Danel Safer and Julie Magno Zito question the gender and age ratios observed in our study and attribute differences between our study on stimulants as a group of drugs and studies elsewhere on methylphenidate (in particular) to the inclusion of diet pills in our questionnaire. The inclusion of diet pills along with other prescribed stimulants was noted in our discussion section as a limitation of the present study. However, as Safer and Zito comment in their letter, if stimulant diet pills are rarely prescribed to minors, then one would not expect the inclusion of diet pills to greatly influence male–female ratios of prescribed stimulants. In contrast, nonprescription diet pills are preferentially used by females.1 Johnston and colleagues examined nonprescribed diet pills and stay-awake pills (caffeine, ephedrine and/or phenylpropano-lamine) among American adolescent students. They concluded there was little distortion in their estimates of stimulants due to the marked decline in the US of the annual prevalence of the use of diet pills (from 20% in 1982 to 9.6% in 1998), in the presence of an increasing trend in amphetamine use.1 Also, regarding male–female methylpheni-date ratios, the gap between males and females has been narrowing.2 Safer and Krager showed a narrowing of the ratio from 1:12 in 1981 to 1:6 in 1993 in middle school.3 Robison and colleagues reported a narrowing of the male–female ratio for children aged 5 to 18 years, from 5.4:1 in 1990 to 3.1:1 in 1995.4
Safer and Zito state that the prevalence of stimulant treatment in our study was 50% higher in 10th grade than in 7th grade. The estimate of past-month medical stimulant use, which is more likely to be accurate, shows no significant difference in the prevalence of medical stimulant use in 7th compared with 10th grade (p > .05). Of note, Zito and colleagues found that the largest increase in methylphenidate utilization had occurred among high-school aged youth of 15 to 19 years.2 Our item on past-year medical use was analyzed primarily to determine the relationship between medical and nonmedical stimulant use. The medical and nonmedical drug use items, symmetrical by design, date back to 1991 in the Nova Scotia Student Drug Use Survey and earlier in the case of the Ontario Student Drug Use Survey. Due to the 12-month recall period and discontinued therapeutic regimens and trials of therapy, the past-year prevalence estimate can be expected to be less accurate than the past-month estimate. However, this should not invalidate the association between past-year medical and non-medical stimulant use. In effect, our study revealed a relationship between medical and non-medical stimulant use based on several indicators, including the past-year use items.
Finally, marked geographic variability has been observed in methylphenidate utilization.2,5,6 Whereas the Nova Scotia Prescription Monitoring Program provides some insight into methylphenidate utilization in Nova Scotia, we do not have comparable information for the other 3 Atlantic provinces. We do know significant differences exist in prevalence of use of several substances among adolescent students in the 4 provinces.7,8 Clearly, many factors could have influenced the age and gender ratios observed in our study.
Christiane Poulin Department of Community Health/ Epidemiology Clinical Research Centre Halifax, NS