Did the under-reporting of meth/amphetamine use increase in a general population survey in Australia as negative media coverage increased?

Abstract

Aim: To test (1) if there was a change in self-reported lifetime prevalence of meth/amphetamine use by birth cohort and (2) if the extent of under-reporting of meth/amphetamine use was associated with the proportion of the population who nominated meth/amphetamine as a drug problem.

Design: Observational study using seven waves of repeated cross-sectional nationally representative household surveys between 2001 and 2019.

Setting: Australia.

Participants: Participants were from three birth cohorts: 1951-60 (age 68-77 at the 2019 survey; n = 29 458; 55% female), 1961-1970 (age 58-67; n = 29 859; 57% female) and 1971-1980 (age 48-57; n = 28 758; 59% female). Data were weighted to align the sample to the Australian population.

Measurements: Past year meth/amphetamine use; under-reporting of lifetime meth/amphetamine use in each birth cohort, year and survey stratum (operationalised as the difference between self-reported lifetime prevalence in 2001 and that of each subsequent year); proportion of the population who nominated meth/amphetamine as a drug problem in each birth cohort, year and survey stratum. Under-reporting was regressed on the proportion of people holding negative attitude towards meth/amphetamine. Survey year and birth cohort were adjusted for.

Findings: Between 2001 and 2019, the lifetime prevalence of meth/amphetamine decreased from 6.1% (95% CI = 5.3-6.9) to 1.7% (95% CI = 1.2-2.2) in the 1951-1960 birth cohort (p < 0.001), from 13.0% (95% CI = 12.0-14.1) to 4.4% (95% CI = 3.7-5.2) in the 1961-1970 birth cohort (p < 0.001) and from 21.4% (95% CI = 19.9-22.9) to 11.2% (95% CI = 10.0-12.4) in the 1971-1980 birth cohort (p < 0.001). The proportion who nominated meth/amphetamine as a 'drug problem' increased significantly in all three cohorts (all p < 0.001) and the degree of under-reporting of meth/amphetamine use was significantly associated with proportion of people who nominated meth/amphetamine as the 'drug problem' (b = 0.09, SE = 0.01, p < 0.001).

Conclusion: In Australia, the actual prevalence of lifetime meth/amphetamine use may be two- to fourfold higher than that estimated in the most recent national household surveys (2019). The level of under-reporting is strongly associated with increasing negative attitudes towards methylamphetamine and d-amphetamine use over the same period.

Keywords: amphetamine; bias; meth; methamphetamine; population survey; under-reporting.