A nurse keeps showing up through a difficult month. Colleagues call her resilient. She is also sleeping badly, covering vacant shifts and losing time with her family. The praise describes visible persistence; it cannot tell us whether her health is protected, whether the workplace is sustainable or whether she has an acceptable way to stop. A better account of resilience asks how people and systems adapt under real demands and what the adaptation costs.
Resilience is often sold as toughness: an inner quality that lets a person absorb adversity without changing. Research points to a more complex subject. Some people maintain relatively stable functioning after a difficult event. Others struggle and recover. Some difficulties emerge later. Responses change with exposure, available support, prior circumstances and the outcome being measured. “Adaptive capacity” is a useful umbrella for this changing relation between demand and resources, provided it does not erase the distinct meanings researchers give to resilience.
A trajectory is not a character verdict
Bonanno and colleagues’ review of loss and potential trauma emphasizes heterogeneous trajectories rather than a universal path through distress. A relatively stable course after adversity can be described as resilience in this outcome-oriented tradition; improvement after a period of substantial disruption is often called recovery. Those patterns are observations over time, not moral rankings. Neither a person who struggles nor a person who recovers slowly has failed a test of character.
Even trajectory labels depend on what investigators measure and when. Low measured symptoms during one year do not prove an absence of grief, fear or pain. A person may function at work while experiencing private distress. Another may show disrupted sleep after an event but remain connected to valued relationships. The pattern of one outcome is not the whole life. Researchers must state whether they measured depression, posttraumatic symptoms, daily functioning, well-being or some combination.
Timing matters as well. A brief survey soon after a crisis may capture initial shock. A later survey may capture adaptation, exhaustion or a second stressor. A trajectory inferred from repeated observations is more informative than one questionnaire, but it is still a model fitted to selected measures. The number of apparent groups can change with sampling, follow-up intervals and statistical choices. It is risky to turn a model’s category into a fixed identity for an individual.
Capacity is relational
An individual can have useful skills, including problem solving, emotional flexibility and knowledge of when to seek help. Yet capacity is also created or constrained outside the person. Reliable income, safe housing, accessible care, trusted relationships, fair work conditions and a predictable route to information can alter how much demand a person must carry. Resilience is therefore not a private reserve that one simply discovers in a crisis.
Schäfer and colleagues’ preregistered systematic review examined longitudinal studies of responses to societal challenges and crises in high-income countries. Across fifty included studies, higher income and socioeconomic status, perceived social support, emotion regulation and psychological flexibility were associated with more resilient trajectories. The review also found mixed results for many factors, smaller associations after adjustment, frequent convenience sampling and relatively little study of societal-level resources. This supports a multilevel question while warning against easy causal prescriptions.
Consider two people recovering from the same disruption. One has paid leave, a nearby friend and a stable home. The other has an unstable schedule, no transport to care and a landlord threatening eviction. Both may use coping skills. Their measured outcomes cannot be understood fairly by comparing only their determination. Context does not make individual action irrelevant, but it changes what action is possible and what it costs.
In a school, the same principle applies. A student who asks for help and adjusts a study plan may be using adaptive capacity. If the school offers no accessible explanation, no quiet space and no way to catch up after illness, “be resilient” can become a label for institutional inaction. The program should improve the conditions in which the student’s effort occurs.
Adaptation can include changing course
Toughness language suggests that the successful response is to continue doing the same thing under greater strain. Adaptive capacity may instead involve resting, delegating, withdrawing from a harmful role, asking for support, revising a goal or protecting others by changing a system. A pause can be a functional response. So can leaving an unsafe situation. Persistence is one possible strategy, not the definition of success.
This makes “flexibility” more precise than unbroken endurance, but even flexibility needs limits. A person may face demands that no adjustment can make reasonable. If a job requires permanent overtime, the problem is not insufficient flexibility. If a community experiences repeated displacement, individual coping programs cannot replace secure housing and structural protection.
A practical assessment asks what is within the person’s control, what support is actually available and which changes require institutional action. It also asks whether the adaptation preserves health and valued commitments over time. Short-term continuation can be expensive if it consumes sleep, relationships or opportunities for recovery. Visible functioning may coexist with accumulating burden.
Brain health broadens the frame
Udeh-Momoh and colleagues’ 2025 review treats resilience in brain health as a multidimensional interaction of biological, psychological, social and environmental factors. It argues that research concentrated in the global north cannot simply be exported to majority-world settings with different exposures, resources and cultural understandings. The authors propose integrating exposures across the life course, cultural reserve, community resources, allostasis and whole-body health. Much of this is a research agenda, not proof that one integrated program has already solved brain-health inequality.
The review is a reminder that brain health does not sit outside social life. A biological marker may relate to adaptation while being shaped by sleep, nutrition, environment or access to care. A study that measures only individual coping may overlook the conditions producing the demand. A study that reports an average across countries may conceal meaningful differences in who was recruited and what services were available.
At the same time, a multilevel framework is not permission to infer every causal connection. Exposures, biomarkers, social support and clinical outcomes can be correlated in complex ways. A convincing claim about an intervention needs a design that separates plausible alternatives. The appeal of a broad resilience story should not outrun the evidence for a specific mechanism.
Support is more than a contact count
Social support can be experienced as knowing someone will help, receiving practical assistance or belonging to a reliable group. These are not identical variables. A person with many contacts may lack someone safe to call. A small network can be strong. In the Schäfer review, perceived support was among the factors associated with resilient trajectories, but the authors also emphasize limitations in sampling and adjustment. The association does not demonstrate that adding contacts by itself causes recovery.
Support has to be usable. A hotline that is impossible to reach, a formal care pathway with a long wait, or a manager who says “my door is open” while punishing disclosure can all look like available resources on paper. Evaluating resilience support means checking access, trust, quality and the burden placed on the person seeking it.
Communities also maintain capacity. Local organizations can share information, identify needs and coordinate practical help. Public services can provide transport, food, safe spaces and continuity of care. These are social mechanisms that may make adaptation more possible. They should be assessed directly rather than imagined as a collective personality trait called community toughness.
Measurement changes the story
Some scales ask whether a person believes they can bounce back. Others assess symptoms after an adverse event. Still others combine resources and outcomes into a single score. Those approaches answer different questions. A high self-rating before exposure is not the same as observed functioning during exposure. A low symptom score after exposure may reflect low adversity, strong support, underreporting or a genuinely favorable response.
Good resilience research specifies the stressor. Adapting to bereavement is not the same task as adapting to repeated economic insecurity. It records prior functioning when possible, follows people over time and states what counts as a meaningful outcome. It also examines attrition: those under greatest strain may be least able to complete follow-up surveys. A reassuring curve built from remaining respondents can be misleading.
The same caution applies in practice. A workplace survey may find that many staff “cope well” if those who left are excluded. A school may report strong attendance after a crisis while missing pupils who transferred. A city may celebrate recovery in aggregate although one neighborhood remains without services. Resilience should be examined with distribution and opportunity in view, not only a favorable mean.
There is also a timescale problem. A community may maintain service for several weeks by relying on volunteers, then suffer burnout. A person may meet deadlines during an emergency and only later feel the accumulated effects. If evaluation ends when immediate output resumes, it may mistake a bridge for a sustainable road. Follow-up should track replenishment, not just continued function.
What interventions can responsibly aim to do
An intervention can help people recognize available supports, practice a specific coping skill, plan a manageable next step or ask for assistance. Such programs should describe the outcome they are designed to improve and test it against a credible comparison. They should not promise immunity to distress. Nor should they turn adversity into a required route to growth.
A 2024 umbrella review of resilience-promoting interventions and protective factors gathered a broad literature after adversity. Its breadth is useful for mapping candidates, but umbrella reviews inherit the quality, overlap and heterogeneity of the underlying reviews. A finding that a class of programs is promising does not specify the right program for a particular trauma, population or service system. The outcome, comparison and follow-up still matter.
For a hospital, a resilience program might pair optional individual support with safe staffing, protected breaks and a confidential route to report overload. The institution should measure both worker well-being and demand conditions. If it counts attendance at a coping workshop while ignoring vacancies, it is not testing the full mechanism it invokes. For a school, targeted help after disruption should sit alongside accessible teaching and stable relationships.
Participation should be chosen where possible. A person facing grief or trauma may not want a group discussion or a story about personal growth. Respect includes room to decline an exercise and still receive needed support. In a clinical setting, symptom assessment and treatment decisions belong with qualified professionals, not with a generalized library page.
Programs should also record who cannot participate. A workshop at midday may be inaccessible to shift workers and caregivers. A digital exercise can exclude people without a private device, a stable connection or a comfortable language option. If the participants who benefit are those with the most resources to attend, the program may be useful for them while leaving the central inequality intact. Evaluation should include reach, not just the outcomes of those who completed every session.
There is a further ethical question about demands after a program ends. Suppose a team learns useful techniques for recovering from stressful calls. If management then increases call volume because the team appears better able to cope, a measured benefit could be converted into a new burden. Resilience support must be paired with a commitment not to spend the recovered capacity without consent or fair conditions. This is an institutional design principle inferred from the multilevel evidence, not an effect directly tested by the cited reviews.
Finally, interventions should preserve multiple routes to adjustment. Some people prefer practical help, others a trusted conversation, and others a period of quiet with an option to return later. A narrow program can misclassify a person as disengaged when its method simply does not fit. The question is whether support becomes more usable and whether the exposure itself can be reduced, rather than whether everyone conforms to one preferred coping style.
What the evidence does not show
Research does not show that resilient people never experience distress, that recovery follows one timetable or that remaining outwardly productive proves health. It does not show that adversity is beneficial by default. A trajectory with relatively low symptoms is a pattern in selected outcomes, not a certificate of invulnerability.
The systematic review of crisis trajectories does not show that income, social support or psychological flexibility can simply be assigned to individuals by telling them to be more resilient. Associations do not remove confounding, and many included studies used non-representative samples. The global brain-health review maps important research directions, but it does not validate one universal resilience biomarker or one intervention for all cultures and contexts.
Neither does a social explanation make personal skills meaningless. It means skills should be judged in relation to opportunities and demands. The Institute synthesis here is that adaptive capacity includes the ability to use support and alter conditions, while toughness rhetoric can hide cost. This is an interpretive frame from several evidence streams, not a newly discovered law of psychology.
Return to the nurse
The nurse may have acted with judgment and care through a demanding month. Calling that resilience might be accurate in one limited sense. The next question is whether her schedule, pay, relationships and health permit a sustainable future. If the institution can change the conditions, praising endurance alone is insufficient.
Resilience as adaptive capacity means more than continuing to stand. It means having the resources, relationships and room to respond, recover and sometimes change the situation itself. The best resilience story leaves space for both human agency and the responsibilities of the systems around it.
Research references
- Bonanno, G. A., Westphal, M., & Mancini, A. D. (2011). Resilience to loss and potential trauma. Annual Review of Clinical Psychology, 7, 511–535. View source
- Schäfer, S. K., Supke, M., Kausmann, C., et al. (2024). A systematic review of individual, social, and societal resilience factors in response to societal challenges and crises. Communications Psychology, 2, 92. View source
- Udeh-Momoh, C. T., Migeot, J., Blackmon, K., et al. (2025). Resilience and brain health in global populations. Nature Medicine, 31, 2518–2531. View source
- Bonanno, G. A., & Diminich, E. D. (2013). Annual research review: Positive adjustment to adversity, trajectories of minimal-impact resilience and emergent resilience. Journal of Child Psychology and Psychiatry, 54, 378–401. View source
- Southwick, S. M., Bonanno, G. A., Masten, A. S., Panter-Brick, C., & Yehuda, R. (2014). Resilience definitions, theory, and challenges: Interdisciplinary perspectives. European Journal of Psychotraumatology, 5, 25338. View source
- Abate, B. B., Sendekie, A. K., Tadesse, A. W., et al. (2024). Resilience after adversity: An umbrella review of adversity protective factors and resilience-promoting interventions. Frontiers in Psychiatry, 15, 1391312. View source