A new study from researches at Oxford and Cardiff said drinking three and a half glasses of wine a week may raise the risk of dementia and can harm the brain.

According to the research, men and women who drank a little over one unit of alcohol a day suffered a noticeable decline in their brains’ ability to function over four years.

We asked three doctors in North Texas what they thought about the study. See their responses below:

Dr. Gregg Shalan, Neurologist, Methodist Dallas

“I think that this study does have some merit. We know that too much alcohol can harm the brain and it stands to reason that smaller amounts like those looked at in the study could also have negative effects. One confounding factor in this discussion is wine. Wine contains many other substances in addition to alcohol, some of which may be beneficial. The alcohol, however, shifts the risk/benefit toward the negative direction.”

Diana Kerwin, M.D., Chief of geriatrics, Texas Health Dallas, Director, Texas Alzheimer’s and Memory Disorders

What are your general thoughts about the study?

Prior to this study, moderate alcohol intake has been studied regarding dementia risk and alcohol in “moderation” was not thought to increase risk of dementia and in fact was thought to have some beneficial/preventative effects. Heavy alcohol intake has been known to increase risk. I think this study raises a great point of what defines “moderate” intake and maybe what we have thought of as moderate, is actually too much for brain health and well-being (Too much of a good thing?).

Are you surprised at the findings?

It was surprising that the amount of alcohol that we define as moderate, is likely too much. And the amounts that are possibly moderate or okay when you are younger, do not apply after age 65. The study suggests that everyone should decrease alcohol consumption as they get older to protect brain function.

Would you tell patients to stop drinking wine/alcohol all together?

I have always advised my patients to reduce their alcohol consumption as they get older and also if they are on several medications, reducing alcohol consumption is safest to reduce a risk of a medication interaction.

Anything else you’d like to mention about it?

I think the study is a good reminder in the fact that we may not know what exactly is the safest level of alcohol consumption. So, limiting your alcohol, it appears, is the best way to get some of the benefits without increasing your risks.

Aparna Kotamarti, M.D., Medical Director, Texas Health Fort Worth’s Senior Health and Wellness Center

What are your general thoughts about the study?

It was well understood that increased alcohol usage can increase the risk of dementia.

This study covered a smaller amount that had been felt not to cause much issues.

It was a comprehensive study that studied small amounts of use and it was very thorough and informative.

Are you surprised at the findings?

No surprise, as alcohol is known to have effects on the central nervous system.

Would you tell patients to stop drinking wine/alcohol all together?

Just like we counsel against the use tobacco, I would counsel limiting usage of alcohol.

Anything else you’d like to mention about it?

More studies need to be conducted.

Dr. Brendan Kelley, Neurologist with UT Southwestern’s O’Donnell Brain Institute:

The long-term impact of alcohol use on cognition is controversial. Numerous epidemiologic studies investigating alcohol consumption and the incidence of cognitive impairment or dementia have found reduced risk with light to moderate alcohol consumption. Taken together, these studies have reported a reduced rate of cognitive decline in light and moderate drinkers compared to abstainers and heavy drinkers.

This recently published study addresses an important topic relevant to lifestyle choices and public health. However, the data presented and the manner in which cognitive performance was measured do not provide conclusive evidence of sufficient strength to support clinical recommendations to patients or to the public.

For instance, the study used only a single cognitive measure (computerized measure of reaction time) to gauge cognitive performance. This strategy is not really adequate to assess a person's true cognitive performance, and certainly not adequate to assign a diagnosis of dementia. The reported effect size related to alcohol use was similar in magnitude to the difference seen between men and women and to the impact of educational attainment.