The headline on the federal news release was reassuring: cigarette smoking at historically low levels, alcohol use continuing a long decline, substance use among American adults otherwise steady.
Further down was a number that went the other way. Self-reported heroin use among adults aged 19 to 30 rose from 0.2 percent in 2024 to 0.6 percent in 2025, which investigators described as the first significant uptick and the highest level in about a decade.
The findings come from the 2025 wave of the Monitoring the Future Longitudinal Panel Study, conducted by researchers at the University of Michigan with support from the National Institute on Drug Abuse, and released this week.
What This Survey Actually Measures
Monitoring the Future is not a single survey. The part most people have heard of is the annual cross-sectional survey of 12th graders, which has run since 1975.
The panel study follows those same people afterward. To estimate drug use prevalence among a nationally representative sample of US high school graduates over the past 50 years, the panel has collected data from more than 120,000 individuals who originally participated as 12th graders. It now covers adults aged 19 to 65.
Young adult participants are surveyed every two years. Midlife adults are surveyed every five.
That design produces something most drug surveys cannot: the ability to watch the same cohort age through decades of changing drug availability. It also carries a structural limitation worth stating up front, since the panel is drawn from high school graduates and therefore does not represent people who left school before graduating, a group with a different substance use risk.
The Reassuring Half
Cigarettes have collapsed as a young adult behavior. Just 8.3 percent of adults aged 19 to 30 reported smoking in the past 30 days. Across all age groups measured, past-30-day smoking ranged from that figure to 10.6 percent among adults aged 55 to 65.
Alcohol remains the most commonly used substance among adults, but among young adults, nearly all measures of drinking reached low levels, with past-year, past-30-day, and high-intensity drinking holding steady rather than rising.
Nora Volkow, director of the National Institute on Drug Abuse, framed the split directly, saying, "It is reassuring to continue to see relatively low smoking rates" alongside the continuing decline in adult alcohol use.
For readers who assume American drug use is uniformly worsening, these are the numbers that complicate the picture. Two of the substances responsible for the largest share of preventable deaths in the United States are being used less, not more.
The Numbers That Went the Other Way
Cannabis and nicotine did not follow the same path.
Daily, past-30-day, and past-year cannabis use and cannabis vaping remained at or near all-time highs among adults overall. Adults aged 35 to 50 reported a significant increase in daily cannabis use, from 6.6 percent in 2024 to 8.5 percent in 2025. Nicotine vaping also remained at or near record levels: among young adults, past-30-day nicotine vaping has tripled and past-year use has nearly doubled since 2017. Reported nicotine pouch use over the past year among all adults has more than doubled since 2023, when the panel first asked about these products.
Those increases coincided with the low cigarette figures, a pattern that has defined the past several years. The delivery method changed, not the underlying appetite.
Then there is heroin. The increase among 19- to 30-year-olds is small in absolute terms, from 0.2 percent to 0.6 percent, and it remains below the all-time high of 0.7 percent recorded in 2013. It is the direction and the break in trend that make it notable rather than the magnitude. Nonmedical use of sleeping medications in the same age group also rose, from 2.0 percent to 3.0 percent, while past-year psychedelics use fell from 8.9 percent to 7.2 percent, reversing a decade-long increase.
Two cautions apply. The base rate for heroin use in this population is very low, and small changes in a very small number carry wide statistical uncertainty. And self-report is the least reliable measurement method for the most stigmatized substances, which cuts in both directions across years.
Volkow's reading was that rising rates of cannabis, nicotine, and certain other drug use warrant attention, and that continued monitoring is what allows effective interventions to be developed.
What Self-Report Cannot See?
Every figure here comes from people describing their own behavior on a survey, which shapes what the data can support.
Underreporting is concentrated exactly where the public health stakes are highest. Someone using heroin is less likely to say so than someone who had two beers, and someone in active crisis is less likely to complete a panel survey at all.
Long-running panels also lose participants. Attrition over decades is not random, and people whose lives become chaotic are more likely to drop out, which tends to make later waves look healthier than the original cohort.
The survey measures use, not consequence, and that gap is the largest one here. It does not count overdose deaths, emergency department visits, or substance use disorder diagnoses, and a population can use a drug less often while dying from it more often if the supply becomes more dangerous. That has been the defining feature of the American overdose crisis, and it is the reason a decline in reported use should never be read as a decline in harm.
Federal researchers, who maintain a public set of trend resources, have said continued monitoring is essential precisely because the drug landscape keeps shifting underneath the numbers. The full panel results are published by the University of Michigan. Anyone concerned about their own substance use or a family member's can reach the free, confidential SAMHSA National Helpline at 1-800-662-HELP for treatment referrals and information.
Key Questions Answered
What is the Monitoring the Future Longitudinal Panel Study? It is a long-running survey conducted by the University of Michigan with NIH support that follows people who first participated as 12th graders. It has collected data from more than 120,000 individuals and now covers adults aged 19 to 65, surveying young adults every two years and midlife adults every five.
What did the 2025 wave find about cigarettes and alcohol? Past-30-day cigarette smoking was 8.3 percent among adults aged 19 to 30 and ranged up to 10.6 percent among adults aged 55 to 65. Among young adults, drinking measures held steady at low levels, continuing a multi-year decline in adult alcohol use.
What went up? Cannabis use and cannabis vaping remained at or near all-time highs, with daily use among adults aged 35 to 50 rising from 6.6 percent to 8.5 percent. Nicotine vaping stayed at or near record levels, and nicotine pouch use more than doubled since 2023.
How big was the heroin increase? Past-year heroin use among adults aged 19 to 30 rose from 0.2 percent in 2024 to 0.6 percent in 2025. That is the first significant uptick in about a decade, though it remains below the 0.7 percent recorded in 2013.
Should that heroin figure alarm people? It warrants attention rather than alarm. The base rate is very low, so small changes carry wide statistical uncertainty, and heroin is among the substances people are least likely to report accurately on a survey.
What are the limits of this data? The panel is drawn from high school graduates, so it does not represent people who left school before graduating. It relies on self-report, loses participants over decades in ways that are not random, and measures use rather than consequences such as overdose deaths or emergency visits.
Where can someone get help with substance use? The SAMHSA National Helpline at 1-800-662-HELP is free and confidential and provides treatment referrals and information.
