Americans, and especially those under age 35, are changing their tune on alcohol use, with a growing share endorsing the view that moderate drinking is bad for health — and a new study backs them up.

According to a Gallup poll released Tuesday, almost half of Americans, 45%, say that having one or two alcoholic drinks a day is bad for a person’s health. That’s the highest percentage yet recorded by the survey, which has been conducted 10 times since 2001.

Younger adults are the group most likely to say drinking is bad for health, with 65% in that camp, compared with 37% of adults ages 35 to 54 and 39% of adults 55 and older.

Only 8% of adults reported that they thought moderate drinking had a positive effect on health, an all-time low.

More younger adults are choosing to abstain, too, as nonalcoholic “mocktails” become widely available and people feel less social pressure to drink. A different Gallup poll published last year found that 62% of adults under 35 said they drink alcohol, down 10 percentage points from 20 years earlier.

However, the survey also found a 10 percentage point increase in adults 55 and older who reported drinking, and a new study says that could be harmful for their overall health.

Separately, a new study found that moderate drinking doesn’t appear to benefit the health of older adults, joins a growing body of research countering the once widely held belief that a glass or two of alcohol a day, especially wine, might be good for health. Instead, recent research has revealed that alcohol can be a potent carcinogen and is a potential contributor for a host of other diseases, such as depression and liver and kidney problems.

“Alcohol is a carcinogen and contributes to about 50 different types of death,” said Dr. Timothy Naimi, who directs the Canadian Institute for Substance Use Research at the University of Victoria.“Overall, alcohol is a health hazard,” said Naimi, who was not involved in the study.

Naimi said that fact tends to get lost when people start to talking about moderate drinking or how much alcohol they can have before it starts to be harmful.

When it comes to alcohol, he said, less is more.

“The most consistent finding across all the science is that, generally speaking, the less you drink, the better you are off health-wise,” Naimi said.

The study, which was published Monday in the journal JAMA Network Open, largely came to that same conclusion. It looked at the health outcomes of more than 135,000 adults 60 and older tracked through the UK Biobank registry. At the first study visit, which was between 2006 and 2010, participants were asked detailed questions about their alcohol use, and the researchers used their answers to classify them as occasional or low-, moderate- or high-risk drinkers.

Studies like these, which look back in time to find patterns and relationships between a person’s habits and their health, are subject to biases that can skew their results. The researchers tried to avoid one of the biggest ones in studies of alcohol use by avoiding comparisons with people who said they didn’t drink any alcohol.

Nondrinkers often don’t drink because they can’t, perhaps due to health conditions or medication use. Comparing drinkers with nondrinkers, in that case, could make drinking look less harmful or even beneficial, as some older studies have suggested.

For the new study, occasional drinkers — those who had less than about 20 grams of alcohol a week — served as the reference group. A standard drink in the United States contains about 14 grams of alcohol, according to the National Institutes of Health.

The study also had some limitations. For example, most of the information was self-reported and people don’t always accurately remember what or how much they drink. The study population wasn’t very diverse. More than 94% of study participants were White, so the findings may not reflect the experience of people from other racial or ethnic groups.

The study defined low-risk drinkers as those who reported having up to 10 grams of alcohol daily for women and 20 grams a day for men. Those classified as moderate risk reported having between 20 and 40 grams of alcohol each day for men and 10 to 20 grams for women. High-risk drinkers were those who reported having more than 40 grams of alcohol daily for men and 20 grams each day for women.

When people reported getting more than 80% of their alcohol from a certain type of beverage, such as beer, or drinking in certain situations like with meals, they were classified as having a particular preference, too.

The study found risks at every level of regular drinking. People who fell into the low-risk category were about 10% more likely to die of cancer than those who reported only occasional drinking. Those who were moderate risk had about a 10% to 15% higher risk of dying of any cause and dying of cancer compared with occasional drinkers, while high-risk drinkers were about 33% more likely to die of cancer, heart disease and any other cause than those who drank only every once in a while.

The researchers further scored participants based on where they lived, a proxy for income level, and on their underlying health, with one point assigned for any of 49 health conditions.

The increased risks associated with drinking were more pronounced in people who scored as having lower incomes and more health conditions.

“So that’s an important thing because it should be taken into account, too, for public health interventions in these populations,” said study author Dr. Rosario Ortola, who is an assistant professor of preventive medicine and public health at the Universidad Autonoma de Madrid in Spain.

However, people who said they mostly drank wine or drank with meals had slightly lower risks of cancer and death, despite having socioeconomic and health-related risks, compared with those who drank only on occasion.

“So what we think is that maybe what is causing this beneficial effect is not wine itself or drinking with meals itself, but other factors that we cannot control,” Ortola said.

It could be, for example, that people who drink wine or who only drink with meals appreciate moderation in other areas of their lives too, or perhaps they’re more likely to have other healthy behaviors, like being physically active.

In other words, Naimi said, “Moderate drinking is maybe a reflection of a healthier lifestyle, but it is not its cause, and that’s a very important distinction to make.”

The researchers concluded that exceptions like these may point to benefits apart from alcohol, such as the antioxidants in wine or perhaps spacing drinks at meals. More research is required to understand these differences.