Does age increase the risk of adverse drug reactions?

Abstract

In the 6 years from January 1st 1995 to December 31st 2000, 92 043 spontaneous domestic reports of adverse drug reactions (ADRs) were recorded in the French pharmacovigilance database. Related to the size of the French population (58.5 millions inhabitants, 1999 National Census) this figure corresponds to an average reporting rate of 2.62 per 10 000 inhabitants per year (2.46 for men, 2.79 for women). The use of age-specific reporting rates showed (Figure 1) a marked apparent increase in risk with age: the average ADR reporting rate was thus 1.94 per 10 000 inhabitants and per year before 60 years and 5.13 beyond this age (P < 10−6). This apparent increase of the risk of ADR with age has been already found by several studies [1–3]. Figure 1 Age-specific reporting rates (France 1995–2000) before (crude, ♦) and after (adjusted, ▪) adjusting for drug consumption. As a reference for drug consumption (prescribed or not, reimbursed or not), we used data of from C.R.E.D.E.S (Centre de Recherche d’Etude de Documentation en Economie de la Sante, Paris) [4]. The C.R.E.D.E.S statistics provided the average number of drug packages bought per person in France in 1998 according to age (in 10 year bands), which allowed computation of age-specific reporting rates adjusted for drug consumption. After adjustment, the relationship with age disappeared (Figure 1). The distribution of reporting rates, previously ranging from 1.05 to 7.48 per 10,000, varied from 2.38 to 4.42 per 10,000, with a peak in the 30–39 year old class. This confirms the observation previously made by US and European authors: the greater incidence of ADRs in elderly patients observed in unadjusted studies conducted on the basis of spontaneous reports could be in part explained by exposure to a greater number of drugs [5–9]. It should be emphasized that our study was based on averaged risks of adverse drug reactions and did not discriminate either for seriousness or ADR types (e.g. certain types of ADRs, such as gastro-intestinal toxicity, could have a greater probability of occurring in elderly patients while others could be more frequent in younger patients). Moreover, our results are based on spontaneously reported data and not on the (unknown) actual num-ber of ADR cases in the source-population. Under-reporting, leading to a marked underestimation of incidence rates, might be thought to be not the same for middle-aged or elderly patients, and if greater in the latter might explain, in part, the absence of difference after adjusting for drug consumption. However, the existence of an age-specific under-reporting is not supported neither by our experience nor the literature. This study supports the idea that the main factor for the risk of adverse drug reaction is the number of drug treatments and not the age itself.