Frontotemporal dementia fails differently from the other dementias. The ones before it in this series get missed because they look like ageing. This one gets missed because it looks like a psychiatric illness — and because it arrives in people in their fifties, an age at which almost nobody is looking for dementia.
This is the fifth post in our series on the dementia diagnoses, after mild cognitive impairment, Alzheimer's disease, vascular dementia and Lewy body dementia. It also contains the single strongest piece of evidence in our whole library for why we are careful about cognitive-enhancement compounds — a finding that has nothing to do with marketing and everything to do with a trial result.
We are not doctors, and nothing here is treatment advice.
Frontotemporal dementia is degeneration of the frontal and anterior temporal lobes — the regions governing personality, social behaviour, judgement and language. It splits into two broad clinical pictures:
Behavioural variant (bvFTD). Personality and conduct change first. Disinhibition — saying and doing things the person never would have. Apathy that looks like depression but without the sadness. Loss of empathy, so a previously warm person becomes indifferent to the people closest to them. Compulsive, repetitive behaviour. Changed eating, often a new craving for sweet food. Poor judgement with money.
Primary progressive aphasia. Language goes first. In the semantic variant, word meanings erode — the person can speak fluently but no longer knows what "spanner" refers to. In the nonfluent variant, speech becomes effortful and grammatically broken while comprehension holds up. In the logopenic variant, word-finding fails with frequent pauses.
In both, memory is often relatively preserved early on. A 56-year-old who can tell you what they did last Tuesday, but who has become cruel, reckless or strangely indifferent, does not fit anyone's template for dementia.
Because the symptoms are psychiatric symptoms. Behavioural disinhibition, apathy, compulsions and personality change in a person in midlife is a presentation that leads directly to a psychiatric clinic, and the differential there is depression, bipolar disorder, late-onset psychosis or a personality disorder — not neurodegeneration.
The clearest evidence that this confusion is real and systematic is that researchers have had to build studies specifically to resolve it. A 2026 multicentre study across five centres retrospectively reviewed MRI scans from 297 patients with sporadic behavioural variant FTD and 92 patients with primary psychiatric disorder, with a single rater blinded to the clinical diagnosis applying eight visual atrophy rating scales.1 Every scale showed more atrophy in bvFTD than in the psychiatric group. The orbitofrontal, anterior-temporal and fronto-insula scales discriminated best, and a composite of the three outperformed any scale alone.
Note what that paper is for. It exists because the two conditions are confused often enough, in specialist centres, that a tool to separate them was worth five centres' effort. Its framing is explicit: bvFTD is often misdiagnosed as late-onset primary psychiatric disorder because the symptoms overlap and there is no disease-specific biomarker.
A second 2026 study looked for differences in the autonomic nervous system, comparing FTD patients with primary psychiatric disorder patients and controls.2 FTD patients showed greater problems with temperature regulation (p = 0.026 versus psychiatric patients) and — the most striking difference — altered pain perception in 13% of FTD patients versus 2% of psychiatric patients and 1% of controls (p < 0.001). They also had smaller hypothalamic subregions on MRI, with autonomic symptoms tracking the volume loss.
That is a useful practical signal: a midlife behavioural change accompanied by a changed relationship to pain or temperature is a different proposition from one without.
This is the part worth reading twice.
A 2023 network meta-analysis searched for every randomised controlled trial of drug therapy for the neuropsychiatric symptoms of frontotemporal dementia. It found seven trials, with 243 participants in total.3 For a disease this devastating, that is the entire randomised evidence base.
What it found:
The authors' own conclusion is support for high-dose oxytocin and caution regarding the use of piracetam.
Now, piracetam. Piracetam is the original nootropic — the compound the word "nootropic" was coined to describe, in 1972. It is sold around the world as a cognitive enhancer, it is widely available, and it has a devoted following among people who take things to think more clearly.
In the only network meta-analysis of randomised drug trials in frontotemporal dementia, it made patients significantly, substantially worse — and the magnitude of that harm (+3.48) is three times larger than the magnitude of the best benefit in the same analysis (−1.17). It also measurably increased the burden on the people caring for them.
We have argued in why we don't sell brain blends that a compound with a mechanism story and no testing in the people who would actually take it is not a product, it is a guess. This is what that argument looks like when somebody finally runs the trial. Piracetam is not in our range and never has been. The lesson generalises past piracetam: "it's a nootropic, it supports cognition" is a claim about a mechanism, and mechanisms can run the wrong way in a diseased brain.
Put this post beside the last one and a clinically important contrast appears.
In Lewy body dementia, cholinesterase inhibitors produce a moderate cognitive benefit — standardised mean difference −0.53 across 17 trials — without worsening movement. In frontotemporal dementia, reviews of the pharmacological evidence conclude that the drugs developed for Alzheimer's appear ineffective for FTD's psychiatric symptoms, with the preliminary support instead pointing at serotonergic antidepressants.4
A separate systematic review of 23 studies and 573 individuals with behavioural variant FTD found trazodone produced the largest reduction in neuropsychiatric symptoms, with citalopram, paroxetine, rivastigmine and trazodone each reducing multiple symptoms including disinhibition, hyperorality and depression.5 Set that against the network meta-analysis finding trazodone's adverse event odds ratio of 9.53, and the honest summary is: possibly the most effective option available, and poorly tolerated. Both things are in the literature and both belong in the same sentence.
The broader point: two dementias that can look similar in clinic respond to drugs in opposite directions. That is the strongest argument there is for getting the specific diagnosis rather than the general one.
A 2026 systematic review assessed non-invasive brain stimulation across the FTD syndromes — 27 studies, mostly transcranial direct current stimulation or repetitive transcranial magnetic stimulation.6
tDCS combined with language therapy consistently improved language abilities in primary progressive aphasia, with some evidence the gains held over time. Benefits were most consistent in the nonfluent variant; effects in the semantic variant were more limited and domain-specific. rTMS showed short-term improvements in language and executive function, particularly after stimulating left frontal regions. In behavioural variant FTD the findings were heterogeneous, with social-cognitive outcomes more responsive than global cognition.
That is symptomatic potential, not disease modification, and the review says so. It is also more encouraging than the drug picture, which is a low bar.
We sell nothing that treats, slows or prevents frontotemporal dementia. No ingredient in our range has been tested in it. The genetic forms — associated with the C9orf72, MAPT and GRN genes, with a substantial minority of cases having a family history — are not amenable to anything on a supplement shelf, and the sporadic forms are not either.
The piracetam result is the thing we would most like readers to take from this page, because it applies well beyond this disease. The nootropics category runs on mechanism stories: this compound affects that receptor, therefore it supports cognition. Frontotemporal dementia is one of the few places where somebody put the oldest and most famous of those compounds into a randomised trial in people with a real brain disease — and it came out the wrong way, significantly, with a larger effect than any of the benefits on offer.
Our own range is not exempt from that logic, which is why we keep writing that exercise has better evidence than our supplements do, and why the risk-factor work in modifiable dementia risk matters more than anything we sell.
If what brought you here is a family member in their fifties whose personality has changed — who has become uncharacteristically rude, reckless, indifferent or compulsive — the useful thing to know is that "this is not depression and it is not a midlife crisis" is a reasonable thing to say to a doctor, and that asking specifically about frontotemporal dementia is reasonable too. Average diagnostic delays in this condition are measured in years, and the people who shorten them are usually the family.
Six diagnoses, each written the same way: what it is, what the evidence actually shows, and what we do and do not sell for it.
This article is for informational purposes only and has not been evaluated by the FDA. It is not intended to diagnose, treat, cure, or prevent any disease. We are not doctors. Frontotemporal dementia requires specialist diagnosis, and its genetic forms have implications for families that belong with a genetic counsellor. Nothing here is a reason to start, stop or change any prescribed medication. If someone in midlife is undergoing a marked change in personality, behaviour or language, that warrants medical assessment rather than watchful waiting.