Ask people who use cannabis whether it makes them drink less, and most will say yes. That is what a large New Zealand survey reported, and it is the finding this article was first written around. A harder question is what people’s habits look like when nobody asks them to explain anything: how many days a month they drink, how many days they use cannabis, and which of the two dominates. A fresh study from the United States measures exactly that, and it does so only among women.
The paper, Substance use patterns in a community sample of adult women in the United States: Alcohol, Cannabis, and other drug use, was published online on 18 September 2026 in Addictive Behaviors Reports by Shelly F. Greenfield, Allen J. Bailey and colleagues at McLean Hospital and Harvard Medical School. It is open access and covers 2,296 women aged 18 to 70 who had used at least one substance in the previous 30 days. This update sets it beside the New Zealand survey of 23,500 people published in Harm Reduction Journal in November 2024.
Key Takeaways
In the United States study, 84.8% of the women had drunk alcohol in the past month and 45% had used cannabis, while 18.3% reported heavy alcohol use and 22.6% heavy cannabis use, with “heavy” meaning 16 or more days out of 30. A statistical sorting of the answers produced four groups: women who mainly drink (about 57% of the sample), women who use cannabis heavily and drink to varying degrees (27%), women who use several substances including other illicit drugs (10%), and women who mainly misuse prescription medication (6%). Age pulled the two main substances in opposite directions: heavy drinking was more common in each older age group, heavy cannabis use in each younger one. The sample came from a paid online panel and was deliberately weighted toward frequent users, so these percentages describe the women surveyed, not American women as a whole. The earlier New Zealand survey found that 60% of people who used both cannabis and alcohol said cannabis led them to drink less, but it too relied on a self-selected online sample and on people’s own judgment of cause and effect. Neither study can show that cannabis makes anyone drink less, and the American one did not ask.
Cannabis and Alcohol Use Among Women in the United States: Who Took Part in the Study
The researchers recruited through online panels run by the survey company Sago Health. Between July and September 2024, women aged 18 to 70 were invited to answer four screening questions: on how many of the past 30 days had they used alcohol, cannabis, other drugs, or prescription medication in a way it was not prescribed. Each answer fell into one of four bands: none, 1 to 5 days (which the authors call mild use), 6 to 15 days (moderate) and 16 or more days (heavy). Anyone reporting at least some use could take the full questionnaire, for which they received panel points worth 2 dollars.
Of 4,876 women screened, 2,405 (49.3%) were ruled out, 174 chose not to continue and one response was removed as spam, leaving 2,296. Their average age was 47.1. The team split them into four age groups of equal span: 18 to 30 (319 women), 31 to 43 (642), 44 to 56 (655) and 57 to 70 (680). In the sample, 72.3% identified as white, 18.7% as Black or African American and 11.5% as Hispanic or Latino; 41.4% were married, 63.1% had children and 48.9% worked full time. Money was tight for many: 28.9% said they had gone without medical care they needed in the past year because of cost, and 38.0% said they lacked enough income for basic living expenses at least some of the time in the past month.
One design detail matters for every number that follows. To make sure frequent users were well represented, the team changed the entry rule partway through. After the first 1,495 respondents, only women reporting at least 6 days of use in one of the four categories were admitted; the original rule returned for the last 337 women screened. Of those ruled out, 1,385 reported no substance use at all and 1,020 were turned away during that middle phase for mild use only. The result is a sample with more frequent users than a random draw of American women would produce.
Why study women separately? The authors point out that over the past three decades the gap between men and women in substance use disorders in the United States has narrowed sharply, that women tend to suffer more severe medical and social consequences, and that most women with a substance problem never seek treatment. National surveys show the same shift from another angle. In the long-running Monitoring the Future study, women aged 19 to 30 reported more past-year cannabis use than men of the same age for the first time in 2023, although men aged 35 to 50 still used more than women. And in the CDC telephone survey we covered in our piece on cannabis use trends among U.S. adults, past-month use among women rose from 6.1% to 14.4% between 2016 and 2024, faster than among men.
Four Patterns: Mostly Alcohol, Mostly Cannabis, Several Substances, Prescription Medication
Across the whole sample, alcohol was the most common substance: 84.8% of the women had drunk on at least one day in the past month and 18.3% on 16 days or more. For cannabis the figures were 45% and 22.6%. Prescription medication misuse was reported by 19.1%, with 8.6% misusing on 16 or more days, and 11.2% had used other substances, with 6.1% reporting heavy use.
Averages like these hide how people combine substances, so the team used a method called latent class analysis. Put simply, the software sorts people into groups whose answers resemble one another, without the researchers deciding in advance what the groups should be. Four groups emerged.
The largest, about 57% of the sample, the authors call “Primary Alcohol.” These women drank anywhere from a few days a month to 16 or more, had less than a 20% likelihood of reporting any cannabis use, and did not use other illicit drugs. The second group, 27%, is “Heavy Cannabis with Alcohol”: a high likelihood of using cannabis on 16 or more days a month, with drinking at varying frequencies. The third, 10%, is “Polysubstance Use,” marked by the highest probability of frequent use of other illicit drugs alongside a wide range of alcohol and cannabis use; the authors note that this group probably lumps together several different combinations. The fourth, 6%, is “Prescription Medication Misuse with Alcohol”: moderate to frequent misuse of prescription drugs with some drinking, although this group also had the highest likelihood of not drinking at all.
The authors conclude that more than half of the women used more than one substance. That fits earlier work by two members of the same team arguing that using several substances is the rule among people who use any, not the exception.
For readers interested in the cannabis-or-alcohol question, the two biggest groups are the telling ones. Most women in the alcohol group used little or no cannabis, while women in the heavy cannabis group also drank, at varying frequencies. That is a snapshot of who does what. It is not evidence that one substance pushed out the other, because the survey never asked the women why they used what they used or whether cannabis had changed their drinking.
Age Pulls Alcohol and Cannabis in Opposite Directions
The clearest result in the paper is the age pattern. The table shows what share of each age group fell into each of the four groups.
| Pattern | All women | 18–30 | 31–43 | 44–56 | 57–70 |
|---|---|---|---|---|---|
| Primary alcohol | 57% | 52% | 48% | 56% | 68% |
| Heavy cannabis with alcohol | 27% | 34% | 31% | 28% | 20% |
| Polysubstance use | 10% | 11% | 16% | 9% | 5% |
| Prescription misuse with alcohol | 6% | 3% | 5% | 7% | 7% |
Heavy drinking, counted as 16 or more days a month, was lowest among the youngest women and rose with each older group. Heavy cannabis use did the reverse. Compared with women aged 18 to 30, women aged 57 to 70 had less than half the odds of belonging to the heavy cannabis group (odds ratio 0.44, 95% confidence interval 0.32 to 0.60), while the differences among the three younger groups were not statistically significant.
Other answers point the same way. Among women who had drunk alcohol in the past three months, 30.8% of those aged 18 to 30 drank on three or more days a week, against 43.9% of those aged 57 to 70. Monthly cannabis use ran from 56.1% in the youngest group through 52.9% and 42.9% to 34.7% in the oldest. The authors themselves flag that last number: more than a third of women aged 57 to 70 in this sample used cannabis at least once a month.
There is a twist in the frequency data. Among the 1,167 women who answered a question on how often they use cannabis, 37.6% said once a day or more. That share was 32.1% among women aged 18 to 30 and 41.2% among women aged 44 to 56. In other words, fewer older women in the sample used cannabis, but those who did were not occasional users.
On alcohol, the findings match what national data have shown for years. The authors cite an analysis of the National Survey on Drug Use and Health in which binge drinking among women aged 50 to 64 rose from 6.3% to 9.1% between 2005 and 2014, a relative increase of 44.4%, while the change among men of the same age, from 19.6% to 21.5%, was not statistically significant. A separate study of national health interviews from 1997 to 2014 found binge drinking rising among people aged 60 and over, especially women. The paper lists the health risks of heavy drinking for older women as cognitive impairment, falls, liver disease and breast cancer, citing the U.S. Surgeon General’s advisory on alcohol and cancer risk.
Two smaller findings round out the picture. Cigarettes were far from gone: 25.8% of the women had smoked in the past three months, with the highest rates among those aged 31 to 43 (31.6%) and 44 to 56 (28.1%), and of the smokers 49.0% smoked ten or more cigarettes a day and 65.3% lit the first one within 30 minutes of waking. And prescription medication misuse was concentrated among older women, with 7% of each of the two older age groups falling into that pattern. The paper does not report which cannabis products the women used, although federal survey data show that men are more likely to smoke cannabis while women more often choose edibles, drinks and topicals.
A caution on reading the age pattern: the survey captured one moment. It cannot say whether today’s 25-year-old heavy cannabis users will be drinking more and using less at 60, or whether each generation simply carries its own habits through life.
The New Zealand Survey: Does Cannabis Replace Alcohol and Other Drugs?
A comprehensive study conducted in New Zealand involving 23,500 participants, published in Harm Reduction Journal in November 2024, found that cannabis consumption may act as a substitute for more harmful substances, potentially decreasing the use of alcohol, opioids, and stimulants. This research adds to the ongoing global conversation about the potential impacts of marijuana, particularly in the context of legalization and harm reduction strategies.
The data come from the New Zealand Drugs Trends Survey, an anonymous online questionnaire run by Massey University between January and May 2020. People who had used cannabis and another substance in the same six months were asked one direct question for each substance: does using cannabis mean you use more, less or the same of it?
Key Findings: Cannabis as a Substitution
The study found that 60% of participants reported using less alcohol after starting to use cannabis, while a similar percentage noted reduced use of synthetic cannabinoids. Additionally, 44% stated their morphine use decreased, and 40% said they consumed less methamphetamine. The data suggests that marijuana could play a role in reducing the frequency and quantity of these substances among certain user groups.
However, the impact of marijuana as a substitute varied depending on the substance and user demographics. For instance, nearly 70% reported that cannabis use did not affect their consumption of LSD, MDMA, or cocaine. Notably, 33% of those who used both cannabis and tobacco indicated they smoked fewer cigarettes, although 20% said cannabis use led to increased tobacco consumption.
Demographic Insights
The results highlighted that young adults (aged 21 to 35) were more inclined to report reduced alcohol and methamphetamine use due to marijuana. Conversely, urban dwellers and students were less likely to experience this substitution effect. Adolescents aged 16 to 20 had mixed outcomes, with some reporting increased use of other substances, while others saw a decrease or no change at all. Participants aged 21 to 25 consistently noted lower use of alcohol and other drugs, such as MDMA and methamphetamine.
Gender played a modest part as well. In the statistical model, male respondents had slightly higher odds of saying cannabis led them to drink less (odds ratio 1.14), as did Māori respondents (1.19). Seen next to the American data, that is one more hint that men’s and women’s patterns should not be assumed to be the same.
Broader Implications
These findings contribute to a complex narrative regarding marijuana’s role in potentially mitigating the use of more harmful drugs. The study noted similar mixed outcomes in past research, such as U.S. college studies that observed reduced binge drinking among students aged 21 and over post-legalization, but no significant change among younger peers.
The report also referenced alcohol purchase trends in regions where marijuana is legal. For example, after the legalization of recreational cannabis, Colorado and Washington state saw declines in wine purchases, while Washington reported an increase in spirit purchases, and Oregon observed a decrease.
The wider literature is just as divided. A systematic review of 65 studies cited by the New Zealand team found that 30 supported the idea that cannabis replaces alcohol, 17 found that the two are used together and reinforce each other, 14 found neither, and four found both. On The Cannex we have covered research from each camp, including an Oregon study linking more cannabis shops to less heavy drinking and a laboratory experiment in which heavy drinkers drank less after using cannabis, though not all of them did.
A Complex Relationship
While the study adds valuable insights, the authors emphasize that the relationship between cannabis and other substance use remains nuanced. The potential for marijuana to serve as a less risky alternative for certain drugs is evident, but the extent varies by user profile, geographic location, and legal context.
What the Two Studies Can and Cannot Tell You
Both studies are surveys of volunteers, and both sets of authors say so plainly.
The New Zealand sample was recruited through Facebook advertising aimed at people aged 16 and over with an interest in alcohol, music and nightlife. It is a convenience sample, not a mirror of the population. The 60% figure is people’s own opinion about cause and effect, with no measurement of how much less they drank. And one detail cuts against a simple reading: everyone who used both cannabis and alcohol, including those who said cannabis made them drink less, drank more often than drinkers who did not use cannabis at all. Over six months, the “less” group averaged 42.0 drinking days, the “no impact” group 46.8, the “more” group 58.7, and drinkers who did not use cannabis 38.5. The authors suggest that people who drink but avoid cannabis may differ in ways the survey did not capture, such as age, family and work commitments.
The American study has its own limits. The authors describe it as a cross-sectional online survey of a non-probability sample drawn from one commercial panel, state that it is not meant to replicate national surveys, and warn that the results may not apply to the wider population and cannot establish cause. Beyond what they list, three features of the design are worth keeping in mind. Only women who had used something in the past month were included, so the study says nothing about women who abstain. The mid-study change in the entry rule means that the 18.3% and 22.6% heavy-use figures are not estimates for American women in general. And “heavy” is defined by the number of days, not by how many drinks or how much cannabis was consumed on those days.
Population data add a further note of caution. In the CDC survey analysis mentioned above, binge drinkers had 2.40 times the odds of past-month cannabis use compared with people who did not binge drink. Across a whole population, drinking and cannabis use tend to travel together more often than one excludes the other.
What the new paper does offer is a map of where to look. The authors name three groups they consider under-studied: older women who drink heavily, younger women who use cannabis heavily, and older women who misuse prescription medication. Their practical recommendation is addressed to clinicians, who they say should ask women about all substances, not just one, at every stage of adult life. The work was supported by philanthropic funds at McLean Hospital and a fellowship at the Harvard Radcliffe Institute, and several authors report research grants from the National Institutes of Health.
FAQ
Neither study can answer that. In the New Zealand survey, 60% of people who used both substances said cannabis led them to drink less, but that is a personal impression from a self-selected sample of men and women. The United States study of 2,296 women recorded how often they drank and used cannabis and did not ask whether one affected the other.
About 57% of the women mainly drank alcohol and used little or no cannabis. Another 27% used cannabis on 16 or more days a month and drank to varying degrees. About 10% used several substances, including illicit drugs other than cannabis, and 6% mainly misused prescription medication, sometimes with alcohol.
In this sample, heavy drinking was more common in each older age group, while heavy cannabis use was most common among women aged 18 to 30 and least common among those aged 57 to 70. Because the survey was taken at a single point in time, it cannot tell whether individual women change their habits as they age or whether the generations simply differ.
It depends on the state and on the purpose. As of June 2026, 41 states and Washington, DC allowed medical cannabis and 24 states and Washington, DC allowed adult recreational use. Under federal law, an order published in April 2026 moved FDA-approved marijuana products and marijuana covered by a state medical marijuana license to Schedule III of the Controlled Substances Act; all other marijuana, including cannabis sold in state adult-use shops, remains in Schedule I. A hearing on moving all marijuana to Schedule III was paused by a Drug Enforcement Administration judge on 29 September 2026 with no end date set. Alcohol is legal nationwide for adults aged 21 and over. Rules on possession, purchase and age limits for cannabis differ from state to state, so check the law where you live.
Legal Disclaimer
This article is journalistic coverage of published scientific research and is provided for informational purposes only. It does not constitute medical, legal or professional advice, and nothing in it should be read as encouragement to use cannabis, alcohol or any other substance, or to replace one substance with another. The studies described are observational surveys based on self-reported answers and cannot establish cause and effect. Cannabis laws differ between countries and between U.S. states, differ between state and federal level, and change frequently; readers are responsible for knowing the rules that apply where they live. Anyone concerned about their own or someone else’s substance use should speak to a qualified health professional.
