Genes

Genes influence the way individuals respond to stress. The heritability of PTSD, a psychiatric disorder exemplifying vulnerability to trauma, is known to be approximately 30–40% (Logue Reference Logue, Amstadter and Baker2015). Additionally, studies have emphasised the importance of the stressor itself, both its nature and its severity, in dictating an individual's response to stress. More recently, twin studies have examined the heritability of resilience. Incorporating the number and severity of experienced stressors into statistical models yields a measure of ‘relative resilience’ for each individual – i.e. higher or lower resilience than predicted on the basis of characteristics of the sample as a whole. Using this approach, findings from a large twin study suggest that resilience is equally influenced by genes and environment (Amstadter Reference Amstadter, Myers and Kendler2014). Other twin studies have begun to examine heritability of trait resilience and of psychological characteristics known to be associated with resilience, for example prosocial attitudes.

Variation in a range of stress-response system genes, such as the alpha-2 adrenoreceptor or neuropeptide Y genes, has been implicated in differential responses to stress. For example, slower return to baseline of noradrenaline levels once a stressor subsides (Neumeister Reference Neumeister, Charney and Belfer2005) or lower production of neuropeptide Y in response to threat-related stimuli (Zhou Reference Zhou, Zhu and Hariri2008) result respectively in differential vulnerability or resilience to stressful events. In a key series of studies, variation in haplotypes of the FKBP5 gene, coding for a chaperone protein that modulates hypothalamic–pituitary–adrenal (HPA) axis responsivity to stress, was shown to confer resilience or vulnerability to the development of trauma-related psychopathology in childhood (Watkins Reference Watkins, Han and Harpaz-Rotem2016). Paralleling the history of genetic research in psychiatric disorders, initial genetic studies of resilience focused on candidate genes. Examples include studies of protective alleles involved in modulating noradrenaline, amygdala and hippocampal responses to stress or threat (Zhou Reference Zhou, Zhu and Hariri2008) and others searching for protective alleles in more resilient survivors of childhood trauma (Cicchetti Reference Cicchetti, Rogosch and Sturge-Apple2007).

Beyond confirming the importance of gene × environment interactions, studies suggest that some gene variants influence sensitivity to the surrounding environment, with broad implications for preventive interventions especially during childhood and adolescence (Belsky Reference Belsky2016). More recently, genome-wide association studies (GWAS) of resilience have emerged, facilitated by a growing number of publicly available data-sets and large research collaborations (Stein Reference Stein, Choi and Jain2019).

Epigenetics

Epigenetic studies also illuminate the impact of environmental factors on resilience. Certain environmental exposures throughout the lifespan, such as maternal care, social support and drug use, have been found to alter chromatin structure, thus affecting gene expression and an individual's susceptibility to trauma (Dudley Reference Dudley, Li and Kobor2011). In studies of adults, differences in methylation patterns of candidate genes, for example the serotonin transporter gene (SLC6A4) promoter, have been linked with differential psychological outcomes related to stressors of various degrees (Gottschalk Reference Gottschalk, Domschke and Schiele2020), as well as differential limbic system activity in response to sad and fearful stimuli (Ismaylova Reference Ismaylova, Lévesque and Pomares2018).

Epigenetic changes during developmental years can have permanent effects on the brain. During prenatal and early postnatal life, the developing brain is particularly susceptible to environmental exposures via epigenetic modifications in the hippocampus resulting in long-lasting effects on stress response (Miguel Reference Miguel, Pereira and Silveira2019). During adolescence, exposures to excessive alcohol are associated with epigenetic changes in the frontal cortex, striatum and nucleus accumbens (Pascual Reference Pascual, Boix and Felipo2009). Although much remains to be learned, studies have begun to elucidate the complex interactions between genotype, environment and epigenetics and their influence on resilient phenotypes.

Stress-response systems

Resilience is heavily dependent on the efficient activation and termination of the stress response (Feder Reference Feder, Fred-Torres and Southwick2019), which is mediated by the HPA axis and the locus ceruleus–noradrenaline system and early life experiences. For example, increased stress in childhood (e.g. abuse or neglect) can lead to persistently elevated levels of corticotropin-releasing hormone (CRH), associated with decreased resilience persisting into adulthood (van Bodegom Reference van Bodegom, Homberg and Henckens2017). Chronic stress has also been linked to changes in glutamate-system function in adulthood, associated with synaptic atrophy in the hippocampus and prefrontal cortex (Deyama Reference Deyama and Duman2020). Of note, the hippocampus contains high levels of glucocorticoid receptors, which mediate negative HPA axis feedback, facilitating appropriate and efficient responses to stress. Additionally, the protective role of the dopaminergic system during responses to stress has been investigated, revealing that activation of dopaminergic neurons may facilitate resilience by dampening fear responses. The function of neuropeptide Y signalling is also potentially protective against stress and trauma. Higher levels of this anxiolytic neuropeptide may balance the actions of CRH in the brain, regulating fear responses (Sabban Reference Sabban, Alaluf and Serova2016). A range of other neuropeptide and neurotransmitter systems affect stress responses, including brain-derived neurotrophic factor (BDNF), endocannabinoid, oxytocin and other systems. Preclinical studies have identified novel molecular adaptations in more resilient animals, with translational potential to resilience in humans (Tan Reference Tan, Costi and Morris2020). Taken together, these biological responses to stress will open new avenues for interventions to enhance resilience.

The immune system

The immune system is another mediator of resilience. After a stressful event, glucocorticoids are released and inhibit the production of pro-inflammatory cytokines such as interleukin 6 (IL-6) (Gądek-Michalska Reference Gądek-Michalska, Tadeusz and Rachwalska2013). In animal models, lower circulating levels of IL-6 prior to a stressful event have predicted subsequent resilience to stress (Gądek-Michalska Reference Gądek-Michalska, Tadeusz and Rachwalska2013). In human studies, elevated systemic levels of IL-6 have been identified in individuals with treatment-resistant depression and PTSD (Hodes Reference Hodes, Pfau and Leboeuf2014), whereas dispositional positive affect, a characteristic of resilient individuals, has been associated with lower IL-6 levels (Stellar Reference Stellar, John-Henderson and Anderson2015). Studies have even linked positive affect with protection against developing infections such as the common cold and influenza after participants were deliberately infected with these viruses via nasal spray (Cohen Reference Cohen, Alper and Doyle2006). Although these studies do not purport causality, they shed light on the potentially protective role of positive emotions and their impact on the immune system.

Elevated levels of C-reactive protein (CRP), a marker of peripheral inflammation, are also associated with the development of PTSD (Eraly Reference Eraly, Nievergelt and Maihofer2014). Conversely, lower plasma levels of IL-6 and CRP, comparable to levels in trauma-unexposed controls, have been linked to recovery from PTSD (Gill Reference Gill, Saligan and Lee2013). Tumour necrosis factor alpha (TNF-α), another cytokine produced mainly by macrophages during acute inflammation, has also shown relevance. TNF-α has been associated with depression and TNF-α antagonists have shown promise for the treatment of depressive disorders, although there remains a lack of clarity on their adverse impact on the overall immune system (Brymer Reference Brymer, Romay-Tallon and Allen2019).

Together, these studies highlight the complex and reciprocal interactions between the immune system and stress-response systems in mediating resilience and offer potential opportunities for novel treatment targets.

Neural circuitry

Several studies aim to understand the neural underpinnings of resilience. Functional neuroimaging studies have examined neural circuitry involved in fear, emotion regulation, reward responses and cognitive flexibility, comprising interconnected regions of the amygdala, anterior cingulate cortex, hippocampus, ventral striatum and prefrontal cortex. Such research suggests that resilience is associated with lower activation of threat appraisal regions such as the amygdala, more efficient functioning of prefrontal areas subserving implicit emotion regulation and higher activation of dorsolateral prefrontal regions involved in cognitive control (Scult Reference Scult, Knodt and Swartz2017; Chen Reference Chen, Ke and Qi2018). Additionally, resilience seems to involve higher neural adaptability and more efficient use of neural–emotional resources for adaptive coping (Feder Reference Feder, Fred-Torres and Southwick2019). Of particular relevance is neuroplasticity, the brain's ability to respond to stimuli by reorganising its structure and connections. The discovery that connections between key brain regions can be strengthened or weakened following personal experiences helps contextualise resilience as dynamic over the lifespan.

Heterogeneity among resilience interventions

To date, ‘resiliency training programmes’ are a loosely defined set of interventions aimed to enhance resilience that lack standardisation in format, method of delivery, intervention and control groups, outcome measures and even the constructs used to define and measure resilience (Leppin Reference Leppin, Bora and Tilburt2014). A narrative review of 44 randomised control trials of resilience interventions for adults found that, despite methodological shortcomings, interventions showed small to moderate efficacy at enhancing resilience and mental health at 3-month follow-up (Leppin Reference Leppin, Bora and Tilburt2014; Linz Reference Linz, Helmreich and Kunzler2020).

Among the studies analysed, interventions took place in groups, individual settings or a combination of formats. They occurred in-person, online, by phone or were multi-modal. Participants ranged widely and included teachers, students, soldiers, police officers, people with chronic physical or mental illnesses, traumatised populations and random samples from the general population. Interventions were compared with treatments, no treatment or waiting-list control groups. Sessions varied in length and frequency from a single 40 min session to several 120 min sessions. Programme content ranged across different therapeutic approaches (adaptation training, mindfulness, stress management, cognitive–behavioural techniques, etc.), and some programmes focused on the explicit teaching of phenotypic resilience features such as emotion regulation, optimism or self-efficacy.

Although this heterogeneity has garnered criticism for potentially diluting the field of resilience intervention research (Leppin Reference Leppin, Bora and Tilburt2014), it is important to acknowledge that different aspects of resilience may be differentially valued by different populations; there may be no single ‘one-size-fits-all’ intervention for building resilience.

Child and adolescent interventions

Individuals and groups at different developmental stages warrant distinct age-appropriate and context-dependent interventions. For example, young children exposed to adversity are particularly susceptible to developing mental and physical disorders as a result of the vast neural growth and plasticity during this period (Shonkoff Reference Shonkoff and Garner2012). This plasticity also makes them ideal candidates for focused interventions. Early effective intervention programmes may be caregiver and family based, school based or community based. Long-term school-based programmes for children in Bhutan, Mexico and Peru have shown an increase in well-being and academic performance after students underwent a 15-month intervention that focused on, among other skills, the quality of interpersonal relationships, empathy and altruism, mindfulness, effective communication and emotion management (Adler Reference Adler2016).

Adult interventions

Interventions for adults typically draw on advanced processes of cognition and emotion processing. They range from pre-trauma training and preventive approaches to early post-trauma and long-term programmes.

Prevention

Preventive programmes for mental health may target specific skills, such as cognitive reframing and support-seeking, hardiness training, or relaxation and mindfulness guidance (Horn Reference Horn and Feder2018). Perhaps the most widely used examples are those employed by the US army: master resilience training (MRT) and pre-deployment stress inoculation training (PRIEST). These courses teach soldiers stress control, coping skills, support-seeking and cognitive reappraisal (Griffith Reference Griffith and West2013; Hourani Reference Hourani, Tueller and Kizakevich2016). Soldiers who have completed MRT have reported enhanced coping and fewer symptoms of poor behavioural health during subsequent stressful times (Griffith Reference Griffith and West2013). Additionally, there is evidence that soldiers without baseline mental health problems who complete PRIEST are protected against developing symptoms of PTSD (Hourani Reference Hourani, Tueller and Kizakevich2016). Given the inevitability of exposure to traumatic situations among soldiers, the case for preventive resilience interventions is obvious. With increasing global stress, preventive programmes for general populations might be of great utility to minimise adverse mental outcomes.

Early post-trauma and short-term interventions

Early post-trauma and short-term resilience interventions focus on exposure therapy and the development of effective coping mechanisms. In one study, individuals who suffered from DSM-IV criterion A trauma who engaged in a modified exposure therapy programme in the emergency department hours after experiencing the trauma reported lower symptoms of PTSD and depression 3 months later compared with the control group (Rothbaum Reference Rothbaum, Kearns and Price2012). In a similar study, survivors of sexual assault who were shown a short video teaching coping strategies within 72 h of their assault also reported lower PTSD and depressive symptoms up to 6 months later (Resnick Reference Resnick, Acierno and Waldrop2007). Although these results are promising, further research is warranted to characterise the strengths and pitfalls of these early post-trauma interventions.

Long-term interventions

Long-term resilience training programmes vary greatly, but most aim to promote the development of lasting psychosocial, behavioural and cognitive tools (e.g. social support, optimism, physical exercise, emotion regulation and cognitive reappraisal). One intervention for US combat veterans with PTSD and their partners uses a couples-based treatment called strategic approach therapy (SAT). This therapy targets avoidance symptoms of PTSD by encouraging effective communication, intimacy and anxiety reduction (Sautter Reference Sautter, Glynn and Thompson2009).

Pharmacological interventions

Pharmacological interventions may be employed in addition to psychosocial approaches or as stand-alone treatments to alter fear responses, emotion regulation, memory and attention and to enhance prosocial behaviours (Kelmendi Reference Kelmendi, Adams and Yarnell2016). For example, several drugs have been proposed to increase resilience by regulating the sympathetic nervous system (e.g. neuropeptide Y) and HPA axis (e.g. CRH antagonists, dehydroepiandrosterone), improving neurogenesis, preventing neuronal damage (e.g. antidepressants) and controlling memory formation for traumatic events (e.g. beta blockers). The combination of medication and psychotherapy has been shown to be superior to either type of treatment alone (Kamenov Reference Kamenov, Twomey and Cabello2017).

Summary

We have highlighted the great deal of heterogeneity among resilience training programmes. Given the discussion of various cultural conceptions of resilience, this diversity is necessary but also limits our ability to make valid comparisons across interventions. Importantly, the goal for many of these programmes is not just to mitigate the risks of developing negative psychopathology, but also to teach skills that promote mental health and sustainable well-being. Although many distinct interventions aim to accomplish this, the common thread among them is an emphasis on behavioural change and deliberate practice. Akin to working a muscle, the cultivation and development of resilience takes time, practice and, often, support from others in order to see positive changes in phenotype and underlying physiology.

What else affects resilience?

As discussed, individual responses to stress and adversity are mediated by neurobiological, psychosocial and cultural factors. Although many will recover or endure a brief symptomatic period, some may suffer lasting psychological consequences. The idea that resilience is context dependent means acknowledging the profound impact of past traumatic experiences, pre-existing psychological and medical conditions, and access to resources. In other words, exhibiting and building resilience is easier for some than for others. This notion is particularly salient when examining the diversity of responses to the COVID-19 pandemic.

Studies repeatedly show that the frequency and intensity of stress matter (Hobfoll Reference Hobfoll, Mancini and Hall2011) and those who have experienced childhood trauma or chronic stress are less likely to manifest resilience (Alim Reference Alim, Feder and Graves2008). Individuals with pre-existing psychiatric conditions such as major depressive disorder will also need to work harder to adapt, owing to overwhelming feelings of hopelessness or anhedonia that may prevent their active practice of resilience. In a similar vein, chronic medical conditions or severe injuries may make it more difficult for individuals to employ cognitive and physical coping strategies. The severity of COVID-19-related illness can similarly affect survivors’ resilience.

Finally, an individual's resilience is strongly influenced by external support and available resources. Consider for a moment the effects of COVID-19 for an individual with a stable income, housing and established support systems compared with an uninsured individual who lost their job and lives in a small and crowded multi-generational apartment. Clearly, varying levels of wealth, job security, social support and socioeconomic status make it easier for some to overcome the adversity caused by COVID-19. Early data have shown that individuals and communities most affected are those with the least access to healthcare, job security, community services and other basic resources (Ahmed Reference Ahmed, Ahmed and Pissarides2020). In the USA, communities of colour, including Black, Latinx and Native American populations, are disproportionally affected owing to the aforementioned factors as well as systemic racism within the healthcare system and society (Fortuna Reference Fortuna, Tolou-Shams and Robles-Ramamurthy2020). Resources must be allocated to ensure that these suffering communities are given the greatest chance to survive and thrive. COVID-19 presents an opportunity to create actionable changes that lead to the establishment of a stable ground for all to stand upon.

Another mediating factor that predicts how an individual or community will recover from stress is the stressor itself. The world now faces a universal stressor: COVID-19. Although the pandemic has disproportionate impact depending on racial or ethnic minority group status, geography and personal proximity to the virus, the universal nature of this tragedy may allow for tighter control of a crucial variable (the stressor) in calculating the complex equation that defines resilience (Fig. 2). This presents a unique opportunity for several applications of future research. For example, researchers across the world might choose to observe global responses to further validate personal/individual, social and cultural determinants of resilience. It also poses an opportunity to better understand and eventually predict the trajectories of individuals and groups with distinct neurobiological, developmental, psychosocial and cultural profiles. This research might involve international surveys, intervention programmes, coordinated qualitative interviews, or other forms. Despite variation, this work should strive for a broad, dynamic and global approach.

FIG 2 An integrated model of resilience. Neurobiology and developmental profiles interact with contributions from context and culture as well as progress made from evidence-based interventions. At the core of resilience is the interaction between these three systems and the stressor itself. In this article, we use the COVID-19 pandemic as an example of a universal stressor that may have an impact on diverse populations and cultures around the world in unique and shared ways. Importantly, this relationship moves in both directions as the resilience developed and manifested by individuals can also feed back to further influence their neurobiology, cultural context and practised interventions.

Recommendations and interventions

As the pandemic progresses, we can draw on findings presented in this review to support the development, scaling and implementation of resilience interventions. Examples are summarised in Box 2.

Final thoughts

Although the tendency to engage in resilient behaviours may be easier for some than others, practising such strategies over time can eventually modify neurobiology to reinforce resilient behaviour via neuroplasticity. Importantly, practices should be tailored to specific populations and cultures. In one recent study looking at the psychological impact of COVID-19 on the general public in China, the authors suggest several psychological interventions (Wang Reference Wang, Pan and Wan2020), including globally shared modalities such as cognitive–behavioural therapy (CBT), along with traditional Chinese medicine such as self-administered acupressure and emotional freedom techniques. In another article, researchers from the USA present multiple coping strategies based in both Western and Eastern traditions, including behavioural activation, acceptance-based coping, mindfulness practices and loving-kindness meditation to decrease stress and promote resilience (Polizzi Reference Polizzi, Lynn and Perry2020). As more research emerges, coordinated global efforts are needed to track the effectiveness of these interventions across diverse populations.

Since avoidance is a common response to trauma and there is a high degree of stigma in many communities associated with seeking help for psychological problems, interventions should be made even more public and widely accessible. For example, in settings such as hospitals, where exposure to the pandemic is inevitable, interventions might be offered to employees with a default opt-out rather than opt-in option, to change norms. Additionally, in communities hit hardest, screening for trauma should be augmented (e.g. by primary care physicians, community healthcare workers, teachers, employers and others) with opportunities to connect those who screen positive to resources. We have before us an opportunity to destigmatise and normalise the pursuit of mental healthcare on a global scale. Doing so will require cooperation and resources from local and national governments, healthcare systems, educational institutions, employers and others to ultimately advance the mental health and resilience of our global population.