CBD for Older Adults: Safety, Interactions and What Research Shows

If you are weighing CBD for yourself or a parent, you are not alone: cannabis use among older Americans has risen in recent years (Han 2025). Here is what studies in older people found and what caregivers should know. We do not recommend products or doses.

At a glance

How many older adults use CBD?

In the 2022 National Survey on Drug Use and Health (NSDUH), a federal survey, 14.3% of adults 65 and older reported past-year CBD use, more than the 8.0% who reported cannabis use (Choi 2025). Past-month cannabis use in this age group rose from 4.8% in 2021 to 7.0% in 2023 (Han 2025). Among adults 65 and older who used cannabis in the past year (2021–2024), 19.9% said a doctor had recommended it (Han 2026).

Strength of evidence: these are self-reported survey figures. NSDUH does not include people living in institutions such as nursing homes (Han 2025).

What trials in older adults show

Side effects overall. A meta-analysis pooled 46 randomized trials (6,216 people, average age 58.6). THC-containing products increased side effects such as dry mouth, dizziness and drowsiness. CBD alone did not, but it was tested in only three small studies (about 50 people, average ages under 60, one a single-dose test); no one over 65 was studied (Velayudhan 2021). An update of 58 trials found THC raised problems with mobility, balance and coordination as the dose went up (Velayudhan 2024).

Strength of evidence: moderate for THC side effects; low for CBD alone (three small studies, about 50 people, no one over 65), too little to judge its safety in seniors.

Dementia-related agitation. Our dementia guide counts only two small controlled trials of pure CBD and rates that evidence very weak. In one, 30 people with vascular dementia took 300 mg of CBD a day or placebo for four weeks; behavioral symptoms fell slightly more with CBD on the two main scales (p = 0.05 and p < 0.05), with no difference in thinking or daily function. Three of the eight authors are co-inventors on a CBD-related patent (Pessoa 2026).

Results with products that also contain THC are mixed. In a 60-person trial (average age 79) of an oil with 30% CBD and 1% THC, agitation scores fell by 4 points or more in 60% of the treated group and 30% on placebo; this was significant by the usual test (p = 0.03) but borderline with the continuity correction the authors had also planned (p = 0.06). All eight dropouts were taking the oil, and memory loss and hallucinations were more frequent with it, though not significantly. Two of the nine authors worked for the oil's maker and held stock options (Hermush 2022). In a 75-person, three-week US trial, dronabinol (a prescription synthetic THC) beat placebo on one of two main agitation scales; only sleepiness was more common as a side effect (Rosenberg 2026). A Cochrane review, in which all four trials tested THC or THC-like synthetic drugs, said it cannot be certain whether cannabinoids help or harm people with dementia (Bosnjak Kuharic 2021). A 2026 meta-analysis found sedation about twice as common with cannabinoids as with placebo (Pan 2026).

Strength of evidence: very weak for pure CBD; weak for THC-containing products.

Parkinson's disease. In a 61-person randomized trial of a CBD-rich extract with a little THC, movement symptoms did not improve more than with placebo, and sleep, thinking and daily activities favored placebo (Liu 2024). In 33 people with Parkinson's who act out their dreams in sleep, 14 weeks of pure CBD did not reduce these episodes (de Almeida 2021). More in our Parkinson's guide.

Strength of evidence: weak; small, short trials.

Pain and arthritis. Across all ages, in 15 of 16 randomized trials using pharmaceutical-supplied CBD for pain, CBD was no better than placebo (Moore 2024). In 86 people with knee osteoarthritis (average age 62.8), 600 mg of CBD a day added to acetaminophen for eight weeks reduced pain no more than placebo; side effects and liver enzyme rises were more common with CBD (Pramhas 2023). A 12-week trial in 136 people with hand osteoarthritis or psoriatic arthritis also found no difference (Vela 2022). See chronic pain and arthritis.

Strength of evidence: moderate, pointing to no benefit for pain.

Sleep and anxiety. A 2026 review found that rigorous studies of cannabinoids for sleep that specifically enrolled older adults are "virtually nonexistent," and that evidence for CBD in sleep remains sparse (Draghmeh 2026). We did not find a randomized trial of CBD for anxiety in older adults (as of November 2026). Evidence across ages is in our insomnia and anxiety guides.

Strength of evidence: very low for older adults; there are essentially no trials in this age group.

Why age matters

Many medicines at once. In 2015–2016, 85.0% of US adults 60 and over had used a prescription drug in the past 30 days (NCHS 2019a), and 34.5% of those aged 60–79 had used five or more (NCHS 2019b).

A body that handles drugs differently. With age, the liver and kidneys clear many drugs more slowly, and older people are often more sensitive to blood thinners, heart medicines and psychiatric drugs (Mangoni 2004). A computer modeling study predicted generally higher CBD blood levels in older than in younger adults (Qian 2026). The label of Epidiolex, the FDA-approved CBD medicine, says its trials did not include enough patients over 55 to tell whether they respond differently, and advises caution in dosing elderly patients, who more often have reduced liver, kidney or heart function or take other drugs (Epidiolex label 2026).

Strength of evidence: a review of age-related changes, a computer model (not measurements in older people) and the drug label.

Enzyme interactions. CBD is broken down mainly by the liver enzymes CYP2C19 and CYP3A4, and it is a moderate inhibitor of CYP2C19. The label warns that CBD may make the blood thinner clopidogrel less effective, and may raise levels of medicines such as digoxin, theophylline and tizanidine (Epidiolex label 2026). With warfarin, a systematic review found seven case reports; the INR (a clotting test) rose in six, with very high or unknown cannabinoid doses (Smythe 2023). Check your medicines in our drug interaction tool and see blood thinners and warfarin.

Strength of evidence: drug label for the enzymes; case reports for warfarin.

Sleepiness and falls. CBD can cause sleepiness, and alcohol and other sedating drugs may add to it (Epidiolex label 2026). We did not find any study that looked at CBD and falls (as of November 2026). For THC, in 18 people with dementia (average age 77), 1.5 mg twice a day for three days increased body sway when standing with eyes closed; no falls occurred (van den Elsen 2017).

Strength of evidence: very low; no falls data for CBD.

Liver. The Epidiolex label calls for liver tests (ALT, AST and bilirubin) before treatment and at 1, 3 and 6 months (Epidiolex label 2026). In a trial run by FDA researchers in 201 healthy adults (median age 36), liver enzymes rose above three times normal in 5.6% of those taking CBD for four weeks, at a dose the researchers placed within the range consumers take, and in none on placebo (Florian 2025).

Strength of evidence: moderate for liver enzyme rises; older adults were not studied separately.

The THC problem

In California, emergency visits associated with cannabis among adults 65 and older rose from 20.7 to 395.0 per 100,000 emergency visits between 2005 and 2019 (Han 2023). Among older adults who used cannabis in the past year, 11.4% met criteria for cannabis use disorder (Han 2026).

Strength of evidence: hospital records and a national survey; they track cannabis in general, not CBD.

A "CBD" product can still carry THC:

• THC was found in 18 of 84 CBD extracts bought online (Bonn-Miller 2017).
• THC was found in 35% of 105 topical CBD products; four of them were labeled THC-free (Spindle 2022).
• Among 56 CBD gummy products, 39% contained delta-9 THC, and nine had more than 0.4 mg in a single gummy (Johnson 2025).

Strength of evidence: product testing studies; they show labels are unreliable, not how often people are harmed.

Starting December 11, 2026, unless Congress delays it again, federal law limits final hemp-derived cannabinoid products to 0.4 mg total THC per container (Pub. L. 119-37; Pub. L. 119-103). Both are enacted statutes; details in what changes and when. How to check a certificate of analysis: reading labels.

Nursing homes and long-term care

We did not find a CMS rule or guidance document on CBD use by nursing home residents (as of November 2026). Two clinicians described the tension for medical cannabis (marijuana): skilled nursing facilities must comply with federal law while residents may qualify under state programs (Palace 2019).

In five long-term care facilities in Geneva, an oil with THC and CBD (1:2) did not reduce agitation more than placebo in 25 residents with severe dementia, though as-needed psychotropic medicines were used less (Bianchi 2026). A UK care home study of a THC-and-CBD mouth spray in 29 residents with Alzheimer's was designed to test feasibility, not benefit; no adverse reactions were reported (Albertyn 2025).

Strength of evidence: low; two small trials.

For Medicare questions, see CBD and Medicare.

When to talk to a doctor or pharmacist

Talk with a doctor or pharmacist before starting CBD if you take a blood thinner, seizure, heart, sleep, anxiety or pain medicine, or five or more prescriptions; if you have liver disease; or if you have fallen recently. Bring all products and supplements you use, in their packages, and ask whether liver tests make sense. Caregivers can watch for new sleepiness, confusion, unsteadiness, bruising or bleeding and tell the care team.

Frequently asked questions

References