Off-Label SSRI Use for Repetitive Behaviors in ASD
Mixed-to-negative trial evidence for SSRIs targeting repetitive behaviors in autism, against a specific parent request driven by another family's reported experience.
Most days follow the same shape in the Alvarez house, and that sameness is by design: N.A., 11, autistic, functions best on a schedule his parents have refined over years of trial and error, down to which cereal bowl he uses and which route the car takes to school. What brings them in today isn't the routines themselves, which nobody wants to disrupt, but what happens when they're interrupted — and lately, even when they aren't. He has begun repeating a specific hand motion, a rapid tapping against his thigh, dozens of times an hour regardless of context, along with a verbal repetition of phrases from a show he stopped watching months ago. Neither behavior is new in kind; both have sharply increased in frequency over the past two months, to the point that his teacher has raised it as newly interfering with his ability to complete classwork.
His parents have read, in several places online and heard from another family at his school, that an SSRI helped their own child's repetitive behaviors, and they arrive today specifically asking for one. What they haven't encountered in that reading is the actual controlled-trial record on this exact question, which is considerably less encouraging than the anecdotes driving their request — and the team's job today is to give them that fuller picture honestly, not to simply grant or refuse what they've asked for. N.A.'s mother describes the other family's story in specific, vivid detail — a boy who reportedly stopped a years-long hand-flapping habit within weeks of starting an SSRI — in a way that makes clear how much weight one persuasive, personally-witnessed account can carry against an abstract reference to "the research," especially for parents who have spent years watching strangers' general advice fail to describe their own child. That asymmetry, a single memorable case against a statistic, is worth naming honestly rather than treated as a failure of the parents' reasoning; it is the same asymmetry that makes any single-family anecdote a genuinely poor substitute for a controlled trial, however much more persuasive it feels in the room.
A specific parent request against a mixed-to-negative trial record
I want to give them the actual record before we go further, because it's more specific than "mixed evidence" suggests. King and colleagues' multi-site, placebo-controlled trial of citalopram in 149 children with high levels of repetitive behavior found no significant difference from placebo — roughly a third of children improved in both arms — and the citalopram group had significantly more side effects, including increased energy and impulsivity, hyperactivity, and insomnia. Reddihough and colleagues' later multicentre trial of low-dose fluoxetine, in JAMA in 2019, is more equivocal but lands in a similar place: it did show lower obsessive-compulsive scores at sixteen weeks in the unadjusted analysis, by about two points on a twenty-point scale, but that difference disappeared in the trial's own prespecified adjusted analysis, and its authors concluded the evidence wasn't strong enough to recommend fluoxetine while explicitly declining to rule out benefit for some individual children.
I'd want to be careful not to let two trial averages fully close the door, though. Neither trial excludes the possibility of a real responder subgroup — SSRIs remain genuinely useful for comorbid anxiety in autistic children, which can itself drive or worsen repetitive behavior as a coping mechanism, and N.A.'s own history includes situational anxiety around routine disruption that hasn't been formally worked up.
That's a real distinction, but it's a different clinical question than the one his parents actually asked. Treating anxiety, if it's genuinely present and driving some of this, is a defensible reason to consider an SSRI. Treating the repetitive behavior itself as the primary target, on the strength of the specific trials I just described, isn't supported by what those trials actually found.
The path forward follows directly from that distinction: a formal anxiety evaluation first, since that's the actual evidence-supported indication if it's present, rather than starting an SSRI today framed as a repetitive-behavior treatment the trial record doesn't back. If anxiety is confirmed, the medication conversation happens on genuinely different, better-supported ground than the one his parents walked in with.
Agreed: anxiety evaluation referred, behavioral support intensified through the school in the meantime, SSRI conversation explicitly deferred pending that evaluation's result rather than started today.
His parents were walked through the actual trial data directly, including why a trusted anecdote from another family doesn't override two negative controlled trials on this specific question — they left, by their own account, still hopeful but genuinely better informed than when they arrived.