A total of 502 participants between the ages of 14 and 24 completed the survey, 60% were female and the average age was M = 19.96 years (SD = 2.63; Table 1). Of the respondents, 26.1% were high school students, 17.9% apprentices, 47.4% university students, 6% employees, 0.8% were unemployed, and 0.8% indicated that their current occupation was not listed among the options. Missing data were common and especially prevalent in the HSI questionnaire (n = 389) and CAST questionnaire (n = 399).

Among 14- to 24-year-olds, lifetime prevalence of cigarette use was 59%, while past-month prevalence was 34% (Table 2). Overall, 9% reported smoking cigarettes on a daily basis. With 50% and 52%, respectively, lifetime prevalences for e‑cigarettes and shisha were slightly lower than for cigarettes. Most participants (92%) had consumed alcohol at least once in their life. With 74%, the monthly prevalence was high as well. In total, 28% of respondents met screening criteria for problematic alcohol use. Lifetime prevalence for cannabis was 55%, while monthly prevalence was 25%. Overall, 27% of the participants who completed the CAST met criteria for problematic cannabis use. When comparing a subsample of youth aged 14–18 with data from the 2019 ESPAD study [13], we found that all prevalences in our sample were higher than in the ESPAD study: alcohol lifetime: 85% vs. 89%; alcohol past month: 60% vs. 71%; cigarettes lifetime: 48% vs. 57%; cigarettes past month: 25% vs. 37%; cannabis lifetime: 20% vs. 43%; and cannabis past month: 11% vs. 18%. Lifetime prevalence of many illicit drugs was higher as well. Compared with recent epidemiologic studies, three times as many participants reported ever having taken cocaine or amphetamines [14].

In a cross-sectional study among Austrian apprentices conducted in spring 2021 [20], 12.8% of participants scored above the cutoff for problematic alcohol use in the CAGE questionnaire. With 27.5% in the current sample, problematic alcohol use was more than twice as common (Table 2).

Participants who scored below the cutoff are represented in light gray; those who met cutoff criteria for problematic substance use are represented in dark gray. CAGE Questionnaire for Alcohol Misuse, CAST Cannabis Abuse Screening Tool, HSI Heaviness of Smoking Index

Participants could select various reasons for drinking alcohol. The most prevalent reasons were that it makes social gatherings more entertaining (47%), that it is fun (43%), and that it helps with having fun at parties (37%). Over one-third enjoy the feeling of being drunk (37%). Other reasons included coping, such as cheering oneself up when one is in a bad mood (18%), forgetting problems (12%), or helping with feelings of depression or nervousness (12%). Almost 10% stated that social pressure, fear of being left out (7%), and to be more likable (3%) led them to consume alcoholic beverages.

In line with previous research, the use of alcohol and illicit drugs largely declined during lockdowns. However, a substantial number of participants (n = 75, 44%) reported smoking more cigarettes, compared with 35% (n = 61) who reported smoking fewer cigarettes, and 21% (n = 36) who did not report a change in their frequency of smoking. Alcohol consumption decreased in 56% (n = 213) of participants, while 17% (n = 65) reported no change, and 27% (n = 104) reported drinking more alcohol during lockdowns. Among participants who used illicit drugs, almost half (47%; n = 73) reported using less, 17% (n = 26) reported having used the same amount, and 37% (n = 56) increased their drug use during lockdowns (see Fig. 2).

Change in substance use during lockdowns, reported retrospectively

Most participants stated not having been heavily drunk in the past month (86%). Half of the participants reported not having drunk more than five alcoholic beverages on one occasion, 16% reported having done so once in the past month, 14% twice, 13% 3–5 times, 3% 6–9 times, and 2% 10 or more times.

To assess the effect of time (quarter-yearly increments from spring 2020 to winter 2022/2023) on the frequency of nicotine, alcohol, and illicit drug consumption, repeated-measures ANOVAs with Greenhouse–Geisser corrections were performed. There were statistically significant effects of time on the frequency of alcohol consumption between at least two time points on the use of nicotine, F(2.99,583.59) = 20.03, p < 0.001, alcohol, F(4.1,1309.2) = 40.54, p < 0.001, and illicit drugs, F(3.72,566.11) = 15.18, p < 0.001. Post hoc tests revealed that only the use of alcohol changed significantly at two consecutive points—from spring 2020 to summer 2020 (p < 0.001, d = 0.15) and from spring 2021 to summer 2021 (p < 0.001, d = 0.20)—while there was no significant change of nicotine and illicit drug use when assessing consecutive time points.

Between-subjects Greenhouse–Geisser corrected repeated-measure ANOVAs were performed to assess whether adolescents and young adults with problematic substance use (i.e., scoring above the cutoff for any of the employed screening questionnaires) increased their substance use during the pandemic, compared with youth not reporting problematic substance use. There was a statistically significant difference in frequency of use between at least two time points between the groups for alcohol, F(4.09, 1300.28) = 2.46, p = 0.04, but not for nicotine, F(3.02, 585.98) = 2.47, p = 0.06, and illicit drugs, F(3.77, 568.87) = 2.06, p = 0.09.

Participants who scored above the cutoff on screening measures for problematic substance use reported a higher frequency of substance use at all time points, compared with participants scoring below the cutoff (Figs. 3, 4).

Trends in the frequency of substance use during the pandemic, by age group, for three different substance classes. SP spring, SU summer, FA fall, WI winter

Trends in the frequency of substance use during the pandemic, grouped by score on screening measures, for three different substance classes. SP spring, SU summer, FA fall, WI winter

Participants who reported smoking more cigarettes or drinking more alcohol during lockdowns were not more likely to currently show problematic nicotine use (odds ratio [OR] 1.74, 95% confidence interval [CI]: 0.71–4.25; OR 0.95, 95% CI: 0.37–2.45).

Those who drank more alcohol during lockdowns were more likely to currently exhibit problematic alcohol (OR 3.03, 95% CI: 1.90–4.85) and cannabis (OR 2.60, 95% CI: 1.56–4.33) use.

Individuals who reported using more of one specific substance during lockdowns were more likely to have increased their use of other substances as well (alcohol and cigarettes OR = 2.66, 95% CI: 1.37–5.15; alcohol and illicit drugs OR = 4.87, 95% CI: 2.36–10.04; cigarettes and illicit drugs OR = 3.48, 95% CI: 1.56–7.78).