Effect of Antipyretic Therapy on the Duration of Illness in Experimental Influenza A, Shigella sonnei , and Rickettsia rickettsii Infections

Karen I. Plaisance, Pharm.D., Suneel Kudaravalli, Pharm.D., Steven S. Wasserman, Ph.D.,Myron M. Levine, M.D., and Philip A. Mackowiak, M.D.

Pharmacotherapy. 2000;20(12) 

In This Article


The challenge data analyzed in our investigation offer potentially important insight into possible effects of antipyretic therapy on the course of clinical infections. In the investigation, a striking correlation existed between antipyretic therapy and duration of illness in subjects infected experimentally with influenza A and S. sonnei. No such relationship existed in volunteers infected with R. rickettsii. Influenza A-infected subjects who received antipyretic agents during their illness were sick, on average, 3.5 days longer than those not receiving antipyretic agents (8.8 ± 2.3 days vs 5.3 ± 3.0 day, p<0.001). S. sonnei-infected subjects also exhibited a striking prolongation of illness in association with antipyretic therapy (4.6 ± 2.1 days for those receiving antipyretics vs 1.9 ± 1.6 days for those not receiving antipyretics, p<0.001). Moreover, in both cases, the duration of illness increased pari passu with increasing number of doses of antipyretic agents.

The critical question of whether antipyretic therapy prolonged the illness cannot be answered definitively based on this retrospective obser-vational study. Subjects treated with antipyretic agents might have been ill longer than those not receiving antipyretics simply because the former subjects had more severe illnesses than the latter. The univariate analysis seemed to support this conclusion in that both influenza A- and S. sonnei-infected subjects who were treated with antipyretic agents had significantly higher maximum temperatures during their illness and a greater number of signs and symptoms recorded at the time of their maximum temperature than those not receiving antipyretics. However, when the relationship between relevant clinical variables (including maximum temperature and maximum number of signs and symptoms) and duration of illness was examined by multivariate analysis, only antipyretic therapy exhibited a statistically significant association with duration of illness. Thus, the longer duration of illnesses in subjects treated with antipyretic agents could not be explained solely on antipyretics being given preferentially to influenza A- and S. sonnei-infected subjects with the most severe illnesses.

It is also possible that antipyretic therapy was the result of prolonged illness rather than its cause. If so, one would expect both that subjects receiving antipyretic agents would have longer illnesses than those not receiving antipyretics and that such therapy would be started relatively late in the course of the illness. Conversely, if antipyretic agents prolonged the illness, one again would expect those receiving antipyretic agents to have longer illnesses than those not receiving such agents. In this case, one could expect that antipyretic therapy would be instituted relatively early during the illness because the capacity of an intervention to alter the course of an illness would be greater the earlier it was started during the illness.

The first scenario applied to S. sonnei-infected subjects, over half of whom received their first dose of antipyretic drug after the midway point of their illness. Indeed only 7 of 21 S. sonnei-infected subjects received their initial dose of antipyretic within the time frame corresponding to the mean duration of illness among subjects not receiving antipyretic agents. These findings suggest that antipyretic therapy was the result of prolonged illness in S. sonnei-infected subjects, not the cause.

By comparison, influenza A-infected subjects, for the most part, received their first dose of antipyretic before the midway point of their illness. More important, almost all received their first dose within the time corresponding to the course of illness in subjects not receiving antipyretics. These findings suggest that, at the very least, antipyretic therapy had an opportunity to prolong the illness of influenza A-infected subjects because they received their initial dose before they were expected to have recovered from their illness had they not received such therapy.

Little is known of how antipyretic therapy prolongs the signs and symptoms of viral infections or why viral infections might differ from other infections in this regard. Aspirin inhibited antibody formation and secondary antibody responses,[10,11] the interferon-mediated antiviral state in mouse L cells,[12] incorporation of thymidine into DNA by mitogen- and antigen-stimulated lymphocytes,[13] and both granulocyte adherence to foreign particles[14] and granulocyte phagocytosis.[15] Many of these effects are mediated by inhibition of nuclear factor-kB(NF-kB).[16] On the other hand, aspirin also enhances thymidine uptake in mitogen-stimulated lymphocytes,[17] production of interleukin-2 by mitogen-stimulated peripheral blood lymphocytes,[18] and interferon production by stimulated lymphocytes.[19] By comparison, acetaminophen has relatively modest enhancing[20] and inhibitory effects on immune function.[21]

Antipyretic therapy might prolong some infections because it alleviates fever and, thereby, abrogates the reported potentiating effect of fever on resistance to infection.[2] In our study, S. sonnei-infected subjects exhibited significant reductions in core temperature in association with antipyretic therapy. Influenza A-infected subjects did not, however, in spite of the fact that antipyretic therapy appeared to prolong their illness.

Whatever the mechanisms responsible, antipyretic drugs prolonged the course of both rhinovirus[5] and varicella virus infections.[4] Our study suggests that such agents might have a similar effect on influenza A infections. In view of the potential clinical importance of these preliminary observations, we believe they merit further investigation in a prospective, randomized, placebo-controlled trial.

Manuscript received April 25, 2000. Accepted pending revisions June 30, 2000. Accepted for publication in final form September 15, 2000.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: