Get access

A longitudinal study of medication exposure and xerostomia among older people


W. Murray Thomson, Department of Oral Sciences, The University of Otago, PO Box 647, Dunedin 9001, New Zealand.
Tel.: +64 3 479 7116
Fax: +64 3 479 7113


Objective:  To describe the incidence of xerostomia among a population of older people over a 6-year period, with particular attention to medications as risk factors.

Background:  Understanding the natural history of xerostomia requires longitudinal epidemiological research, but only one study has examined changes in xerostomia over time. While medication is a recognised risk factor for dry mouth, the role of particular medication categories continues to be controversial.

Materials and methods:  Older South Australians (aged 60+) underwent an interview and dental examination at baseline, and these assessments were repeated 2, 5 and 11 years afterward. Medication data were collected at baseline, 5 and 11 years. Xerostomia data were collected at 5 and 11 years using the Xerostomia Inventory (XI) and a standard question.

Results:  Of the 1205 dentate participants assessed at baseline, 669 remained after 5 years, and 246 were assessed at 11 years. Medication prevalence increased over the observation period, such that 94.8% of the cohort were taking at least one medication by 11 years. The prevalence of xerostomia increased from 21.4% to 24.8% between 5 and 11 years (p > 0.05), and the mean XI score increased from 20.0 (SD, 6.7) to 21.5 (SD, 7.9; p < 0.001). Some 14.7% of participants were incident cases of xerostomia, while 11.4% were remitted cases; 10.1% were cases at both 5 and 11 years. After controlling for gender and ‘baseline’ xerostomia severity (represented by the XI score at 5 years), participants who commenced taking daily aspirin after 5 years had over four times the odds of becoming incident cases, while those who commenced taking a diuretic after 5 years had nearly six times the odds of doing so.

Conclusions:  While the overall prevalence of xerostomia increased during the observation period, there was considerable instability, with one-quarter of the cohort changing their status. Medication exposure was strongly associated with the incidence of the condition, with recent exposure to diuretics or daily aspirin strongly predicting it.