Dear colleagues and members of the academic community,
I warmly welcome the introduction of the new #MentalHealth Topic and the opportunity it creates for more visible and connected discussion in this field.
I am writing this because I can no longer discuss the psychological consequences of war only in the language of scales, prevalence rates and confidence intervals.
Lately, I have received invitations from this community to share new research, comment on scientific developments and contribute to professional discussions. I am grateful for them. But some of the data I am processing are also written in sleepless nights, interrupted working days, damaged universities and hospitals, destroyed warehouses, disrupted supply chains, and the constantly changing conditions under which ordinary civilian life has to continue.
In early September, Kyiv endured seven consecutive days of Russian aerial attacks.
During that period, Russian drones struck the National University of Food Technologies in Kyiv, damaging university buildings. President Volodymyr Zelenskyy reported that 160 students were in a shelter. The university itself reported no casualties among students or staff and noted that, because of the security situation, the educational process had already been moved online.
That detail matters.
The absence of casualties does not mean that ordinary civilian life was unaffected.
It means that a university in a major European capital had already reorganised education around the possibility of being attacked.
This is often described through the familiar vocabulary of war-related mental health: fear, anxiety, traumatic stress, sleep disturbance and uncertainty.
All of these are real.
But I believe we may be missing something.
Beyond fear: a hypothesis about coerced loss of everyday autonomy
As a medical scientist studying psychological distress, sleep disturbances and coping among Ukrainian youth and families, I have begun to wonder whether repeated aerial attacks produce another form of psychological harm that we rarely measure directly.
My hypothesis is this:
Repeated attacks and air-raid alerts may damage mental health not only through perceived threat, but through the repeated, externally imposed disruption of everyday autonomy and agency — through the experience that a foreign aggressor can repeatedly override the ordinary decisions through which people organise their own lives.
An air-raid alert does not merely tell me that I may be in danger.
It tells me when I am allowed to sleep.
It determines whether I can continue working.
It interrupts education and changes where students are allowed to be.
It affects whether public transport functions, whether I can take a walk, go to a pharmacy, buy groceries, meet another person, or simply remain where I chose to be.
I can decide at midnight that I need sleep because I have work in the morning.
Minutes later, someone hundreds of kilometres away can effectively revoke that decision.
I can decide that I need food.
Then an attack begins, transport stops, a route becomes unsafe, or a warehouse that supplies shops is destroyed.
I can plan tomorrow.
But tomorrow remains conditional.
The harm is therefore not merely:
“You may die.”
It is also:
“Your time is not entirely yours. Your routines are not entirely yours. Your decisions can be overridden at any moment.”
That distinction matters.
When loss of control has an author
Loss of control is not unique to war.
Illness can take control over a person's day. So can disability, a natural disaster, an accident, economic hardship or many other circumstances that constrain human choice.
But there may be an important psychological distinction here.
Illness does not decide that I should wake at 3 a.m.
A storm does not intend to prevent a child from attending school.
An accident does not repeatedly adapt its behaviour in order to overcome the systems designed to protect me.
Here, the disruption has an external human source.
The conditions of my ordinary life are being altered by the actions of a foreign aggressor over whom I have no control and to whom I have given no authority over my life.
That source may matter psychologically.
The experience is not only:
“Circumstances prevent me from doing what I intended.”
It is also:
“Another actor can cancel my decision.”
Self-Determination Theory identifies autonomy — the experience of one's behaviour as volitional and self-endorsed — as one of three basic psychological needs, alongside competence and relatedness. Importantly, the theory distinguishes between simply lacking need satisfaction and actively experiencing need frustration. Autonomy frustration involves pressure, conflict and the experience of being pushed in an unwanted direction.
Albert Bandura approached the same territory from another direction. In his agentic account of Social Cognitive Theory, human agency depends on the capacity to influence the nature and quality of one's own life. Self-efficacy is not simply optimism; it concerns the belief that one's actions can produce intended effects.
But what happens when a person remains perfectly capable of acting, yet the environment repeatedly prevents intention from becoming action?
What happens when the lesson learned is not:
“I cannot cope.”
but:
“My decisions are provisional because another actor can cancel them.”
This may represent an additional pathway of psychological injury.
Threat can produce fear.
Unpredictability can produce anxiety.
Repeated attacks can fragment sleep and maintain physiological arousal.
Repeated externally imposed interruption may also produce autonomy frustration, anger, humiliation, helplessness, defiance and erosion of perceived agency.
These mechanisms interact.
What remains unclear is whether autonomy frustration and erosion of perceived agency constitute an additional pathway within this interconnected response — one that is currently under-recognised in war-related mental-health research.
A mechanism familiar to another discipline
There is an uncomfortable interdisciplinary asymmetry here.
Military theory has long treated coercion, morale, will and decision-making as legitimate objects of strategic analysis.
The purpose of coercion is not necessarily limited to physically destroying an opponent's capabilities. It can also involve creating conditions under which behaviour changes, resistance becomes more difficult, or decisions are made under externally imposed pressure.
Mental-health research often looks at the other end of this process.
We measure fear.
We measure anxiety.
We measure traumatic stress.
We measure sleep disturbance.
We measure whether people continue to function.
But there is a question between the military mechanism and the psychological outcome that may receive too little attention:
What does repeated externally imposed control over ordinary life do to human agency itself?
The fact that people understand the mechanism does not neutralise it.
I can recognise that my autonomy is being constrained. I can analyse it through Self-Determination Theory, Bandura's concept of agency, or the military literature on coercion.
That awareness does not restore the ability to make the interrupted decision real.
Understanding why someone has taken control over part of my day does not give that part of the day back.
The occupation of ordinary life
For me, the current pattern carries an uncomfortable sense of déjà vu.
At the beginning of the full-scale invasion, Kyiv faced the possibility of encirclement and the disruption of essential supply routes. Food, medicine, transport and communication were not abstract concepts. They became the architecture upon which survival depended.
Four years later, the weapons and military circumstances have changed, but disruption of the systems that make civilian life possible remains painfully familiar.
Educational institutions have been damaged since the earliest months of the full-scale invasion. By 2026, thousands of educational facilities had been damaged or destroyed, including schools and universities.
Health care has been subjected to the same sustained pressure.
By May 2026, WHO had verified more than 3,000 attacks on health care in Ukraine since February 2022, around 80% of them affecting hospitals, outpatient clinics and other health-care facilities.
One of the most widely known examples is Okhmatdyt, Ukraine's largest children's hospital in Kyiv.
On 8 July 2024, a missile directly struck the hospital complex. Intensive-care, surgical and oncology services were damaged; children receiving cancer treatment and dialysis had to be evacuated.
This was not the first time the war had reached the hospital. During the first days of the full-scale invasion in 2022, shelling near Okhmatdyt killed a child and injured others, and subsequent missile interceptions over the hospital damaged its buildings while it continued to receive wounded children.
Whether every individual strike on a civilian institution was deliberate is an important legal and military question.
It is not necessary to resolve that question in order to recognise the psychological mechanism described here.
For the person living inside this environment, the immediate reality is that external violence can determine whether a university teaches, whether a hospital provides treatment, whether transport moves, whether a child goes to school, whether I can sleep, and whether a plan made for tomorrow will still be possible tomorrow.
And this is not limited to institutions.
Recent attacks have destroyed or damaged warehouses storing food, books, medical supplies, children's goods and materials used in everyday production and construction.
In August, a Russian strike destroyed the warehouse complex of Lasunka, a major Ukrainian ice-cream producer, together with products prepared for distribution.
Another Russian drone strike destroyed approximately eight million books at a warehouse belonging to Ranok, one of Ukraine's major publishers and producers of school textbooks, shortly before the beginning of the new academic year.
In September, a warehouse belonging to KIDDISVIT, a major distributor of children's toys and goods, was destroyed in Kyiv. The company reported significant losses and had to reorganise its logistics.
On the same day, warehouses belonging to a long-established Kyiv producer of plywood and board materials were hit, damaging stocks of timber, building materials and industrial equipment.
A food-processing plant in Bila Tserkva producing instant noodles and other convenience foods was also damaged in Russian attacks.
Food.
Books.
Schools.
Universities.
Hospitals.
Medical supplies.
Children's toys.
Materials for homes.
Movement through one's own city.
Sleep inside one's own home.
These events differ in scale, immediate consequence and military context.
But psychologically they may contribute to a single environment: an environment in which ordinary civilian agency is continuously made conditional.
A metropolis can remain physically functioning while millions of ordinary decisions within it are made conditional by recurrent external threat.
And this is where statistics about “escalation” and “de-escalation” can become misleading.
An attack does not cease to affect human life when the number of missiles or drones falls below some previous peak.
A city does not psychologically return to normal between two alerts.
A child whose education is organised around shelter availability does not experience “de-escalation” as an abstract military trend.
Neither does the person who has slept in fragments for a week.
Neither does the researcher attempting to analyse data while listening for the next alert.
Infrastructure is autonomy made collective
There is another reason this question matters.
The ability to compensate for lost control is not distributed equally.
I can sometimes resist disruption.
I can change a route, postpone a task, reorganise my work, improvise, become stubborn.
None of this restores full autonomy, but it can allow me to reclaim a small part of it.
A child has fewer options.
An elderly person may have fewer options.
A bedridden patient has fewer options.
A person dependent on electricity for medical equipment may have almost none.
An animal cannot understand why familiar routines suddenly collapse, why the walls shake, or why the person it depends on suddenly behaves differently.
The more dependent a living being is on functioning civilian systems, the less capacity there is to compensate when those systems fail.
This is why damage to food supply, electricity, transport, health care and education is not merely damage to “infrastructure.”
Infrastructure is autonomy made collective.
It is the network of systems that allows millions of individuals to make ordinary decisions and reasonably expect those decisions to remain possible.
I decide to travel, and transport makes that decision actionable.
I decide to study, and an educational institution makes that decision actionable.
I decide to obtain medicine, and supply chains, electricity, pharmacies and health systems make that decision actionable.
I decide to buy food, and production, storage, logistics and retail make that decision actionable.
Personal agency does not exist in isolation.
It depends partly on a functioning social environment that allows intention to become action.
When that environment is repeatedly disrupted, what is damaged is not only convenience.
The relationship between choice and consequence is damaged too.
What are people being forced to adapt to?
I am not describing this only as a researcher observing a population under chronic threat.
I live inside the same conditions I study.
I know what it means to work after fragmented sleep, to make provisional plans, to reorganise ordinary tasks around alerts, and to continue functioning while knowing that decisions about my own day can be overridden from outside.
This does not make my observations more objective.
But it allows me to describe a dimension of exposure that is easy to miss when resilience is inferred simply from the fact that people continue to work, study, care for others and fulfil responsibilities.
Our own research suggests that adaptation under prolonged threat is heterogeneous rather than uniform.
Distinct adaptation clusters can be observed.
Those findings address one question:
How do people respond?
The question I am raising here is slightly different:
What, precisely, are they being forced to adapt to?
What should we be measuring — and when?
War-related mental-health research therefore needs to ask more than:
How frightened are you?
How often do you experience symptoms of anxiety or post-traumatic stress?
How badly has your sleep been disrupted?
These remain important questions.
But perhaps we should also ask:
How much control do you feel you have over the organisation of your day?
How often are your intended actions involuntarily interrupted?
Can you reliably make plans for tomorrow?
Do you feel that your own decisions still determine what happens in your everyday life?
How often are you forced by external threat to act against your immediate choice?
When this happens, what do you experience: fear, anger, humiliation, helplessness, defiance — or something else?
And perhaps another question is necessary:
Does it matter who or what is taking that control away?
These are not semantic distinctions.
If autonomy frustration and erosion of perceived agency form a partially distinct pathway between repeated attack exposure and psychological distress, then we may be overlooking part of the mechanism through which prolonged war affects health.
But recognising a potentially important mechanism does not mean that it should be measured immediately or at any cost.
The exposure is ongoing.
People are not reflecting on a completed traumatic event from a position of safety.
They are still living inside it and adapting to changing conditions in real time.
And every questionnaire is, to some extent, an intervention.
It directs attention.
It selects which experiences deserve examination.
It asks a person to name, compare and organise what may otherwise remain partly unexamined.
That can be useful.
It can also be burdensome.
In my own research, I try to design questionnaires so that they allow some reflection while the overall experience remains containing rather than destabilising.
The sequence matters.
The tone matters.
The amount of attention directed toward distress matters.
And the participant should not leave the research encounter carrying more psychological weight than they brought into it.
This becomes especially important when studying autonomy.
If our research concerns the loss of autonomy, the research process itself should not reproduce that loss.
We should be cautious about asking people who are still under pressure to repeatedly examine how much control they have lost simply because the construct is scientifically interesting.
Some responses that might appear maladaptive outside this context may currently be helping people function, endure or preserve a sense of self.
Research should not require people to dismantle the strategies that are helping them survive.
The scientific question remains important.
So does the question of timing.
Perhaps some mechanisms need to be named now, observed carefully, and measured more fully when doing so no longer risks becoming another demand imposed on people whose lives already contain too many demands from outside.
Normalisation is not adaptation
There is a dangerous temptation outside Ukraine to interpret repetition as normalisation.
If attacks continue long enough, they become shorter news items.
Numbers replace names.
An air-raid alert becomes another notification.
A damaged school becomes another line in a daily summary.
A destroyed warehouse becomes another business loss.
But frequency does not transform coercion into normal life.
Nor should the ability of Ukrainians to continue working, studying, publishing papers, raising children and caring for one another be mistaken for evidence that the pressure has become harmless.
Adaptation is not consent.
Functioning is not evidence of absence of harm.
And resilience should never be used to make suffering less visible.
The academic community has a particular responsibility here.
We study public health, education, human behaviour, development, resilience and human rights.
We should not allow the systematic disruption of these very foundations to disappear into background noise simply because it has lasted too long to remain novel.
What concerns me is a world capable of watching people progressively lose control over sleep, food, education, movement, treatment, work, ordinary purchases and the possibility of planning their own tomorrow — and gradually deciding that this is simply how those people live now.
It is not.
And it must never become normal.
Acknowledgement: This text was developed in dialogue with GPT.