In a dynamic, unpredictable, stimulus-rich environment that includes harmful and dangerous elements, the development and learning of defense strategies are crucial for increasing an individual’s probability of survival and fitness. The individual’s aim may therefore become self-preservation, to limit potential damage through strategies that allow for better environmental understanding and, accordingly, the prediction and prevention of possible dangers (exploration), but also through basic, habitual survival strategies (exploitation) (Mobbs et al., 2024). Consistent with this goal, potentially negative and harmful events are known to capture attention more readily and to be weighted more heavily than positive events during evaluative, decisional, and subsequent action-implementation processes (Norris, 2021; see also Lazarus, 2021 for a different evolutionary perspective). This asymmetry—often summarized by the general principle “bad is stronger than good”—is commonly referred to as the negativity bias (Baumeister et al., 2001).

Despite considerable individual differences in its behavioral expression (Norris et al., 2011), most individuals display this bias across multiple contexts, in response to broad categories of stimuli (e.g., images, sounds, words; Norris, 2021), and over the entire circle of lifespan from infancy (Chae and Song, 2018; Xie et al., 2019) to older adulthood (Norris, 2021), although some evidence suggests the emergence of a positivity effect in late adulthood (Carstensen and DeLiema, 2018). Neuroscientific evidence points to a neural basis for the negativity bias, reflected in heightened responses to negative compared with positive stimuli (Huang and Luo, 2006; Yoshida et al., 2021).

In particular, manifestations of the negativity bias can be observed in two prominent phenomena in cognitive neuroscience such as loneliness (Bellucci et al., 2025) and loss aversion (Norris, 2021). Although typically conceptualized and investigated as distinct constructs within separate research traditions, in this perspective paper we advance a novel conceptual link between them, arguing that they may involve partially shared neuro-cognitive mechanisms.

Loss aversion is a well-established construct in behavioral economics and psychology referring to the greater psychological impact of prospective losses relative to equivalent gains in decision-making under risk (Kahneman and Tversky, 2013; Kahneman et al., 1991; see Figure 1 top left). Originally described in the context of economic decision-making (Kahneman and Tversky, 2013), it is typically investigated through controlled experimental paradigms involving mixed-gambles (e.g., Arioli et al., 2023) in which participants decide whether to accept or reject gambles offering an equal 50% probability of either winning or losing varying monetary amounts. Identifying the indifference point in this mixed-gamble task allows to estimate a “loss aversion” coefficient, commonly denoted as lambda (λ; Molins and Serrano, 2019; Sokol-Hessner and Rutledge, 2019), that captures the extent to which losses are overweighted relative to gains during valuation. Robust empirical evidence estimates λ at around 2 (1.8–2.1 in Brown et al.’s, 2024 meta-analysis; but see also Yechiam and Zeif, 2025 and Walasek et al., 2024), suggesting that losses exert roughly twice the subjective impact of equivalent gains. From an evolutionary perspective, loss aversion may have evolved as an adaptive self-protective mechanism (Li et al., 2012), because the costs of events such as injury or resource depletion, would have outweighed the benefits of comparable gains.

Extending beyond behavioral economics, loss aversion appears to shape decision-making across multiple domains, including non-monetary contexts, confirming its role in broader valuation processes (Rick, 2011). Consistent with the view that it reflects a deep-seated tendency, individual differences in loss aversion have been associated with multiple neuro-physiological markers, including metrics of arousal (Sokol-Hessner et al., 2009; Sheng et al., 2020), interoception (Sokol-Hessner et al., 2015), as well as brain activity (Canessa et al., 2013, 2017; Tom et al., 2007), connectivity (Arioli et al., 2025c) and structure (Arioli et al., 2025a).

Loneliness refers to the distressing feeling that one’s social relationships fall short of what one desires (Perlman and Peplau, 1981; for a recent review see Bellucci et al., 2025), reflecting in the subjective sense of social isolation and disconnection (see Figure 1 top right). Loneliness is typically evaluated with variants of the UCLA (University of California, Los Angeles) Loneliness Scale (Russell et al., 1978), a self-report questionnaire including items assessing perceived deficits in close and supportive relationships, such as the perceived lack of trusted confidant(s) and of mutual aid, protection, and acceptance. Neuropsychological research has typically examined this condition and its effects by contrasting individuals with high vs. low loneliness (Cacioppo et al., 2015), and/or by testing associations between loneliness and specific behavioral or neural responses (Arioli et al., 2025b). While most individuals do not report feeling lonely at a given moment, studies on older adults in high-income countries suggest that approximately 1 in 4 experience some degree of loneliness at least occasionally (Chawla et al., 2021; see also Cacioppo and Cacioppo, 2018 for an overview). Previous research has shown that the experience of loneliness is significantly shaped both by genetic predispositions (Abdellaoui et al., 2019) and environmental influences (e.g., supportive co-housing that does not undermine independence; Bower et al., 2023).

Since social interactions can enhance survival and reproductive success but also entail risks such as exposure to violence, the motivating effect of loneliness towards social connection is expected to reflect in the implicit hyper-vigilance to negative social cues, that may then be coded as possible social threats signaling rejection or isolation (Spithoven et al., 2017). The recognition of such side-effects of social connection shaped the view that loneliness evolved not merely as a biological drive to reconnect, but as an adaptive mechanism that triggers implicit vigilance and the avoidance of social threats to prevent social damage (Cacioppo and Cacioppo, 2018). This tendency is often grounded in the aforementioned evolutionary cost–benefit logic, i.e., that the costs of failing to detect a potential threat outweigh those of missing a social opportunity (Cacioppo et al., 2014). Accordingly, loneliness may automatically trigger heightened vigilance for - and avoidance of—negative social events, alongside more self-focused and protective behavioral tendencies (Cacioppo and Cacioppo, 2018), which can paradoxically hamper social reconnection (Cacioppo and Hawkley, 2009; Holt-Lunstad et al., 2015; Bzdok and Dunbar, 2020). Converging neuroscientific evidence provides complementary support, showing that the experience of loneliness is associated with distinctive activation patterns in brain networks associated with attentional and socio-cognitive processing (Lam et al., 2021).

Any possible link between loss aversion and loneliness depends on the presence of shared underlying mechanisms. We therefore discuss the cognitive processes involved in both phenomena, their neural correlates (see Figure 1 bottom), and, finally, how their extreme manifestation relates to functional impairments commonly associated with depression.

Before turning to the broader implications of this novel perspective, we clarify that, despite our hypothesis of shared underlying neural mechanism, loss aversion and loneliness are distinct phenomena. This distinction is highlighted by several key differences.

First, although both phenomena have been associated with psychopathological conditions (including depression; Erzen and Çikrikci, 2018; Sediyama et al., 2020) and are negatively associated with well-being (Binte Mohammad Adib and Sabharwal, 2024; Koan et al., 2021), they differ in their typical expression. Loss aversion is a decision-making tendency observed in most individuals, and does not inherently entail clinical distress or economic dissatisfaction. Rather than representing a general tendency, loneliness is a condition experienced by individuals who perceive their social relationships as unsatisfactory or even defective. Therefore, while both phenomena may lead to “missed opportunities” (foregone gains in the case of loss aversion, and reduced access to social connection in the case of loneliness), only the latter is inherently associated with an explicit desire for (social) change. Thus, while loss aversion could be largely explained by a negativity bias within decision-making, loneliness represents a far more complex phenomenon that cannot be reduced only to a manifestation of this bias. Indeed, loneliness is primarily characterized by explicit social dissatisfaction, coupled with a cognitive profile of social negativity bias (e.g., hyper-vigilance for social threats) that paradoxically promotes withdrawal and isolation, thereby hindering social reconnection, and supporting the consolidation of social dissatisfaction (Bellucci et al., 2025).

Second, although both concepts involve a preferential processing of negative information, this tendency can be easily quantified in the economic context of loss aversion (e.g., λ ≈ 2), whereas it is more difficult to operationalize in the social domain of loneliness. The objective difference between outcomes such as −5 and +4 is constant in monetary decision-making, allowing to estimate the extent to which an individual’s subjective weighting deviates from the objective “expected value” (Knutson et al., 2005). In contrast, in social contexts it is far more difficult to assign precise objective values of negativity or positivity to events, and/or to quantify how subjective evaluations deviate from such a benchmark in lonelier individuals. This challenge is further complicated by the fact that hyper-vigilance toward negative stimuli in loneliness is often implicit, such that the explicit judgments may fail to capture the underlying implicit evaluations.

Reflecting this difference in formal tractability, loss aversion is typically operationalized as a well-defined asymmetry in the subjective weighting of losses and gains (negative and positive outcomes, respectively). By contrast, the hyper-vigilance for social threats in loneliness is assessed through more heterogeneous methods, including comparisons across social negative and social positive stimuli (following the negativity bias framework, e.g., Bangee et al., 2014; Shin and Kim, 2019), or between social negative stimuli and neutral (e.g., Arioli et al., 2025b) or non-social negative stimuli (e.g., Cacioppo et al., 2015), as well as group comparisons assessing the processing of negative social stimuli in lonely vs. non-lonely individuals (e.g., Du et al., 2022).

Finally, adopting an ecological perspective entails investigating behavior in contexts that allow to capture the entire process - from stimulus processing and evaluation to decision itself and its subsequent outcome - typically involving actual consequences. In the case of loss aversion, this requirement is relatively straightforward to meet, as most studies provide performance-based monetary incentives. Conversely, in the domain of loneliness, simulating a social context with genuine interpersonal exchanges and real consequences is extremely challenging, resulting both in experimental constraints and ethical concerns.

These distinctions do not limit the present perspective, and rather represent critical considerations for experimental design and result interpretations in future research.

Loss aversion and loneliness are well-established constructs in neuroscientific and psychological research. Traditionally, they have been studied as distinct phenomena grounded in separate domains, i.e., decision-making and social interaction, respectively. In this perspective paper, we challenge this separation and propose a novel unifying framework in which these constructs are partially supported by shared neuro-cognitive mechanisms.

We first reviewed loss aversion and loneliness separately, emphasizing the negativity bias as a common driver of both. We then detailed convergent cognitive and neural mechanisms, showing that both are associated with heightened attentional sensitivity to negative stimuli and/or experiences (be they financial losses or social threats), anticipation of adverse outcomes, and difficulties in emotion regulation. These cognitive and affective biases are reflected in specific neural signatures, involving structures associated with the processing of negative emotions, prospective rewards and punishments, as well as emotion regulation via cognitive control. From a clinical standpoint, high levels of loneliness and loss aversion also share convergent associations with depression, suggesting an overlapping profile at their extreme levels.

Following this overview, we discussed the key differences between the two constructs, particularly regarding how they are operationalized in experimental psychology and cognitive neuroscience. While loss aversion is a general tendency in decision-making that can be quantified using mixed-gamble paradigms, loneliness is a subjective experience of social disconnection, which makes valid—and yet ecological—assessment more complex and methodologically heterogeneous. Crucially, these distinctions are essential for evaluating the proposed analogy between the two constructs: rather than discouraging the search for their shared mechanisms, they help define conceptual boundaries and refine experimental approaches. In turn, this analysis enables a clearer conceptualization of both the common and distinctive facets across these aspects of human behavior likely shaped by the negativity bias. Accordingly, we should move beyond a purely “negative” lens focused on methodological obstacles, and rather emphasize the substantial gains of investigating their common basis, and the extent to which they outweigh the methodological challenges inherent in such interdisciplinary research.

Importantly, the present paper offers a theoretical perspective aimed at stimulating empirical research into the potential link between loss aversion and loneliness. Future studies in this field may employ both cross-sectional and longitudinal experimental paradigms to assess the common and distinctive factors shaping these phenomena, and their susceptibility to experiential, personological and/or demographic factors, as well as to experimental manipulations. A comprehensive empirical investigation of the shared mechanisms underlying the constructs of loss aversion and loneliness could not only provide novel insights into the clinical manifestations of depression, but also deepen our understanding of human behavior and of the neuro-cognitive mechanisms shaped by negativity bias and self-preservation across decision-making and social interaction.