Loneliness in schizophrenia: Just loneliness

Abstract

Loneliness is reported by the majority of people with schizophrenia (∼80%), yet the cognitive and neural mechanisms underlying loneliness in this population are incompletely understood. Similar to the general population, loneliness in schizophrenia is associated with cognitive biases toward mistrusting others, as well as structural barriers to social activity. In addition, some evidence suggests that predisposing genetic and environmental factors, as well as some of the neural changes and the poor cardiometabolic health linked with the illness, are associated with loneliness in schizophrenia. However, much evidence supports a transdiagnostic model of the causes and consequences of loneliness, with similar factors playing a role in loneliness in schizophrenia and the general population. Currently there are no validated treatments targeting loneliness associated with schizophrenia, although interventions focusing on cognitive bias modification, positive psychology, mindfulness, or interpersonal synchrony have shown promise. Taken together, the current state of the field suggests that future research on schizophrenia should include measurements of social isolation and loneliness. Overall, there is a need for novel intervention approaches that can interrupt the cycle of mutually reinforcing neurocognitive biases and environmental conditions linked with loneliness in schizophrenia and other populations.

Loneliness may represent both a contributing cause and a consequence of schizophrenia. Studies have identified links between loneliness in schizophrenia and mistrust‐related cognitive biases, the functioning of brain areas involved in social perception, and poor physical health, which are similar to observations in non‐clinical, general population samples. Thus, interventions that can reduce loneliness may have wide‐ranging benefits for both mental and cardiometabolic health in individuals with schizophrenia and other populations.

INTRODUCTION

Loneliness, or perceived social isolation, is an experience of distress that reflects a discrepancy between one's actual and desired levels of social connection. ¹ , ² , ³ It is a universally experienced subjective state that is somewhat benign or even adaptive when it is relatively brief, but it becomes harmful when it is long‐lasting. Although some amount of loneliness is experienced by a segment of the general population throughout the world, ⁴ chronic states of loneliness are particularly common in people who have been diagnosed with mental illnesses such as depression ⁵ , ⁶ and schizophrenia. ⁷ , ⁸ , ⁹ , ¹⁰ Although the precise relationships between these conditions and loneliness remain incompletely understood, the reciprocal relationship between loneliness and depression has been well established ⁵ , ¹¹ , ¹² and is consistent with the generally aversive and debilitating nature of persistent loneliness. Related to this, the high rates (∼80%) of loneliness reported by people with a diagnosis of schizophrenia ⁷ , ¹³ may be linked, to some extent, to the high rates of depression in the disorder. ¹⁴ , ¹⁵ , ¹⁶ Yet few clinicians inquire about loneliness in this population, ¹⁷ and there are no available treatments that have proven efficacy in reducing loneliness in people with serious mental illnesses such as schizophrenia. This gap in knowledge and clinical practice has likely arisen from a lack of commonly held knowledge regarding the psychological, neurobiological, and environmental underpinnings of loneliness in schizophrenia. Thus, this narrative review describes existing models of loneliness in schizophrenia and compares those to the overall literature on loneliness, in an effort to contribute to closing these gaps. One central question considered in this review is whether or not loneliness in schizophrenia has unique features that distinguish it from loneliness experienced by people without this condition. Alternatively, are many of the drivers of loneliness in this population universal?

CONSENSUS MODEL OF LONELINESS AND THE PUBLIC HEALTH PROBLEM

Loneliness can be defined as the subjective experience of distress that arises from a discrepancy between (1) the expected or desired amount or quality of social connection or bonds and (2) the perception of the actual amount or quality of those connections or bonds. ¹ , ² , ³ It is important to note that loneliness is sometimes, but not always, associated with objectively lower than expected levels of social contact (i.e., social isolation). In other words, loneliness and social isolation are typically correlated, but not strongly, on a population level; ¹⁸ , ¹⁹ , ²⁰ , ²¹ thus, one can be lonely without being isolated and isolated without being lonely.

Loneliness is thought to be an adaptive response in the short term, in that it can trigger behaviors that lead to a correction of a perceived deficit in social bonds, restoring the protective social connections that are known, from an evolutionary perspective, to promote the likelihood of survival. ²² , ²³ , ²⁴ Typically, the loneliness signal is brief in duration, disappearing once the perceived deficit in social bonds has been corrected. However, if this deficit is not corrected, the loneliness signal persists over time, which then can become harmful to one's overall mental and physical health due to the multiple systemic, physiological sequelae (e.g., the release of stress‐related hormones and immunological mediators associated with this persistent lonely state). ¹² , ²⁵ Exactly how this occurs is poorly understood. However, according to this “social homeostasis” hypothesis, ²⁶ loneliness triggers a fight or flight stress response within the body, with an accompanying cascade of hormonal and immunological responses, including the release of cortisol and inflammatory cytokines. ²⁷ , ²⁸ , ²⁹ Thus, chronic loneliness may involve experiencing a persistently elevated state of stress.

It has been well established that rates of this type of harmful, chronic loneliness have increased across the world during the past several decades. ³⁰ , ³¹ In response, what has been called a “loneliness epidemic” has become a major public health concern in many countries, in part because of the robust associations observed between isolation and loneliness and poor mental and physical well‐being. ³² , ³³ , ³⁴ , ³⁵ Specifically, large epidemiological studies have shown that social isolation and loneliness are linked to a wide range of poor health outcomes, including elevated rates of cardiovascular disease and stroke and earlier mortality, ³⁶ , ³⁷ as well as numerous psychiatric conditions. ¹² , ³⁵ It has been proposed that poor health behaviors and elevations in levels of stress and systemic inflammatory responses that accompany isolation or loneliness are responsible for these associations. ²⁸ , ²⁹ , ³⁸ , ³⁹ , ⁴⁰ However, thus far, the sequence of physiological events that may link the changes in stress and inflammatory responses associated with isolation and/or loneliness and specific health outcomes, such as cardiovascular disease and shorter lifespans, are not well understood.

LONELINESS IN SCHIZOPHRENIA

Schizophrenia is a disorder that is associated with particularly high rates of loneliness and social isolation. ⁷ , ¹⁰ , ¹³ Unfortunately, there are many misconceptions about this psychiatric condition, which affects ∼0.5%–1.0% of the world's population. ⁴¹ , ⁴² Schizophrenia is often initially recognized and diagnosed following the emergence of psychotic symptoms, also referred to as the positive symptoms of schizophrenia (e.g., delusions and hallucinations, which are beliefs and perceptions, respectively, that are not based in reality and can co‐occur with disorganized thinking or behavior). Although experiencing some of these positive symptoms (for at least 1 month) is one criterion for receiving the diagnosis of schizophrenia, the typically more disabling aspects of the illness are its negative symptoms, which include anhedonia (impairments in anticipating, seeking, or experiencing pleasurable experiences, including those involving social activities and interactions) and related deficits in motivation, which are often associated with some degree of diminished physical and social activity. ⁴³ Schizophrenia is also variably associated with mild‐to‐moderate deficits in cognition (e.g., in working memory, episodic memory, attention, and processing speed), as well as in certain domains of social cognition (e.g., facial affect recognition, theory‐of‐mind processing, and social attributions ⁴⁴ , ⁴⁵ ).

Potential associations between loneliness and the symptoms of schizophrenia

Given that many of the symptoms of schizophrenia are social in nature—paranoid delusions typically involve incorrect assumptions about the malevolent intent of others, and negative symptoms often manifest as impairments in social drive and behavior—one important, largely unanswered question about loneliness in schizophrenia is whether or not it is closely linked to these core symptoms of the illness.

In addition, a related variable that must be accounted for in this research is the social isolation commonly experienced by people with this condition. ⁴⁶ , ⁴⁷ Social isolation for people with schizophrenia is often closely linked to structural barriers that reduce opportunities for social contact or connection, such as being unemployed, financial hardship, or the stigma and social exclusion associated with having the illness. The degree to which loneliness in people with schizophrenia represents a direct consequence of the widespread social isolation associated with the illness is unclear.

But how do social isolation, loneliness, and the negative symptoms of schizophrenia interact within individuals? In the past, deficits in affective and verbal expression observed in people with schizophrenia (referred to as affective flattening and alogia by clinicians and classified as negative symptoms) led to the assumption that people with this illness do not experience distress when socially isolated and thus do not experience loneliness. However, this assumption has not been supported by surveys finding that over 75% of people with schizophrenia describe feeling lonely, ⁷ , ¹³ and many cite reducing loneliness as a top priority for their treatment. ⁷ These findings are consistent with a body of research showing that deficits in emotional expression in schizophrenia are not accompanied by deficits in internal emotional experience. ⁴⁸ , ⁴⁹ In other words, people with schizophrenia have emotional responses (feelings) that are similar to those without the illness, including feelings of loneliness when their social needs or expectations are not met. Thus negative symptoms and feelings of loneliness can coexist and, in fact, have been found to be correlated in severity in some studies. ⁵⁰ , ⁵¹ , ⁵² Negative symptoms such as social withdrawal may exacerbate social isolation and loneliness, or vice versa; thus, treatments aimed at reducing each of these symptom types may be broadly beneficial for these related experiences.

Another question about the relationship between the symptoms of schizophrenia and loneliness is whether some of the cognitive biases known to be common in lonely people in the general population, such as a sensitivity to rejection or “social threat”, ¹² which can perpetuate states of isolation and loneliness, are more common in people with schizophrenia, as they may overlap mechanistically with some of the most common symptoms of the disorder, such as paranoia. Supporting this possibility, associations between loneliness and paranoia have been detected consistently. ⁵² Related correlations among social anxiety, paranoia, and loneliness have also been reported. ⁵³ , ⁵⁴

Another question is whether the cognitive impairment associated with schizophrenia contributes to loneliness. A few studies have found preliminary evidence for associations between measures of cognition and loneliness or social isolation, ¹³ , ⁵⁵ consistent with findings in older adults; ²⁹ , ⁵⁶ , ⁵⁷ however, this question requires further investigation.

Taken together, it is important to note that, although it is clear that both the positive and negative symptoms of schizophrenia are linked to loneliness, the direction of these associations remains poorly understood.

Loneliness and cardiometabolic health in schizophrenia

The robust associations found in the general population between social isolation and/or loneliness and poor cardiometabolic health ³⁶ , ³⁷ may be relevant for understanding the frequent co‐occurrence of loneliness and poor health outcomes in people with schizophrenia. It has been well established that having schizophrenia is associated with elevated rates of cardiovascular disease, diabetes, and obesity and a diminished lifespan (by two decades on average), ⁵⁸ , ⁵⁹ in comparison to rates in the general population. Although numerous hypotheses about these associations have been proposed, ⁶⁰ their underlying biological mechanisms remain unknown. An important, testable prediction arises from these findings: In light of the strong associations found between loneliness and poor cardiometabolic health in the general population, and the high rates of loneliness in schizophrenia, it is possible that loneliness contributes to the poor cardiometabolic health associated with schizophrenia. Supporting this possibility, one study found a significant correlation between rates of loneliness and of cardiometabolic disease in people with schizophrenia. ⁶¹ The poor physical health outcomes associated with schizophrenia are likely due to a range of factors, ⁶² including certain social determinants of health, such as the low socioeconomic status and reduced access to healthcare associated with having schizophrenia, and some of the common symptoms of the illness, including poor insight regarding one's health needs and cognitive impairment, as well as secondary effects of treatment (e.g., the metabolic side effects of antipsychotic medications). ⁶³ , ⁶⁴ However, the experience of loneliness may also contribute directly to the medical morbidity and earlier mortality associated with schizophrenia.

There is also evidence that some of the risk for poor physical health associated with schizophrenia may be intrinsic to (and pre‐date the onset of) the illness; shared genetic risk factors for schizophrenia and cardiometabolic diseases, such as diabetes, have been identified. ⁶⁵ , ⁶⁶ Intriguingly, similar relationships have been found for the genetics of isolation/loneliness and schizophrenia. One study found a significant association between a polygenic risk score for loneliness and social isolation and genetic risk for schizophrenia, ⁶⁷ and another study found correlations between a polygenic risk score for social isolation behavior and genetic risk for several psychiatric disorders, most strongly with autism spectrum disorder, followed by schizophrenia and then depression. ⁶⁸ Thus, these studies suggest that loneliness and poor physical health in schizophrenia may each have a genetic component. The effects of these predispositions may then be exacerbated by interactions between them and the secondary effects of the illness over the course of the lifespan. Loneliness may negatively impact health, and symptoms and the real‐world consequences of having schizophrenia may negatively impact both loneliness and health. However, to fully understand the complex antecedents of these outcomes in people at risk for or experiencing schizophrenia, these models must be tested prospectively in longitudinal studies.

In summary, the various potential causes of loneliness and isolation in schizophrenia and in the general population are likely to be complex and heterogeneous, with psychological, environmental, and perhaps genetic components (Figure 1). To inform future examinations of these candidate contributions to loneliness in schizophrenia, some of what is known about these components is reviewed below, with unresolved questions discussed in turn.

FIGURE 1.

Contributors and correlates of loneliness in schizophrenia. Potential contributors to loneliness (purple box) are listed in white boxes and correlates or consequences in red boxes (pink indicates that the direction of effects is likely bidirectional). See Lim et al., ¹⁰ for a related model. HPA, hypothalamic–pituitary–adrenal; SI/L, social isolation/loneliness.

UNRESOLVED QUESTIONS ABOUT LONELINESS IN SCHIZOPHRENIA

Is loneliness a cause or a consequence of schizophrenia?

One long‐standing model of schizophrenia, originally proposed by Ralph Hoffman in the 1970s, suggested that the earliest (pre‐illness) symptoms of schizophrenia include, most prominently, social withdrawal, accompanied by social isolation and loneliness, and that these states of social deprivation lead to “social deafferentation” ⁶⁹ —the deficit in social input is followed by compensatory changes in the brain that are analogous to changes that can occur following loss of a limb (e.g., phantom limb syndrome). According to this model, this socially deafferented state can be followed by a compensatory emergence of psychotic symptoms with social content (e.g., hearing human voices conversing, or experiencing paranoid or self‐referential ideas that focus on the relationship between the self and others). These symptoms with social content serve to fill in the deficit in social contact (the social isolation) experienced by the at‐risk individual. Subsequently, a reciprocally‐reinforcing maladaptive loop is thought to emerge in some individuals, in which the psychotic symptoms lead to further withdrawal from others, which then leads to more social deafferentation and greater vulnerability to psychotic symptoms. For example, low‐level paranoid thoughts involving mistrust of others can become increasingly inflexible and resistant to revision over time if the beliefs lead to avoidance of others and thus are not challenged, and ultimately corrected, during interactions with other people that fail to confirm the paranoid ideas.

A related hypothesis is the “social defeat” model of schizophrenia, which proposes that early experiences of social adversity, including childhood abuse, social rejection, or exclusion (including bullying, discrimination, and social exclusion related to one's race/ethnicity, immigration, or socioeconomic status), increase the risk for the development of psychotic ⁷⁰ , ⁷¹ and negative ⁷² , ⁷³ , ⁷⁴ symptoms via the development of negative (defeatist) beliefs about the self. This model is also consistent with the recent evidence that greater levels of social fragmentation within neighborhoods (calculated based on factors such as rates of single‐person households and residential mobility that reflect levels of social cohesion or support) are associated with an increased incidence and prevalence of psychotic disorders and with an earlier onset and greater severity of symptoms, even after controlling for socioeconomic deprivation. ⁷⁵ , ⁷⁶ , ⁷⁷ , ⁷⁸

These models are in line with the overall hypothesis that certain predispositions, and early experiences and conditions that foster social disconnection and reinforce negative beliefs about the self or others, may increase a person's vulnerability to developing the symptoms of schizophrenia or exacerbate existing symptoms. Consistent with this overall model, studies have found evidence that experiences of loneliness mediate the well‐established association between having a history of childhood trauma and psychosis‐related outcomes (including a diagnosis of a psychotic disorder, ⁷⁹ or severity of psychotic symptoms ⁸⁰ or subclinical psychotic experiences ⁸¹ ). In addition, similar links between early social adversity and characteristics related to negative symptoms, such as low motivation and a tendency toward social withdrawal, may be mediated by loneliness and social defeat‐related cognitive biases (e.g., low self‐esteem, negative self‐schemas), ⁵¹ as well.

These are appealing models of the early evolution and maintenance of the symptoms of schizophrenia and, specifically, the relationship between isolation/loneliness and these symptoms. In addition, the recent evidence mentioned above, that genetic variants linked to social isolation and loneliness overlap with those linked to a risk for schizophrenia, ⁶⁷ , ⁶⁸ supports the possibility that a partially genetically determined tendency toward social withdrawal (possibly related to behavioral traits such as introversion or behavioral inhibition) may contribute to the development of schizophrenia. However, the empirical evidence supporting these types of models—which suggest that some degree of pre‐existing social isolation, and the stress or loneliness associated with it, may play a causal role in the etiology of schizophrenia—is mixed. Significant cross‐sectional correlations between social disconnection (social isolation or loneliness) and psychotic symptoms (both subclinical and clinical) have been consistently observed. ⁵² , ⁸² , ⁸³ , ⁸⁴ , ⁸⁵ , ⁸⁶ , ⁸⁷ , ⁸⁸ , ⁸⁹ However, the direction of this overall association remains unclear; some longitudinal studies have found evidence that loneliness precedes the onset of subclinical psychotic symptoms, including paranoia, ⁵⁴ , ⁹⁰ , ⁹¹ , ⁹² whereas other studies have found evidence for the opposite direction of effects ⁹³ , ⁹⁴ or bidirectional effects. ⁹⁵

Supporting a causal role of loneliness in the later emergence of paranoia, several studies using experimental designs in non‐clinical populations found that increases or decreases in loneliness were followed by subsequent increases or decreases in paranoia, respectively. ⁹⁶ , ⁹⁷ However, one limitation of this overall line of research is that the majority of these studies were conducted in general population samples focusing on individuals with varying degrees of subclinical psychotic symptoms, often referred to as psychotic experiences or psychotic‐like experiences. These types of mild, subclinical, psychotic‐like symptoms have only a modest association with true clinical psychosis, as the majority (>80%) of people experiencing these phenomena do not develop schizophrenia or any psychotic disorder (although their risk for these conditions and other mental illnesses is 5‐ to 10‐fold higher than the average level of risk in the general population). ⁹⁸ , ⁹⁹ , ¹⁰⁰ Therefore, it is necessary to conduct additional longitudinal studies of these social deafferentation or social defeat models in large samples or in individuals who are at a very high risk for developing psychotic disorders.

It is notable that the predicted close relationship between loneliness and the most common psychotic symptom, paranoia, is conceptually related to the general psychological model of loneliness applied to the general population, which emphasizes the role of mistrust, social threat, and related maladaptive beliefs about the intentions of others. ²² In addition, one behavioral manifestation of the mistrust associated with loneliness, an increase in the physical distance that a person prefers from another (a need for greater interpersonal distance or personal space), has been linked to loneliness ¹⁰¹ , ¹⁰² and has been consistently observed in people with schizophrenia. ¹⁰³ , ¹⁰⁴ , ¹⁰⁵ , ¹⁰⁶ , ¹⁰⁷ Thus, taken together, mistrust of others, in all of its manifestations (psychological and behavioral), appears to be linked to loneliness in a dimensional fashion across many populations. In other words, it is possible that paranoia, an extreme expression of mistrust of others, and possibly other symptoms of schizophrenia, operate in the maintenance of loneliness in a way that is similar to other types of common maladaptive biases about the intentions of others frequently observed in the general population.

How do the deficits in social cognition observed in people with schizophrenia contribute to loneliness?

Another hypothesis about loneliness in schizophrenia is that the deficits in social cognition commonly observed in many people with the illness contribute to loneliness (and/or that social isolation and loneliness contribute to these deficits in social cognition). Given that deficits in social cognition have been consistently linked to impairments in social functioning in people with schizophrenia, ⁴⁵ , ¹⁰⁸ , ¹⁰⁹ it is reasonable to ask whether such deficits are related to social isolation and/or loneliness as well. However, thus far, the evidence for this proposed relationship has been mixed, with the majority of studies finding little evidence for a relationship between social cognitive impairment—as measured by performance on validated social cognition tasks—and loneliness in schizophrenia. ¹¹⁰ , ¹¹¹ , ¹¹² , ¹¹³ However, one recent study did find, in a subgroup of individuals with schizophrenia with prominent social cognition deficits, a relationship between impairments in social cognition and loneliness. ¹¹⁴

It is important to highlight that the findings of these studies appear to depend somewhat on how social cognition, a heterogeneous category of mechanistically distinct processes, is defined. The types of social cognition that have been linked to loneliness in schizophrenia are those that reflect maladaptive beliefs about the self or other, such as guilt, low self‐esteem, ¹¹⁰ biases toward perceiving social threat, ¹¹¹ or self‐reported aspects of social cognition that rely on self‐awareness. ¹¹² These types of associations are in line with the pattern of findings in general population samples of associations between loneliness and poor self‐esteem, sensitivity to social threat, and heightened anticipation of rejection or exclusion. ¹² , ²² The similarities among the findings of studies conducted in schizophrenia and general population samples suggest that there are some common, fundamental processes occurring in diverse populations that perpetuate loneliness.

What is known about the neurobiology of loneliness in schizophrenia?

Although there is a large literature based on both animal models and humans that has examined the overall neurobiology of social isolation and loneliness using a variety of techniques, little is currently known about the specific brain circuits associated with social isolation and loneliness in people with schizophrenia. However, it is likely that the findings of neuroimaging studies of loneliness conducted in healthy samples are relevant to this question, particularly given that many of the brain networks that have been implicated in the mechanisms of loneliness are thought to be altered in schizophrenia. Neuroimaging studies of loneliness (which have been reviewed extensively elsewhere ²⁹ , ¹¹⁵ , ¹¹⁶ , ¹¹⁷ ) have reported associations between loneliness and the structure or functioning of brain circuitry involved in recalling or perceiving social information, such as regions of the default network and medial and lateral temporal lobe areas, and brain circuitry involved in the detection of rewarding or salient information in the environment, such as the striatum, anterior insula, and other components of the ventral attention and salience networks. ²⁹ , ¹¹⁸ In addition, studies conducted in animal models of isolation have suggested that the hypothalamus and areas of the midbrain (including the ventral tegmental area [VTA]) play a central role in triggering a range of responses of the brain and body to social isolation, including the attentional vigilance and physical arousal that typically accompany social isolation. ²⁶ , ¹¹⁹ The findings of these studies are consistent with evidence that early social adversity (e.g., neglect, abuse, social rejection, and exclusion) may lead to sensitization of the dopamine‐containing areas of the midbrain (e.g., the VTA) ¹²⁰ and these areas may then become over‐responsive during states of social isolation.

Regions involved in social perception and autobiographical memory, such as the amygdala, hippocampus, and regions of the default network, appear to be particularly involved in responses of the brain to social isolation and loneliness, or in their secondary effects on the brain. For example, lower volume, ¹²¹ , ¹²² gray matter density, ¹²³ responsiveness, ¹²⁴ and connectivity ¹²⁵ of the amygdala have been linked to social isolation. Consistent with the well‐known role that the amygdala plays in social behavior, ¹²⁶ its sensitivity to levels of social activity in animal models, ²⁹ and the evidence that the amygdala is altered in disorders characterized by impairments in social functioning, including schizophrenia and autism, ¹²⁷ , ¹²⁸ , ¹²⁹ , ¹³⁰ a range of changes in the amygdala likely occurs when humans are more socially isolated and experiencing lower levels of social stimulation. ¹³¹ , ¹³² , ¹³³

The default network (which includes the medial prefrontal cortex, posterior cingulate gyrus, and lateral parietal and temporal cortical areas) is also involved in social perceptual processes, such as mentalizing, social attributions, and self and other representations ¹³⁴ and appears to play a role in the experience of loneliness. For example, recent studies have found that, in comparison to non‐lonely people, lonely individuals show less similarity in responses of default network regions (specifically the medial prefrontal cortex) to representations of the self versus others ¹³⁵ and less similar patterns of activation of the default network between themselves and other people. ¹³⁶ In addition, one large study (N = ∼40,000) found that several default network regions were larger and exhibited increased within‐network functional connectivity in lonely compared to non‐lonely individuals. ¹³⁷

One region that is often included in the default network (given its strong connections with many of the neocortical areas of the default network ¹³⁴ ), the hippocampus, has been consistently implicated in prior neuroimaging studies of loneliness. Several studies have found evidence for negative correlations between loneliness and the volume of the hippocampus and other regions with close connections to the hippocampus. ¹³⁸ , ¹³⁹ , ¹⁴⁰ , ¹⁴¹ Loneliness has also been associated with altered functional connectivity and task‐elicited responses of the hippocampus. ¹⁴¹ , ¹⁴² , ¹⁴³ , ¹⁴⁴

These associations may be related to a tendency of lonely individuals to engage in internally focused cognitive processes that rely on the default network and hippocampus, such as rumination, reviewing the past (autobiographical memory retrieval), mental simulations, ¹⁴⁵ and mapping social relationships. ¹⁴⁶ , ¹⁴⁷ These processes could in part represent compensatory mechanisms that attempt to reproduce experiences of social intimacy. ¹⁴⁸ Given that loneliness has been linked to increases in levels of cortisol and peripheral physiological measures of stress in humans, ¹² a reduction in hippocampal volume in lonely individuals may also reflect changes in hippocampal morphology akin to the effects of chronic stress on the brain that have been previously detected in studies of stress conducted in humans ¹⁴⁹ and animal models. ¹⁵⁰ , ¹⁵¹

How are these neuroimaging studies of loneliness in healthy, general population samples relevant to understanding the neurobiology of loneliness in schizophrenia? Although the precise answer to this question is unknown, clues can be inferred from some of the overlap of the loneliness and schizophrenia neuroimaging literatures. Although neuroimaging studies of schizophrenia have identified abnormalities in numerous regions of the brain, findings of changes in the structure and function of the hippocampus are among some of the most consistent findings. ¹⁵² , ¹⁵³ , ¹⁵⁴ For example, an early meta‐analysis of structural magnetic resonance imaging (MRI) studies conducted in individuals with schizophrenia found that reduced volumes of medial temporal lobe structures, including the hippocampus, were commonly reported in the illness (observed in 74% of the 193 studies reviewed). ¹⁵⁵ Consistent with this finding, a recent analysis conducted by the Enhancing Neuroimaging Genetics through Meta‐Analysis (ENIGMA) consortium in more than 2000 individuals with schizophrenia and a similar number of control subjects found reductions in the volume of a range of brain regions in the schizophrenia group versus the control group, with the largest effect size (d = −0.46) in the hippocampus. ¹⁵⁶ Studies have generally shown that the hippocampus is smaller and less responsive during cognitive tasks (e.g., episodic memory retrieval), shows weaker habituation in response to emotionally salient stimuli, and exhibits elevated resting activity in individuals with schizophrenia in comparison to control subjects. ¹⁵⁴ , ¹⁵⁷ , ¹⁵⁸ However, given that loneliness has also been associated with changes in the hippocampus, and loneliness is common in schizophrenia, it remains unclear what role loneliness may play in the abnormalities in the hippocampus and other brain regions that have been observed in schizophrenia.

Although no well‐powered studies have addressed this question to date, one recent study investigated the neural correlates of loneliness in people with psychotic illnesses, including schizophrenia and affective psychoses, using task‐based, functional MRI (fMRI) and found evidence that loneliness in the psychotic disorder sample (N = 40) was linked to over‐responsivity of the hippocampus to dynamic social stimuli (moving images of human faces), even after controlling for social isolation, depression, and social anhedonia. ¹⁵⁹ Similar but weaker correlations were observed in the healthy control group in this study, and, in a subsample of the full cohort, baseline hippocampal responses predicted increases in loneliness 1 year later. Another recent study measured the functional connectivity of the default network in individuals with schizophrenia and Alzheimer's disease and in two healthy control groups (a younger and older control group, age‐matched to the schizophrenia and Alzheimer's disease group, respectively). This study found that, across the full sample, default network connectivity was negatively correlated with loneliness. ¹⁴⁴ Taken together, these two neuroimaging studies of loneliness in individuals with psychotic disorders found similar relationships between loneliness and variation in brain function (hippocampal responses and default network connectivity, respectively) in people with and without a diagnosis of a psychotic disorder. These findings are consistent with the hypothesis that the neural correlates of loneliness are similar across different populations and show a transdiagnostic, dimensional relationship with loneliness.

Given that the hippocampus appears to play a central role in loneliness and is also often found to be affected in schizophrenia, one outstanding question is, which functions of the hippocampus might play a role in loneliness in people with schizophrenia or in general? The role of the hippocampus in the experience of loneliness is currently unknown; however, clues have emerged from what is known about its central role in episodic memory processes. The hippocampus is believed to organize information about prior experiences via multidimensional cognitive maps ¹⁶⁰ that rely on predictive coding mechanisms. ¹⁶¹ , ¹⁶² These include maps of social distances (in affiliation and power) between the self and others. ¹⁴⁶ , ¹⁴⁷ These findings suggest that the hippocampus could play a role in generating prediction error signals that represent discrepancies between expected (based on stored social maps) and actual (perceived) levels of affiliation/social connection (see Figure 2 for a schematic illustration of this hypothesis). The subjective and bodily experiences of distress associated with such a discrepancy may be generated by regions that receive hippocampal inputs, such as the amygdala and hypothalamus, and the brainstem targets of the descending projections of these regions. Awareness of this distress may be mediated by activity in frontal cortical areas. In light of the evidence that prediction error signaling is altered in schizophrenia, possibly contributing to delusional beliefs such as paranoia, ¹⁶³ it is possible that abnormal prediction error signaling also plays a role in generating loneliness‐related signals in schizophrenia and other populations.

FIGURE 2.

A schematic of the social homeostasis model and factors that may contribute to expected and actual (perceived) levels of social connection in people with schizophrenia and other populations. The social homeostasis model ²⁶ proposes that discrepancies between expected versus perceived levels of social connection are detected by the brain, which subsequently generates responses aiming to reduce such discrepancies. The hippocampus, other closely connected regions of the default network, and brain regions known to be involved in prediction error signaling (i.e., in identifying these discrepancies), such as the midbrain, striatum, and prefrontal cortex, may be involved in these comparator processes. Detected discrepancies (expected > actual levels of social connection) are thought to trigger a cascade of behavioral and systemic physiological responses and the subjective feeling of loneliness. Successful and unsuccessful corrections may each lead to additional responses of the brain (e.g., reward signals and persistent stress responses, respectively). Various factors that may influence the brain's calculation (i.e., the perception) of the actual and expected levels of social connection are listed within the orange circles. Actual levels of social connection can reflect the quantitative amount of social contact or activity, as well as the quality of those relationships and the degree of intimacy or attunement with others; cognitive biases that lead to greater mistrust and assumptions about the negative intentions of others can reduce the perceived level of social connection (independent of the amount of social activity or number of social contacts). The expected level of social connection may reflect genetic and related personality or temperament factors (e.g., introversion, behavioral inhibition), as well as positive and negative early and/or current experiences within the social environment, such as experiences of childhood social adversity (e.g., abuse, bullying, exclusion), discrimination, or social fragmentation within one's local environment.

Another important consideration is the well‐known link between the hippocampus and stress. The hippocampus and closely connected regions such as the amygdala and prefrontal cortex appear to play a role in the neural responses to stress ¹⁶⁴ , ¹⁶⁵ , ¹⁶⁶ and are also sensitive to secondary, damaging effects of stress (e.g., loss of synapses and volume), ¹⁵⁰ including the stress associated with social isolation. ²⁹ Consistent with this literature, there is evidence that levels of social activity moderate these associations in animal models ¹⁶⁷ , ¹⁶⁸ and humans. For example, one study found that correlations observed between hippocampal volumes and poverty within a psychosis risk sample were absent in the individuals with higher levels of social activity. ¹⁶⁹

Taken together, it is clear that much more additional research is needed to understand the precise role of the hippocampus, the larger default network, and other brain regions in the sequence of neurophysiological events that ultimately gives rise to the subjective and physiological states associated with loneliness in schizophrenia and other populations.

What are the environmental factors contributing to social isolation and loneliness in schizophrenia?

Although it is possible to be lonely and not isolated (perhaps when experiencing a diminished level of intimacy or trust in one's relationships) or isolated and not lonely (e.g., when experiencing some form of peaceful solitude), often these two experiences (loneliness and isolation) coexist. Thus, a substantial contributor to loneliness across all populations is social isolation (objectively low levels of social contact or bonds) and the factors that lead to social isolation. When social isolation is chronic and one's expectations for a certain level of social connection are consistently not met (one's automatic, predefined level of expectation or “set point” for social connection), loneliness can become chronic, because the expectation and perceived reality consistently fail to match. In line with the social homeostasis model, ²⁶ one potential way to resolve this discrepancy is with a downward adjustment of one's set point. This might happen if someone is able to come to terms with and accept their level of isolation and develop a greater ability to tolerate or even enjoy solitude. However, it is unclear whether and how this type of adjustment occurs.

Social isolation is common among people with schizophrenia, ⁴⁶ , ⁴⁷ and as described above, may pre‐date and contribute to the risk for developing schizophrenia. ⁷⁷ This isolation may represent a stronger or more common contributor to loneliness in this population than in the general population. As described above, there is some evidence that conditions fostering social isolation and disconnection (genetic, social adversity, and neighborhood/environmental factors) precede the onset of schizophrenia. However, social isolation may impact the evolution of schizophrenia at multiple points during the illness trajectory. In addition to the potential early effects of social isolation on illness onset, many of the environmentally‐based contributors to the isolation experienced day‐to‐day by people with schizophrenia represent secondary effects of having schizophrenia, such as being unemployed (the majority of people with schizophrenia are not fully employed due to the disability caused by the illness ¹⁷⁰ , ¹⁷¹ , ¹⁷² ), as work is a major source of social connection for many people. A lack of employment and an associated reduction in income can lead to an inability to participate in certain social activities that cost money (e.g., going out to dinner, movies, clubs). Moreover, many people with schizophrenia report experiences of social exclusion or discrimination related to the stigma associated with having a serious mental illness. ⁷ Experiencing this type of discrimination can reinforce beliefs about the negative intentions of others and foster mistrust and social withdrawal, as well as maladaptive beliefs about the self, including internalized stigma/diminished self‐esteem, which have been linked to loneliness in schizophrenia. ¹⁷³ , ¹⁷⁴ , ¹⁷⁵

Thus, people with schizophrenia may be at a greater risk for loneliness due to the potentially synergistic, mutually reinforcing effects of (1) the presence of cognitive biases (toward mistrust), related to symptoms of the illness, such as paranoia, that perpetuate loneliness; and (2) structural barriers to social interactions that people with this illness are particularly vulnerable to.

The causes of social isolation for people with schizophrenia may be distinct from those of other populations, which vary substantially (e.g., the elderly may be isolated due to relationship losses and/or physical impairments; young people may be isolated due to an over‐reliance on online interactions in their relationships). However, the effects of isolation on the neural, cognitive, and physiological mechanisms of each individual's social homeostasis may be similar across these populations.

Are there validated, available treatments for loneliness in schizophrenia?

Given the associated suffering and the multiple, potentially detrimental effects of loneliness on health and well‐being for people with schizophrenia, one might expect that loneliness would be a major focus of treatment of the illness. However, thus far, few interventions have been designed and tested to specifically address loneliness in people with psychotic disorders, and none have been widely implemented. However, a few behavioral interventions aiming to address loneliness and isolation in this population have been piloted and studied. For example, online social interventions, involving chat rooms or peer‐to‐peer interactions, typically moderated by a facilitator, have shown mixed results. Several studies of these programs initially found some preliminary evidence for beneficial effects of these interventions on loneliness and isolation in people with psychotic disorders, ¹⁷⁶ , ¹⁷⁷ but these effects were not confirmed in subsequent, more rigorously controlled effectiveness trials. ¹⁷⁸ Also, one study found that cognitive behavioral therapy for psychosis (CBTp) did not reduce loneliness. ⁹⁴ In contrast, positive psychology interventions, which focus on identifying positive emotions and strengths and expressing gratitude and kindness, have shown promising effects on loneliness in early pilot studies. ¹⁷⁹ , ¹⁸⁰ In addition, a novel approach currently undergoing testing is to use responses to ecological momentary assessments to personalize feedback and recommendations provided to lonely individuals. However, one randomized controlled trial of this type of intervention conducted in people with schizophrenia and related disorders showed that both the group receiving such feedback and the no‐feedback group experienced reductions in loneliness, suggesting that the in‐the‐moment assessments alone may have been beneficial, increasing a sense of social connectedness. ¹⁸¹

Another novel approach being explored for ameliorating loneliness in people with schizophrenia is based on what is known about interpersonal or motor synchrony, which is defined as the alignment in time of the behaviors (e.g., smiling, nodding, bodily gestures, postures, or verbal output) of two or more interacting individuals. ¹⁸² Schizophrenia has been associated with impairments in synchronization of movements during social interactions (interactional synchrony), as well as deficits in basic mimicry and imitation processes. ¹⁸³ Intriguingly, there is evidence suggesting that some of the neural processes mediating synchrony across motor, emotional, and cognitive processes (i.e., similar motor behaviors, feelings, or ideas, respectively) are shared; thus, these distinct types of synchronous processes may reinforce each other. ¹⁸² In other words, greater motor synchrony may promote emotional synchrony and perhaps the experience of social connectedness across people via increasing the activity of neural circuitry supporting shared aspects of these functions. Interventions that involve synchronous actions across a group of people (e.g., group activities involving coordinated movements such as dancing, cooperative games, playing music, or singing) can test this model. ¹⁸⁴ For example, one 8‐week pilot study showed that singing in a choir—which requires precise behavioral coordination across participants and has been associated with enhanced well‐being ¹⁸⁵ —was followed by reductions in loneliness and depression in people with schizophrenia. ¹⁸⁶ The similarities between the social homeostasis model of loneliness ²⁶ and the feedback‐loop framework of interpersonal synchrony ¹⁸² —each includes a comparator function (comparing expected vs. actual social bonds or social alignment) followed by, if necessary, corrective action or by reward signaling when the discrepancy between expected and actual levels is low—suggest that increasing interpersonal synchrony may lead to reductions in loneliness, but this hypothesis remains to be rigorously tested.

Based on the social homeostasis model, novel interventions for loneliness could focus on (1) increasing the actual/perceived level of social connection or (2) reducing the expected level of connection. Given that cognitive biases leading to misattributions of negative intentions to others (or signals suggestive of the presence of social threat/rejection) may interfere with the perception of actual/current connection to others, therapeutic approaches that attempt to modify such biases may be effective. Consistent with this, studies conducted in general population samples testing interventions targeting loneliness have suggested that it is more effective to target maladaptive cognitions that perpetuate loneliness than to focus on improving social skills or increasing opportunities for social contact. ¹⁸⁷

One meta‐analysis of studies of interventions targeting loneliness in general population samples found an overall improvement in the quality of social contact (a reduction in loneliness or increase in social connectedness) across 32 randomized controlled trials. ¹⁸⁸ Psychological interventions, primarily cognitive behavioral approaches that addressed maladaptive attributional biases and avoidance of social situations, showed the greatest benefit, particularly among younger populations. Interventions that focused on increasing social contact had no effect on loneliness and social connectedness, although those programs successfully increased the amount of social contact. These findings highlight the dissociation between the mechanisms underlying these two types of outcomes (isolation vs. loneliness). ¹⁸⁸ There is also evidence that group interventions, with opportunities for interaction and participation, are more effective than individual treatment. Online, technology‐based interventions have shown mixed effects, with non‐technology‐based interventions appearing to be superior, but more data are needed. ¹⁸⁸ , ¹⁸⁹

Alternatively, the expected arm of the homeostasis model of loneliness could be targeted by training lonely individuals to increase their capacity to tolerate or enjoy solitude and participate in a variety of activities that do not depend on relationships or interactions with other people (e.g., being in nature, solitary hobbies, spiritual practices), using mindfulness‐based practices. Consistent with this possibility, some initial studies have found evidence that mindfulness‐based interventions may be effective in reducing loneliness. ¹⁹⁰ , ¹⁹¹ , ¹⁹² , ¹⁹³

CONCLUSIONS: A TRANSDIAGNOSTIC MODEL OF LONELINESS

In summary, the information available to date suggests that the causes of loneliness in schizophrenia are complex and heterogeneous, with internal (psychological) and external (structural) contributors playing roles to different degrees across individuals, with pre‐existing vulnerabilities (genetic, affective/cognitive biases) contributing to an unknown extent. Taken together, the majority of the available data suggest that the causes of loneliness in schizophrenia are not qualitatively different from those in the general population, although the frequency or strength of contributions of certain factors may be higher or lower in people with schizophrenia and related illnesses, in comparison to individuals without these conditions. Recent studies using advanced ecological momentary assessment techniques to measure loneliness in people with schizophrenia have supported this transdiagnostic model, identifying dimensional relationships between loneliness and day‐to‐day experiences that are similar across people with and without serious mental illnesses. ¹⁹⁴ , ¹⁹⁵ One implication of this transdiagnostic model is that the study of loneliness in people with schizophrenia, a well‐defined population with very high rates of loneliness, may offer a useful vehicle for understanding the mechanisms of loneliness overall. Another implication is that treatments that are found to be effective for reducing loneliness in the general population may be adapted, possibly with only minor modifications, for people with schizophrenia. In addition, future studies can also test the alternative possibility that loneliness has unique manifestations and correlates in people with schizophrenia, ¹¹² which can guide the development of specialized interventions for loneliness in this population.

Current data also strongly suggest that social isolation and loneliness represent both contributing causes and consequences of schizophrenia, with both early and ongoing effects during the illness trajectory. Thus, prevention and treatment efforts focused on combating loneliness in this population should take this into consideration—targeting multiple points during the evolution of the illness. Overall, it is clear that loneliness is a major challenge for people with schizophrenia due to the associated harmful effects on health and well‐being. Loneliness may be more common or severe in people with schizophrenia due to the combined effects of (and synergy between) cognitive biases toward mistrusting others that are associated with the most common symptom of schizophrenia, paranoia, and the very real barriers to social connection often faced by people with this illness.

To briefly summarize a few recommendations for future research on this topic, there is a need for studies aiming: (1) to identify the neural pathways involved, and the sequence of neurophysiological and cognitive processes that occur, during the detection of discrepancies between expected versus actual levels of social bonds and the cascade of neural and systemic responses to that discrepancy; (2) to conduct therapeutic trials aiming to reduce loneliness, by testing cognitive bias modifications, positive psychology, and/or mindfulness‐based interventions, as well as those focused on enhancing interpersonal synchrony; (3) to implement such interventions at multiple phases of illness and pre‐illness; (4) to consistently measure isolation and loneliness in future mechanistic and intervention studies of psychotic (and other) illnesses, since these common experiences may impact the findings and interpretations of such studies (e.g., play a role in neural, systemic inflammatory/stress‐related, or cardiometabolic abnormalities); and (5) to employ, when feasible, a transdiagnostic approach in investigating loneliness.

Lastly, the well‐established associations between loneliness and poor cardiometabolic health—two major challenges for people with schizophrenia that require novel approaches to prevent and treat—provide a strong rationale for conducting in‐depth research on this association and testing whether interventions that reduce loneliness lead to improvements in cardiometabolic health in this population and others. Similarly, the associations between loneliness and changes in the structure and function of areas of the brain believed to be central to the pathophysiology of schizophrenia highlight the potential contribution of loneliness to the health of both the brain and the body in people with this illness. Although causal relationships underlying these associations have yet to be identified, their implications for health and well‐being provide a clear justification for further investigation. Research at all levels of inquiry, including genetics, physiology, basic neurobiology, behavioral and neuroimaging studies, clinical trials, epidemiology, and public health, is needed to close the many gaps in our understanding of this important aspect of the human experience.

CONFLICT OF INTEREST STATEMENT

The author declares no conflicts of interest.

Holt, D. J. (2025). Loneliness in schizophrenia: Just loneliness. Ann NY Acad Sci., 1553, 96–111. 10.1111/nyas.70022

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.