Suicide warning signs can be quieter in our kids
If you need help right now: call or text 988, the Suicide and Crisis Lifeline, or chat at chat.988lifeline.org. Free, confidential, 24/7/365.
In Kansas, the Family Mobile Crisis Helpline takes calls and texts at 1-833-441-2240 for anyone 20 and under, 24/7/365, and it can send a mobile team out to you.
Johnson County residents can also reach the same crisis team directly at 913-268-0156. Wyandotte County is 913-788-4200. If your child is in immediate danger, go to your nearest emergency department.
The sun is setting and your kid has been in their room since abandoning their shoes at the door at 4:15pm. You find yourself staring at the woodgrain of their door, lost in thought. You already asked if they were okay and got a yes, delivered in the tone that lets you know they are finished talking.
So you are on the other side of the door like Anna, except they definitely do not want to build a snowman. Or answer a question. The silence feels cold and spikes your anxiety.
Or is it nothing, and they just need the space to recharge, like they always have.
Signs of withdrawal are harder to notice when your kid is "like that." They have always been a bit "intense," maybe an old soul, maybe they just need to shut down a bit after school. They have always needed a couple of hours to decompress before reliving the day and starting homework.
That conundrum is complicated and that is what this post is about.
The list was written for a different nervous system
You have seen the standard suicide warning signs. Talking about death or wanting to die. Withdrawing from friends and activities. Giving away possessions. Big mood changes. Hopelessness. Increased substance use.
That list is correct and it saves lives. It was also built around how distress tends to surface in a nonautistic, non-ADHD nervous system, and there are three places where it quietly fails a neurodivergent kid.
It assumes distress shows on a face. The 988 Suicide and Crisis Lifeline hosts an autism-specific guide that says this outright: many autistic people can report suicidal thoughts without necessarily showing the same degree of distress a non-autistic person would show. Reduced facial expression is a feature of how a lot of autistic people communicate. It tells you nothing about how much pain is in the room. If you are waiting to see it on your kid's face before you take it seriously, you may be waiting for something that is never going to arrive.
It assumes withdrawal is obvious. A kid who has always needed to disappear into their room to recover from a school day does not trip a withdrawal alarm when they start disappearing for a different reason.
It assumes the kid will say something. Some kids will. Some kids have never in their life been able to put "I am not okay" into words on demand.
That last one is the trap. The whole list works by asking you to spot a change. A kid who has never been able to say it is not going to start now, so there is no change to spot. Their silence on the worst night of their life looks exactly like their silence on an ordinary school night. You are being asked to notice a difference the kid has no way of producing.
There is a fourth version of this problem, and it comes home in a report card. Your kid's teacher says they are doing great. A pleasure to have in class. No concerns.
Masking is the work of looking fine, and it runs on the same energy as everything else. Tracking adult faces for whether you are in trouble. Copying what the other kids do at lunch so you do not get it wrong. Holding still in a room that is too loud and too bright for six hours. None of that shows up on a report card, and all of it comes out of what they had for the day.
There is some research connecting camouflaging to suicidal thinking. While it is early, and the effects are small, and none of it has been done in children, what sits underneath it is better supported: the more constantly you have to hide who you are to move through a place, the more that place starts to feel like a trap with no way out. Researchers call that defeat and entrapment. Your kid might call it the average school day.
The kid who holds it together all day and falls apart at 4:15 made it to a safe place first, and the falling apart is what safety looks like from the outside.
What to watch instead is your own kid
The 988 guide is free, and the last page is a one-page summary you can print. Its own threshold statement is the most useful sentence in it: the signs count when they represent a marked increase or change from what is usual for that individual.
Not different from other kids. Different than usual for yours.
That reframes the whole exercise. You are not scanning a checklist. You are noticing that something moved, in a kid whose baseline only you know.
What changes in how they act. A sudden increase in withdrawal, measured against last month rather than against other kids. Self-harm getting more frequent or more severe. Anxiety or depression symptoms getting worse, which comes with a warning attached in the guidance itself: do not attribute worsening depression or anxiety to autism or ADHD alone. That misattribution is one of the most common ways these kids get missed.
Watch more closely if the withdrawal arrives alongside your kid losing interest in something they used to love. The guide tells the story of a girl who stops making her clay animals and does not replace that with anything. This one deserves some care, because the research on a dropped interest as a signal on its own is thin, and it is not a red flag the evidence supports yet. But for a lot of kids the interest is the regulating structure of the whole day. The research here is autism-specific, and we see the same thing in kids who are ADHD only. When it goes quiet at the same time everything else goes quiet, ask.
What changes in what they can say. A kid who is normally verbal stops being able to say what is wrong, in speech or in writing, especially on demand. Loss of language under acute distress is one of the ten signs on that list. When a kid goes quiet in the middle of getting upset, most of us read it as the storm passing, calm, de-escalation. Sometimes it is. Sometimes the words are gone and the distress is not.
Also worth knowing: something happened that you might not rate as a big deal. The guide calls this a current traumatic event, reported by the kid or by someone else. Watch the "someone else" part, because what knocked your kid sideways is often something no adult in the building logged as an event at all. The friendship that ended. The teacher who changed the seating chart. The thing that got said in the group chat. We cannot measure its weight on our scale. We have to use theirs.
A lot of our kids have also been saying painful things about themselves for years. I don't belong. Nobody likes me. Everyone thinks I'm weird. Those are statements about how they are treated and where they fit, and they hurt to hear.
"I don't want to be in the world" belongs in a different category. So does "everyone would be better off," and "what's the point of any of it." Those are statements about whether to keep being here at all. If you have gotten used to the first kind, the second kind can slide right past you sounding like more of the same.
What changes in what they think about. Thoughts that loop for hours. In the largest study available, about one in eight autistic people described suicidal thinking that ran eight hours or longer at a stretch, and roughly a fifth described five hours or more. Most of us are calibrated to something much shorter. If your kid tells you they are stuck on something and cannot get off it, take the word "stuck" literally.
A new focus on death or suicide that is separate from an existing interest. This distinction takes knowing the kid. A teenager who has been deep in a dark series for two years is not the same as a teenager who started reading about methods last month.
Searching, planning, stockpiling. Internet searches for methods, detailed plans, saving up medication.
And hopelessness, which mostly shows up as a blank where the future should be. Ask what they think next summer looks like. If there is nothing there, that is information.
How to ask so you get an answer
Ask the direct question. "Are you thinking about killing yourself?" Not "are you having dark thoughts," not "you're not thinking about doing anything, are you." Concrete and specific, especially with autistic kids, who will answer the question you actually asked instead of the one you were gesturing at.
If the reason you have not asked is the fear that asking will put the idea there, that has been studied more than once and it does not appear to work that way. Asking does not appear to create the thought or make an existing one worse, and adolescents in one pooled analysis reported small reductions afterward. It is a modest body of research and some people do find the conversation itself hard. Neither of those is a reason not to ask.
Ask shoulder to shoulder, somewhere they can leave. On a walk, doing dishes, feeding animals. Looking at a face costs a kid bandwidth they need for the answer, and a kid who cannot exit a conversation is being cornered whether you meant it that way or not.
Give a way to answer that is not speech. A number from one to ten. A thumbs down. A text from the next room. Writing it down and handing you the paper. A kid who cannot say it out loud can very often still tell you.
This matters double for a kid with a Pervasive Drive for Autonomy profile (PDA, previously called Pathological Demand Avoidance), where a direct question can land as a demand and the nervous system answers the demand before it answers the question. Say out loud that they do not have to answer right now and that you will be around later. Then be around later.
Ask about the interest. "Are you still into Warriors/Roblox/Pokemon?" is a real check-in question in many homes. Ask it often enough that it is normal.
Keep your face steady. Steady is different from blank. Your regulation is what makes it safe for them to tell you the true version instead of the version that protects you.
Ask again later. One no is one moment, not a clearance.
Reduce access. Lock up medication. Lock up firearms, or move them out of the house. This is one of the few things in the entire literature with population-level evidence behind it, and it is entirely in your control.
Build the safety plan with them, not for them. Their sensory strategies, their interest, their people, their words. A plan written over their head does not get used at 11pm.
If you do one thing after reading this, print the one-page summary at the end of the 988 guide and put it where you will see it. It is free, it is the last page, and it is written for exactly this.
The part that isn't on the list
You are going to read all of that and start running it against your kid, and some of it is going to fit, because some of it fits most neurodivergent kids on a bad week. That does not mean you missed something. It means you are doing what we just asked you to do, and we want to be honest that it costs something.
Watching this closely can be hard. You will check the door more than you used to. You will hear a tone in a one-word answer and lose an hour to it. You will wonder whether you are being careful or being paranoid, and there is no clean line between those two when it is your own kid.
You are allowed to be scared of this.
Tell one other adult what you are watching for, so you are not the only person holding the baseline. A partner, your sister, the school counselor, your own therapist. This is not a competence problem, and needing someone to carry part of it is not a failure of nerve. Nobody should be doing this alone at 11pm with a closed door between them and the answer.
The last thing
Standing at that door running through all of it is not paranoia and it is not helicoptering. It is what it looks like to know your kid well enough to notice that something is different, and to have nowhere good to put that noticing.
You already have the baseline nobody else has. You built it by paying attention for years.
Ask the question.
Summit Ranch is not a crisis service. We do not provide emergency stabilization, and we are not the right call if your child is in danger right now. Use 988 (call, text, or chat), the Kansas Family Mobile Crisis Helpline at 1-833-441-2240, or your nearest emergency department.
Research
988 Suicide & Crisis Lifeline. (2021). Warning signs of suicide for autistic people. https://988lifeline.org/wp-content/uploads/2023/01/Warning-Signs-Resource-Sept-2021-2.pdf
The source for the warning signs section. Hosted by 988 and written by Lisa Morgan, Sarah Cassidy, Mary Donahue, Brenna Maddox, Teal Benevides, and Mirabel Pelton. Its subtitle describes it as "an autism-specific resource based on research findings and expert consensus," and it states plainly that the signs need more evaluation. Ten signs, each with scenarios, plus a printable one-page summary at the end. Loss of a previously intense interest appears in the scenario under sign 1 rather than as a sign of its own, which is why this post treats it as supporting detail and not a flag on its own.
Cassidy, S. A., Bradley, L., Cogger-Ward, H., et al. (2021). Development and validation of the Suicidal Behaviours Questionnaire, Autism Spectrum Conditions. Molecular Autism, 12, 46. https://molecularautism.biomedcentral.com/articles/10.1186/s13229-021-00449-3
Source for the duration finding. Of 689 respondents, 12.99% of autistic participants reported suicidal thinking lasting more than eight hours, and about 22% reported five hours or more.
Cassidy, S., et al. (2023). Is camouflaging autistic traits associated with defeat, entrapment, and lifetime suicidal thoughts? Suicide and Life-Threatening Behavior, 53(4), 572-585. https://onlinelibrary.wiley.com/doi/full/10.1111/sltb.12965
Connects camouflaging to suicidal thinking through defeat and entrapment. Cross-sectional, small effects, adult samples only, which is why this post frames it as a lead rather than a finding.
Blades, C. A., Stritzke, W. G. K., Page, A. C., & Brown, J. D. (2018). The benefits and risks of asking research participants about suicide: A meta-analysis. Clinical Psychology Review, 64, 1-12.
Source for "asking does not plant the idea." Eighteen studies, small but significant reductions in ideation after being asked, with adolescents showing roughly twice the reduction adults did. No study has tested this specifically in neurodivergent young people.
Zalsman, G., Hawton, K., Wasserman, D., et al. (2016). Suicide prevention strategies revisited: 10-year systematic review. The Lancet Psychiatry.
Source for reducing access to means. 164 studies. Well supported at the population level. The review's own conclusion is that no single strategy clearly stands above the others, which is why this post does not call it the most effective thing available.
