Choline: Selezneva et al., 2026
In a five-year Russian study of 122 first-degree relatives of Alzheimer's patients with subjective cognitive complaints, those who received four 3-month courses of choline alfoscerate (1200 mg/day) every 1.5 years showed stable or improved cognition on clinician ratings, while 16.7% of untreated controls progressed to mild cognitive impairment. This is an open-label observational study, no randomization, no placebo, which limits causal claims.
Key takeaways
- 122 relatives of Alzheimer's patients with subjective cognitive complaints were followed for 5 years; 56 received intermittent choline alfoscerate therapy, 66 did not.
- 96.8% of the treated group showed "marked or moderate improvement" on clinician global impression scales by study end; 16.7% of untreated controls progressed to mild cognitive impairment.
- This is an open-label study, no placebo group, no blinding, no randomization, which means expectancy and selection bias are uncontrolled.
- The dosing protocol (1200 mg/day for 3 months, repeated every 1.5 years) is intermittent rather than continuous, which is an unusual but pragmatic approach for long-term prevention.
- Choline alfoscerate is well-studied for dementia in Europe but remains underutilized in U.S. cognitive aging research.
The study
Selezneva, Roshchina, and Gavrilova published this five-year observational study in Zhurnal nevrologii i psikhiatrii imeni S.S. Korsakova in 2026 (PubMed link). They recruited 122 first-degree relatives (children or siblings) of Alzheimer's patients who had subjective cognitive impairment (SCI), self-reported memory or thinking problems without objective deficits on testing. Fifty-six subjects received four courses of choline alfoscerate (400 mg three times daily for 3 months) spaced every 1.5 years; 66 received no treatment. Both groups were assessed at 9 time points over 57 months using a battery of 10 cognitive scales, including the Clinical Global Impression scale. The treatment group was evaluated before and after each course, plus 3 months post-course. The control group was assessed at matched time points. This was not a randomized controlled trial, subjects were assigned to treatment or control by unspecified criteria, likely personal preference or clinician judgment.
How to read this study
What this paper does well: The five-year duration is rare and valuable, most choline trials run 6-12 months. Following high-risk individuals (relatives of AD patients) is the right population if you want to catch early decline. The repeated-measures design with 9 assessment points captures trajectory, not just snapshots. The intermittent dosing protocol is pragmatic, continuous supplementation for five years is expensive and adherence-challenged; cycling on and off every 1.5 years is a realistic strategy.
What this paper is missing: This is an open-label study with no randomization and no placebo. That means we don't know whether subjects chose treatment based on motivation, health literacy, or socioeconomic status, all of which predict better cognitive outcomes independent of any pill. The 96.8% "improvement" on clinician global impression could reflect expectancy bias from both raters and patients who knew they were being treated. There's no blinding, so the investigators knew who got choline and who didn't when scoring cognition. The abstract does not report dropout rates, which in a five-year study are critical, if healthier subjects stuck with treatment and frailer ones left, the improvement could be selection artifact.
How I'd weight this paper: I treat this as hypothesis-generating, not confirmatory. The 16.7% progression rate in controls is a useful baseline for future trials, and the absence of harm over five years is reassuring. But without randomization or placebo control, I can't separate the drug effect from the act of being monitored, cared for, and enrolled in a long-term study. This is evidence to justify a proper RCT, not evidence to change behavior on its own.
What they found
In the treatment group, clinician global impression scores improved after each 3-month course and remained improved at the 3-month follow-up assessments. By the end of five years, 96.8% of treated subjects were rated as showing "marked or moderate improvement." Specific cognitive test scores were not detailed in the abstract, only the global impression rating.
In the control group, cognitive function declined progressively. At 57 months (just under five years), 11 of 66 untreated subjects, 16.7%, had progressed from subjective cognitive impairment to mild cognitive impairment (MCI), which is the clinical stage before dementia where objective deficits appear on testing. The abstract states that "deterioration in cognitive functioning was identified" but does not provide numerical test score changes.
No serious adverse events were reported in the treatment group. The abstract does not specify dropout rates, reasons for discontinuation, or whether any subjects in the control group later received treatment.
What it means for the average man
If you have a parent or sibling with Alzheimer's and you're noticing memory slips, this study suggests choline alfoscerate might be worth considering as a long-term preventive strategy. The 16.7% progression rate in untreated relatives is a useful baseline, roughly one in six people with subjective complaints and family history will progress to measurable impairment in five years.
The intermittent dosing protocol, 3 months on, 15 months off, repeated, is more realistic than daily supplementation forever. But because this wasn't a placebo-controlled trial, we don't know how much of the benefit came from the choline versus from being enrolled in a structured monitoring program. The absence of harm over five years is the clearest takeaway: 1200 mg/day for 3-month cycles appears safe in this population.
Choline alfoscerate is available in the U.S. as Alpha-GPC. If you're considering it, the dose here was 400 mg three times daily, 1200 mg total, which is higher than most over-the-counter recommendations. Start with physician consultation if you're on cholinergic medications (donepezil, rivastigmine) to avoid overlap.
The caveats
The lack of randomization and blinding means this is an observational study masquerading as an intervention trial. We don't know whether the people who accepted treatment were already healthier, more motivated, or had better social support than those who declined. All of those predict better cognitive outcomes regardless of supplements.
The abstract provides no dropout data. In a five-year study, dropout is often 30-40%. If the frailer subjects left early and the healthier ones stayed, the 96.8% improvement rate would be an artifact of survivor bias. We also don't know what "marked or moderate improvement" means quantitatively, was it a 2-point gain on a 30-point scale, or clinically meaningful change? Global impression scales are subjective and vulnerable to expectancy bias when raters aren't blinded.
The study was conducted in Russia and published in a Russian-language journal. This doesn't invalidate the findings, but Western regulatory agencies will want replication in a pre-registered, placebo-controlled RCT before recommending choline for cognitive decline prevention.
Frequently asked questions
Should I trust a study without randomization or placebo control?
Not as much as a randomized controlled trial, but it's not worthless. Open-label studies like this one can show safety, establish feasibility, and generate hypotheses for proper trials. The problem is we can't separate drug effect from selection bias, expectancy, or the benefits of being monitored. If people who chose treatment were already healthier or more motivated, they'd do better regardless of the pill. Randomization is the tool that balances those hidden factors across groups. Without it, correlation doesn't mean causation.
What is choline alfoscerate, and how is it different from regular choline?
Choline alfoscerate, sold in the U.S. as Alpha-GPC, is a choline donor that crosses the blood-brain barrier more efficiently than choline bitartrate or phosphatidylcholine. Once in the brain, it's converted to acetylcholine, the neurotransmitter involved in memory and learning. The "alfoscerate" structure (glycerophosphocholine) makes it more bioavailable than cheaper choline salts. European studies in dementia consistently use 1200 mg/day, which is higher than most U.S. supplement labels.
Does having a parent with Alzheimer's mean I'll get it too?
Not necessarily, but it does increase risk. If one parent has late-onset Alzheimer's (after age 65), your lifetime risk is roughly 20-30% versus 10-12% in the general population. If both parents have it, or if onset was early (before 65), risk is higher. The ApoE4 gene is the strongest known risk factor, one copy roughly triples risk, two copies increase it 8-12 fold. But most Alzheimer's is multifactorial: genes load the gun, lifestyle pulls the trigger. Cardiovascular health, sleep, exercise, and cognitive engagement all modulate risk.
Is 1200 mg/day of Alpha-GPC safe long-term?
This study suggests it is, no serious adverse events were reported over five years of intermittent use. The most common side effect in choline trials is gastrointestinal upset (nausea, diarrhea), which is dose-dependent. A 2021 observational study flagged a possible association between choline supplementation and increased stroke risk, but that data was from food frequency questionnaires, not controlled trials, and the mechanism is unclear. If you have cardiovascular risk factors, discuss with your physician before high-dose choline.
Sources
- Selezneva N, Roshchina I, Gavrilova S. Efficacy of choline alfoscerate treatment for subjective cognitive impairment in first-degree relatives of patients with Alzheimer's disease: Results of a comparative 5-year study. Zh Nevrol Psikhiatr Im S S Korsakova. 2026. PubMed.
- Parnetti L, et al. Choline alphoscerate in cognitive decline and in acute cerebrovascular disease: an analysis of published clinical data. Mech Ageing Dev. 2001;122(16):2041-55.
- Budson AE, Solomon PR. Memory Loss, Alzheimer's Disease, and Dementia: A Practical Guide for Clinicians. 3rd ed. Elsevier; 2022.
These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.
