Move the degrees. See what rises with them.
The same thermal load discharges into five outputs. Drag from a mild day toward a hot one. The bar shows direction and relative magnitude.
Only the first line has an alert threshold. The other four climb the same scale without one.
The figures are illustrative: they apply published slopes (where they exist, e.g. +1.3%/°C for NSW presentations) from a 22°C baseline to show direction and relative magnitude, not to predict any individual's absolute risk. Sleep, decision and body appear as an index because their effect doesn't reduce to a single per-degree slope.
Not only heatwaves
Risk seems tied to extreme peaks. The NSW study shows a single hot day loading as much as a heatwave day, with the slope starting at moderate temperatures. It's the signature of a physiological mechanism.
At 21.9°C, the period mean, the state saw about 45.7 presentations a day. Each extra degree adds around 1.3%, and the effect grows from mild temperatures up. The lead author puts it plainly: "this is about hot weather itself, not climate change."
What air-conditioning moves
The strongest known moderator is cooling. Across 10 million U.S. students who retook the PSAT, a school year 1°F hotter cuts 1% of that year's learning, but only in schools without air-conditioning. Turn the AC on and the slope flattens, here and across the page.
The detail that isolates the mechanism: only school-day heat lowers scores, while hot summers and weekends have no effect. Heat disrupts instruction and sleep time. Cooling becomes cognitive infrastructure, and its unequal distribution shifts the burden onto those without access: sleep is eroded most in poorer countries, in older adults and in women.
One slope, seen from six disciplines
What gives the thesis weight is the convergence. Psychiatry, sleep science, education economics, behavioural economics, conflict research and public health arrive separately at the same monotonic relationship between degrees and the brain.
What is fact, what is reading
Measured fact
- NSW youth suicidality presentations rise about 1.3% per degree, from moderate temperatures up.
- Heat objectively shortens measured sleep across 68 countries.
- Without AC, a hotter school year lowers scores; with AC, the effect nearly vanishes.
- Interpersonal violence rises, on average, with heat across dozens of studies.
Reading and hypothesis
- A shared biological mechanism, through thermoregulation, sleep and mood, plausibly links these outputs.
- If heat messaging included the mind, prevention could act at the same thresholds.
- Cooling could be treated as mental-health infrastructure.
- The burden falls unequally, on those without access to cooling.
What this page does not say
Rigour here also means marking the limits hard, especially on a subject that touches suicidality.
Correlation, not simple causation
The associations are real and replicated, but heterogeneous. We don't claim a degree causes an outcome in a person. These are population slopes, with uncertainty.
Physical heat, not climate distress
The measured effect comes from physical body temperature, not from anxiety about climate change. These are different things; this page addresses only the first.
A health signal, not an alarm
We follow responsible-reporting guidelines. No methods, no sensational language, no acts framed as a response to a trigger.
Magnitudes differ widely
1.3% per degree and 13% in extreme heat are different numbers, and neither is a doubling. Every figure stays tied to its source and unit.
If you or someone close needs support now
- Romania: DepreHUB, 0800 801 200 (24/7)
- Romania, youth: Child Helpline, 116 111
- UK and Ireland: Samaritans, 116 123
- EU: emergency 112
Questions this page opens
Why does a heat warning tell you to drink water and avoid exertion, but nothing about sleep, mood or irritability?
If cooling is the strongest known moderator, should school air-conditioning be treated as public-health infrastructure?
Who absorbs the thermal load on the mind, and what does the map of cooling access look like in your city?
What would a mental-health alert threshold, calibrated on the same scale as the body alert, look like?