Loneliness has been described as a public health crisis, compared to smoking in its effects on lifespan, and declared an epidemic by health ministers in multiple countries. But until recently, science could say little about what loneliness actually is inside the brain: what neurons fire, what circuits activate, what the biological difference is between being alone and feeling alone. A wave of new research is beginning to change that.

The neurons that register isolation

In 2025, Harvard neuroscientist Catherine Dulac published a study that may be the most precise look yet at what loneliness does to the brain at the cellular level. Her team at the Howard Hughes Medical Institute isolated female mice for five days, with brief social reunions on alternating days, and used genetic tools to map exactly which neurons became active during isolation and which ones fired during reunion.

They found two distinct clusters in the hypothalamus, a brain region involved in regulating hunger, thirst, and other survival drives. One cluster activated during isolation and produced an aversive, hunger-like state: an internal craving for company. A second cluster activated during reunion and connected to the brain’s dopamine reward system, producing something that felt like relief or satisfaction. The two circuits worked in opposition, like the push and pull of appetite and satiation.

“Loneliness is not a mood. It is a drive, like hunger,” Dulac said. “It has a specific neural substrate, and when that drive is satisfied, the brain registers it as a reward.”

Touch as the only resolution

One of the most striking findings from Dulac’s experiments was what did and did not resolve the isolation signal. When mice were separated from each other but could still see, smell, and hear their companions through a transparent barrier, the loneliness neurons kept firing as if full isolation were in effect. Only physical contact, specifically tactile interaction with another mouse, quieted the circuit. The researchers compared this to being able to see food through glass: awareness of the thing you need is not the same as having it.

This may help explain why social media use, which provides visual and textual contact but not physical presence, so often fails to relieve loneliness and in some studies appears to worsen it. If the loneliness circuit evolved to respond to physical co-presence, digital proximity may not speak the same language.

Loneliness as a survival signal

Parallel work from Kay Tye’s lab at the Salk Institute and MIT has identified loneliness-related circuits in the dorsal raphe nucleus, a brainstem region involved in social motivation. Tye’s team found that isolating mice caused a specific population of dopamine neurons to become hyperactive, priming the animal to seek social contact urgently. When these neurons were artificially activated in grouped mice, the mice behaved as if they had been isolated, seeking out company even when company was already present.

Together, the Dulac and Tye findings point toward a view of loneliness not as an emotional failure or a personality trait, but as a biological alarm system. Social species need each other to survive. A brain that registers isolation as aversive and social contact as rewarding has a selective advantage over a brain that does not. Loneliness, on this view, is evolution’s way of keeping social animals social.

What chronic loneliness does to the body

The biological effects of prolonged loneliness are well documented and extend far beyond mood. Chronic loneliness is associated with elevated cortisol and inflammatory markers, disrupted sleep architecture, impaired immune function, and increased risk of cardiovascular disease. A widely cited analysis by Julianne Holt-Lunstad found that social isolation increases mortality risk by roughly 26%, comparable to smoking 15 cigarettes a day.

The new cellular research offers a possible mechanism: a loneliness drive that never gets satisfied keeps the stress-related circuits in the hypothalamus chronically activated. The same system that evolved to push animals toward social contact in the short term becomes damaging when the craving goes unmet for weeks or months or years.

The hardest question

What the neuroscience cannot yet answer is why some people feel lonely in a crowd and others feel fine in extended solitude. Individual differences in the sensitivity of these circuits, shaped by genetics, early experience, and learned patterns of relating, probably account for much of that variation. Dulac’s lab is now looking at how early social deprivation changes the development of loneliness circuits in ways that may persist into adulthood.

What the research does establish, with increasing clarity, is that loneliness is not a character flaw or a sign of weakness. It is what a working brain feels when one of its core biological needs goes unmet. That reframing may matter less for the science than for how we treat the people around us who are quietly running on empty.

Source

  • Dulac, C., et al. (2025). Hypothalamic circuits mediating the aversive state of social isolation. Howard Hughes Medical Institute / Harvard University.
  • Tye, K.M., et al. (2018). Social reward and loneliness circuits in the dorsal raphe nucleus. Cell.
  • Holt-Lunstad, J., et al. (2015). Loneliness and social isolation as risk factors for mortality. Perspectives on Psychological Science.