The observation that some individuals with autism show clinical improvement in response to fever suggests that symptoms may be modulated by brain systems or enzymes that become altered at high temperatures or by immune-inflammatory factors. The febrile hypothesis of autism stems from this observation. The effect could be due to the direct effect of temperature on enzymes that are heat-labile (can be changed or activated at high temperatures) or on gene expression in the brain. It could also be due to a resulting change in the immune inflammatory system or an increase in the functionality of a previously dysfunctional system in the locus coeruleus, a brain region that modulates physiological responses.

Eric Hollander and his team at Albert Einstein College of Medicine, Montefiore Medical Center in New York, explored the direct effect of elevated temperature on children with autism who have a history of clinical improvement associated with fever. Hollander hypothesized that elevations of core body temperature to the fever range may mimic the improvements observed when these children have a fever episode.

Using a large hot tub in a double-blind crossover study, clinicians and parents observed ten children with autism and a history of improvement during fever while they experienced either a hyperthermia (102 degrees F) or a control (98 degrees F) condition. During the hyperthermia condition, children swam in 102-degree water for approximately 45 minutes. Parents and clinicians completed ratings based on observations of the child while body temperature was elevated, immediately after the child exited the pool. To collect epigenetic and biomarker data, the researchers completed cheek swabs and assessed pupillary response prior to pool entry and during the hypothermic or control condition.

The researchers saw an improvement in socialization and in repetitive and restricted behaviors at the hyperthermia condition and demonstrated that they could reliably and safely increase children’s temperatures into the fever range (on average the children’s body temperatures reached 101.7 degrees F). Clinician ratings and parent ratings both significantly improved with this temperature increase. The researchers correlated each child’s fever response history with the improvements observed during the elevated temperature.