported taking a total of fewer than 20 tablets during their lifetime. Subjects also answered questions about changes in their pattern of use and in the patterns of aches and pains prompting the use of analgesics.
Statistical Analysis
We used unconditional logistic regression to model odds ratios and 95 percent confidence intervals as measures of the association between analgesic use and chronic renal failure while controlling for potential confounders. Nonusers of a given analgesic served as the reference category for all comparisons related to that analgesic. We initially considered marital status; body-mass index (the weight in kilograms divided by the square of the height in meters); consumption of caffeine, alcohol, and illicit drugs; and the presence of hypertension, angina, claudication, kidney stones, or gout as potential confounders. The final model was based on the scientific literature and the statistical significance of explanatory variables in the model, as assessed by the likelihood-ratio test.14 The model contained terms for sex, age (in 10-year age groups), smoking status (lifetime use of <100 or 
100 cigarettes), level of education (
9 years, 10 to 12 years, or 13 years), regular use of other analgesics, and when applicable, a term for the interaction between acetaminophen use and aspirin use. We excluded from the model 23 patients with renal failure (2.5 percent) and 29 control subjects (2.9 percent) because of missing information for one or more covariates.
To assess the dose–response relation, we performed tests for trend on groups of ordinal variables by assigning each group a score equal to the mean value, fitting the resulting scores into the model,15 and assessing statistical significance by the likelihood-ratio test.
We examined the associations between analgesic use and overall chronic renal failure and between analgesic use and disease-specific types of chronic renal failure. Results for both sexes are combined, since associations for men did not differ materially from those for women. To exclude the possibility that analgesic use was prompted by conditions that were precursors to renal failure, we performed analyses in which reported use during the 5 to 10 years before the interview was disregarded.
Results
We identified 1189 eligible patients of whom 69 (6 percent) died shortly after they were determined to be eligible for the study. The remaining 1120 patients were approached, and 926 participated. A total of 918 patients provided information on exposure to analgesics (77 percent of the 1189 eligible patients). A total of 111 patients declined to participate, and 83 had severe diseases that precluded participation. Of 1330 eligible control subjects, 998 participated. A total of 980 controls provided information on exposure to analgesics (74 percent of the eligible controls); 221 declined to participate, 56 could not be reached, and 55 had diseases that precluded participation. Half of the patients with chronic renal failure were interviewed within 1 month of inclusion, and 95 percent were interviewed within 12 months.
Characteristics of the Study Subjects
Sixty-five percent of the subjects were men. Among both patients and control subjects, the mean age of the men at the time of the interview was 58 years, and the mean age of the women was 57 years. The patients with chronic renal failure lived alone more often than the control subjects did, had fewer years of education, were somewhat more likely to smoke, and were less likely to drink alcohol.
The distribution of patients with chronic renal failure according to underlying disease is shown in Table 1. The median serum creatinine levels were 3.8 mg per deciliter (336 µmol per liter) among men and 3.2 mg per deciliter (281 µmol per liter) among women. The estimated glomerular filtration rate16 ranged from 2 to 53 ml per minute. The median values for men and women were 22 and 19 ml per minute, respectively.
|
|
Acetaminophen Use in the Absence of Regular Aspirin Use
Among subjects who did not use aspirin regularly, the regular use of acetaminophen was associated with a risk of chronic renal failure that was 2.5 times as high as that for nonusers of acetaminophen. The risk increased with an increasing cumulative lifetime dose (P for trend <0.001) (Table 3). The average dose used during periods of regular acetaminophen use also correlated with risk (P for trend <0.001) so that those who took 500 g or more per year (1.4 g per day) during periods of regular use had an odds ratio for chronic renal failure of 5.3 (95 percent confidence interval, 1.8 to 15.1). The duration of use was unrelated to risk (data not shown). The odds ratios associated with regular use were greater than 1.0 for all types of renal failure (Table 4), but they were significant only for renal failure classified as diabetic nephropathy and that associated with systemic disease or vasculitis. Dose-related trends in risk were significant for all diagnoses except glomerulonephritis and nephrosclerosis (data not shown).
|
|
Among subjects who did not use acetaminophen regularly, the regular use of aspirin was associated with a risk of chronic renal failure that was 2.5 times as high as that for nonusers of aspirin (Table 3). This risk increased significantly with an increasing cumulative lifetime dose of aspirin (P for trend=0.01) and with an increasing average dose during periods of regular use (P for trend=0.004), but not with an increasing duration of use. Among those with an average intake of 500 g or more of aspirin per year during periods of regular use, the odds ratio for chronic renal failure was 3.3 (95 percent confidence interval, 1.4 to 8.0). The strongest association was found between aspirin use and chronic renal failure classified as associated with "other renal disease," but the odds ratio was greater than 2.0 for all types of chronic renal failure except that linked to systemic disease or vasculitis (Table 4). The dose-related trends in risk, however, were nonsignificant for all types except glomerulonephritis (P=0.008) (data not shown).
Acetaminophen Use in Addition to Aspirin Use
The odds ratio for chronic renal failure among regular users of both acetaminophen and aspirin was 2.2 (95 percent confidence interval, 1.4 to 3.5) when regular aspirin users served as the reference group. The trend toward greater risk with an increasing cumulative lifetime dose of acetaminophen was statistically significant (P=0.03), with a risk that was 2.4 times as high (95 percent confidence interval, 1.4 to 4.4) for subjects who had consumed a total of more than 500 g of acetaminophen than for those who had used aspirin only. The types of chronic renal failure most strongly associated with regular use of acetaminophen in addition to aspirin were renal failure linked to diabetes (odds ratio, 2.8; 95 percent confidence interval, 1.5 to 5.4) and renal failure linked to systemic disease or vasculitis (odds ratio, 5.1; 95 percent confidence interval, 1.5 to 17.6), but estimates of relative risks of approximately 2.0 were found for all types of chronic renal failure (data not shown).
Aspirin Use in Addition to Acetaminophen Use
The odds ratio for chronic renal failure was 1.6 (95 percent confidence interval, 0.9 to 2.7) among regular users of both aspirin and acetaminophen as compared with users of acetaminophen only. The trend toward greater risk with an increasing cumulative lifetime dose of aspirin was significant (P=0.02). The pattern of associations between aspirin use and specific types of chronic renal failure among users of acetaminophen was similar to that found among aspirin users without regular acetaminophen use, but there were lower point estimates for the risk associated with diabetic nephropathy (odds ratio, 1.7; 95 percent confidence interval, 0.8 to 3.4) and that associated with chronic renal failure classified as other renal disease (odds ratio, 1.4; 95 percent confidence interval, 0.5 to 3.7).
Use of Other Analgesics
More patients than control subjects used other nonnarcotic analgesics as well as aspirin and acetaminophen (Table 2), but after adjustment for acetaminophen use and aspirin use, no associations between the use of these other drugs and the risk of chronic renal failure remained. The adjusted odds ratios associated with the regular use of propoxyphene, nonsteroidal antiinflammatory drugs other than aspirin, codeine, and pyrazolones were 1.0, 1.0, 0.9, and 1.3, respectively.
Self-Reported Pain
More patients than control subjects reported having headaches (12 percent vs. 7 percent), joint pain (8 percent vs. 5 percent), and toothaches (4 percent vs. 2 percent). Headaches occurred most often among patients with hereditary renal disease (20 percent), "other" renal disease (17 percent), diabetic nephropathy (12 percent), and systemic disease or vasculitis (15 percent). As expected, many patients in the last category (30 percent) also reported joint pain. No clear relation was found between the frequency of self-reported pain or any quantitative measure of analgesic use and the glomerular filtration rate in our patients (data not shown).
Analyses Based on Different Periods of Use
Symptoms associated with renal disease may prompt the use of analgesics. Indeed, a greater proportion of patients than of control subjects had begun to use analgesics regularly (35 percent vs. 25 percent for acetaminophen and 37 percent vs. 28 percent for aspirin) within five years before the interview. We therefore performed additional analyses in which we disregarded the most recent exposure. The odds ratios for chronic renal failure associated with either acetaminophen use or aspirin use were only slightly attenuated in these analyses (Figure 1). The attenuation among subjects using both analgesics was moderate (20 percent) and was similar for acetaminophen and aspirin (data not shown).
|
To separate the effects of drugs from the effects of the underlying disease, we analyzed the association between analgesic use and the risk of renal failure among 67 control subjects with diabetes and 324 patients with chronic renal failure due to diabetes. The odds ratio associated with regular acetaminophen use in the absence of regular aspirin use was 4.0 (95 percent confidence interval, 1.0 to 16.1) and that in the presence of regular aspirin use was 2.4 (95 percent confidence interval, 0.7 to 8.1). The odds ratios associated with regular aspirin use in the absence and presence of regular acetaminophen use were 1.4 (95 percent confidence interval, 0.7 to 3.1) and 0.6 (95 percent confidence interval, 0.1 to 3.2), respectively.
Discussion
In this study, as in others,6,8 regular use of either acetaminophen or aspirin or of both was associated in a dose-dependent manner with an increased risk of chronic renal failure. The odds ratios among regular users exceeded 1.0 for all types of chronic renal failure, albeit not always significantly. The regular use of other nonnarcotic analgesics was not associated with an increased risk of chronic renal failure in general.
The study base was well defined, and objective diagnostic criteria were used to identify renal failure. In contrast to previous case–control studies,7,8,17,18,19,20 our study enrolled patients with chronic renal failure relatively early in the course of their disease, so the reported extent of exposure to analgesics is more likely to pertain to a causally relevant period. Photographs of products and their packaging were used to improve the reliability of subjects' recall.21
A recent study conducted in the Physicians' Health Study cohort revealed no positive association between analgesic use and the risk of moderate renal dysfunction (serum creatinine concentration, >1.5 mg per deciliter [133 mmol per liter]); in fact, acetaminophen use appeared to be slightly protective.22 That study may have reduced the possibility of recall bias by ascertaining information about exposure to analgesics before renal dysfunction was diagnosed. Nonetheless, its results must be interpreted cautiously. The level of exposure to analgesics was ascertained retrospectively in that study, after a follow-up period of 14 years, and serum creatinine testing was performed no longer than 1 to 2 years after the assessment of exposure and involved only half of the initial cohort. We used considerably higher cutoff levels of serum creatinine to define renal disease. Preexisting renal or systemic disease was present in all our patients, suggesting that such disease has an important role in causing analgesic-associated chronic renal failure. Subjects with such conditions were unlikely to be recruited into the Physicians' Health Study or to remain in that study until the end of the follow-up period. Although the Physicians' Health Study may have had insufficient power to detect a true association between the heavy use of analgesics and the risk of clinically significant renal failure,23 it supports the finding that persons without preexisting disease who use analgesics have only a small risk of end-stage renal disease.
We found no indications that the severity of the renal failure affects analgesic consumption. Symptoms of diseases that predispose patients to renal failure may, however, lead to an increased use of analgesics, thus possibly introducing a protopathic (reverse causality) bias24 into our assessment of the effects of analgesics. Overall, however, the associations between the use of analgesics and the risk of chronic renal failure were not consistently stronger among subjects with underlying diseases causing frequent aches and pains. In addition, analyses restricted to subjects with diabetes produced risks associated with acetaminophen use that were similar in magnitude to those found in analyses of all patients and control subjects. Hence, although protopathic bias cannot be ruled out, its effect is likely to be small.
A strength of our study is the analyses based on different periods of exposure, in which we tried to separate analgesic use of causal importance from that possibly triggered by the renal disease or its prodromal symptoms. These analyses resulted in only minor reductions in the estimates of relative risk. Such analyses must, however, be interpreted cautiously,25,26,27 particularly when any potential latency time between the exposure to analgesics and chronic renal failure is unknown.
Acetaminophen use and aspirin use had similar effects on the overall risk of chronic renal failure. This lack of specificity might suggest noncausal associations. However, in experiments in animals, both analgesics have induced irreversible renal injury.28,29 In addition, increased risks were not observed with analgesics other than acetaminophen and aspirin.
The possibility of detection bias must be considered. Patients with diabetes or systemic disease, particularly those with substantial pain, would most likely undergo serum creatinine tests more often than healthy controls. We set the threshold for the serum creatinine concentration at a level at which renal failure is clinically significant and typically leads to referral to a nephrologist; this case definition helps to allay any concern about detection bias.
Blinding the interviewers to the case or control status of subjects was deemed impossible. This weakness constitutes a possible source of bias. However, the interview covered exposure to a multitude of substances, and interviewers were unaware of the study hypotheses. Moreover, they were trained to follow the questionnaire carefully in a standardized manner regardless of the subject's case or control status. The somewhat longer interview time for the patients with chronic renal failure than for the controls may be explained by their generally poorer condition, their more complicated medical history, and their greater use of drugs. Therefore, we do not believe that information bias has affected our results in any important way.
Although our study was not designed to confirm or refute the existence of a specific disease entity attributed exclusively to the use of analgesics, it should be noted that analgesic nephropathy, as defined by Elseviers,30 was not observed in the population we studied. Misdiagnosed cases, if there were any, were likely to be assigned to the "other renal disease" category.31 However, the associations with acetaminophen use and aspirin use did not stand out as exceptionally strong among the subjects in this category.
In conclusion, our results are consistent with exacerbating effects of acetaminophen and aspirin on chronic renal failure, practically regardless of accompanying disease. However, it is impossible to rule out bias caused by the consumption of these analgesics for symptoms of the conditions that predisposed patients to renal failure.
Supported by the International Epidemiology Institute, an independent biomedical research organization, which in turn received funding from McNeil Consumer Products. McNeil was not involved in any aspect of the study design, data collection, or interpretation.
We are indebted to the physicians caring for patients with renal diseases in Sweden for their participation; to the staff of the hospitals and laboratories that were involved in this study and to the Swedish Society of Nephrology for their participation and support; to Mrs. Annika Åkerström for coordinating the study; to Dr. Margaret McCredie and Dr. John H. Stewart (University of Otago, New Zealand) for their valuable comments on this report; and to Professor Hans-Olov Adami (Karolinska Institute, Sweden) for his advice and guidance.
Source Information
From the Department of Medical Epidemiology, Karolinska Institute, Stockholm, Sweden (C.M.F., E.E., P.L., P.W.D., C.-G.E., O.N.); the Department of Renal Medicine, Huddinge University Hospital, Huddinge, Sweden (C.M.F., C.-G.E.); the International Epidemiology Institute, Rockville, Md., and the Department of Medicine, Vanderbilt University Medical Center, Nashville (J.P.F., L.B.S., L.L., W.J.B., J.K.M.); and the National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Atlanta (M.M.Z.).
Address reprint requests to Dr. Fored at the Karolinska Institute, Department of Medical Epidemiology, Box 281, SE-171 77 Stockholm, Sweden, or at michael.fored{at}mep.ki.se.
References
| |||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Acetaminophen, Aspirin, and Renal Failure
Campo A., Thurlow W., Fored C. M., Nyrén O.
Extract |
Full Text |
PDF
N Engl J Med 2002;
346:1588-1589, May 16, 2002.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
