ArcaMax

DEAR DR. ROACH: I'm a 70 year old woman in good health, other than osteopenia and digestive problems that plague many seniors. I went through a sleep study and was found to have somewhat severe periodic limb movement. I do wake up frequently, always use the bathroom, and can be awake for a while before going back to sleep.

I was prescribed 100 mg of gabapentin, which does help. But I worry about dementia since the frequency and length of time of taking gabapentin determines the risk. My watch does indicate a longer, deeper sleep when I take it, but there's a trade-off. Do I get the sleep from gabapentin (which I wonder is more sedation than true deep sleep) or continue without it? I'm taking iron every day to boost my ferritin levels. -- K.S.

ANSWER: The evidence that gabapentin increases the risk of dementia is mixed, with some trials showing a risk, while others don't. None of the trials that suggested an increase in risk were prospective trials, which are considered the highest-quality trial.

You are right that the length of time of taking the medication could be linked to dementia, but most studies have found that the dose is what increases the risk. The dose that you're taking is a miniscule dose. My experience is that it takes a minimum of 300 mg three times daily for most people to get relief from pain, which is one of the most common reasons to prescribe the drug. (In a study in people with nerve pain after shingles, the goal dose was 1,200 mg three times daily -- 36 times the dose you are taking!)

You also need to consider that poor sleep and not getting enough sleep are both bad for your brain. Since you have evidence that your sleep is better with gabapentin, you should consider this when weighing the risks and benefits. In my opinion, the very low dose of gabapentin is very unlikely to significantly increase your dementia risk. The benefit that it provides to sleep likely outweighs any existing risk, which is likely to be small.

DEAR DR. ROACH: A recent column concluded that there are no current treatments for long COVID. But GLP-1 agonists have shown to be promising, and if I suffered from long COVID, I'd certainly pursue it. I'm hoping you can address this. -- J.R.

ANSWER: I'd say that the use of GLP-1 agents like semaglutide or tirzepatide is hoped to be effective, as there isn't enough evidence that I can call promising at the time that I'm writing. Scientists who study long COVID think that the multiple mechanisms of GLP-1 drugs suggest a potential benefit, but in absence of any good data, I can't recommend them.

It's often the case that when a drug is found to be very effective for one or more indications, it's tried very widely for other conditions. Sometimes there is marked success, but statistically, it's unlikely. GLP-1 agents have many indications, but they remain expensive and not well-covered by insurance, except for clear indications that are approved by the Food and Drug Administration. (Sometimes, even then, it can be hard for me and my patients to get insurance to cover them.)

========

Dr. Roach regrets that he is unable to answer individual letters, but will incorporate them in the column whenever possible. Readers may email questions to ToYourGoodHealth@med.cornell.edu.

(c) 2026 North America Syndicate Inc.

All Rights Reserved