The review by Pape [1] appropriately elaborates the persistent problem of judging the reliability of self- or peer-reported drug and alcohol use among youth. Characterizing the magnitude of a ‘misperception’ in the reports of teens about their peers' substance use, however, presumes confidence in knowing ‘actual’ rates and patterns of use. As Pape recounts, significant effort has been directed at attempts to account for and minimizing bias, denial, confabulation, exaggeration and other factors and influences on self-report by comparing various methods of collecting information that approximate ‘actual’ use among youth, and the perceptions about that use.

The approach we reported recently was to obtain objective biological markers of drug use against which self-report and caregiver report of teen use could be compared [2]. The results from our sample of 211 teens whose hair specimens were assayed for cocaine and opiate metabolites were that teens' hair was 52 times more likely to be positive for cocaine (33.7%) than was indicated by their self-reported frequency/recency of cocaine use. No teen reported opiate use, while 6.6% of samples were positive. In this study, parents also significantly under-reported their own cocaine and opiate use.

This approach was effective in our cohort of urban African American 14-year-olds for cocaine and opiates because of the availability of reliable biomarkers in hair for metabolites of those drugs. Cocaine metabolites, in particular, are identifiable for months after use. In contrast, biomarkers for marijuana smoking may be less reliable and typical markers of alcohol drinking may be non-specific or tend to be short-lived, although detection of fatty acid ethyl esters in hair is now being recommended to address these limitations [3,4].

The point is that if our data with long-term biological markers in hair are a more reliable indicator of actual use, at least for cocaine and opiates, then indeed the best self-reports are greatly underestimating the prevalence of substance use by youth. Further, the reports of peers about other teens may be more faithful indicators of the extent of cocaine use than self-report. While youth are clearly under-reporting their own use of cocaine, alcohol and other drugs, and while youth reports of drug use by their peers is greater than any self-report, our results suggest that youth may not be exaggerating differences between self-report and other report to the degree Pape concluded. Rather, teens and youth may be more faithful reporters of their peers' drug use than their peers are about themselves.

Declarations of interest