Loneliness can take many forms, but when it persists alongside depression or anxiety, researchers describe it as “toxic”. A new study of 455 adults seeking mental health support found that childhood neglect uniquely predicted anxious loneliness in adulthood, while childhood abuse uniquely predicted depressed loneliness.
The research was led by Sasha Rudenstine of The City College of New York. It was published in Psychology and Psychotherapy: Theory, Research and Practice and points to the need for prevention efforts targeting different developmental pathways.
In 2023, the US Surgeon General declared an epidemic of loneliness and isolation. Yet not all loneliness is alike. Most people feel lonely at times, often in response to a move, a breakup or another life transition. These passing experiences typically push people back towards social contact.
Researchers describe a different and more dangerous form of loneliness as “toxic”. It applies when isolation becomes chronic and is accompanied by depression or anxiety, locking people into a self-perpetuating cycle of withdrawal and distress.
Although depression and anxiety often co-occur with loneliness, they reflect distinct emotional states with different relational consequences. Most prior studies grouped them together or treated each in isolation. The new analysis instead treats anxious and depressed loneliness as distinct phenotypes, asking which adversities push someone towards each. The distinction matters clinically: a treatment that helps one form may do little for the other.
The clinical sample comprised 455 adults seeking psychological evaluation or therapy at a community-based mental health clinic in New York City. The average age was just under 28, with roughly seven in ten participants under 30 and almost two-thirds women.
The sample was racially and ethnically diverse: 46% White, 16% Black, with the remainder spread across East and South-East Asian, mixed-race and ethnically defined identities. Data were collected between September 2017 and March 2020.
Participants completed several validated self-report measures before treatment. These included a loneliness scale, the depression and anxiety subscales of the Brief Symptom Inventory (a screening tool for psychiatric symptoms), the adverse childhood experiences questionnaire, a trait emotional intelligence measure, and three items measuring perceived social support.
Loneliness, depression and anxiety scores were each split at the sample mean to mark people as high or low. Anyone scoring high on both loneliness and anxiety was classified as having anxious loneliness; anyone high on both loneliness and depression was classified as having depressed loneliness.
The team then ran two logistic regression analyses. This approach lets researchers ask which factors independently raise or lower the odds of each loneliness phenotype, while accounting for the others.
Adverse childhood experiences were grouped into three categories. Abuse covered verbal, physical and sexual maltreatment, while neglect covered emotional and material deprivation. Household dysfunction covered events such as parental divorce or having a family member with mental illness.
About a third of participants (33%) fell into the anxious loneliness group, and 37% into the depressed loneliness group. Both rates are high, reflecting that the sample was drawn from people already seeking mental health support. The headline finding was that the two phenotypes had different developmental signatures.
Childhood neglect more than doubled the odds of anxious loneliness, with no comparable effect on depressed loneliness. Childhood abuse showed the mirror image: it more than doubled the odds of depressed loneliness while leaving the odds of anxious loneliness unchanged.
Household dysfunction, the third category, did not independently predict either outcome once the other variables were taken into account.
Trait emotional intelligence, which captures a person’s perceived ability to recognise and regulate emotions, was strongly protective against both phenotypes. Higher scorers on the seven-point scale were roughly four times less likely to fall into the anxious loneliness group. They were also roughly seven times less likely to fall into the depressed group.
Perceived social support played a more selective role. People who felt they had reliable people to turn to in times of need were about half as likely to fall into the depressed loneliness group. Social support did not separately reduce the odds of anxious loneliness once emotional intelligence was accounted for.
The study has several important limitations. Cross-sectional data cannot establish that childhood adversity caused either phenotype rather than coincided with it.
Memories of childhood adversity were collected by self-report and may be subject to recall biases, particularly for people in current psychological distress. Depression and anxiety were assessed using a screening tool rather than a clinical interview. The high-distress groups should therefore be read as reporting clinically meaningful symptoms rather than confirmed diagnoses.
The sample was drawn from a single urban clinic in New York and skewed towards women and adults under 30. Findings may not generalise to older or rural populations, or to people who have not sought mental health care.
The authors propose that the divergent pathways reflect what each adversity does to a developing emotional system.
Neglect, they argue, leaves a child anxiously scanning for caregivers who are unreliable, fostering a hypervigilant style that later manifests as anxious loneliness in adulthood. Abuse, by contrast, attacks the developing sense of self, leaving people prone to shame, self-blame and the internalised negativity that drives depressed loneliness.
Either pattern can become self-sustaining, since both forms of adversity disrupt the early development of emotional and social skills that would normally buffer against later distress.
The most actionable implication is that tailoring matters. Someone whose loneliness is anxiously coloured may benefit from work on emotion regulation and the fear of rejection. Someone whose loneliness sits alongside depression may benefit more from work on self-concept and the ability to feel supported.
Crucially, both protective factors identified in this study are modifiable. Trait emotional intelligence has been shown to improve with training in earlier studies, and perceived social support shifts when relationships shift. Even an adult whose childhood loaded the dice can change the trajectory through cultivated emotional skill and meaningful connection.
The study, “Contributors to co-occurring chronic loneliness and depression and anxiety and the role of adverse childhood experiences,” was authored by Sasha Rudenstine, Jasmyne Pierre, Talia Schulder, and Pedro Goldberg.
