London School of Hygiene & Tropical Medicine

Over the past five years, we have examined how mental health changed during COVID-19, why those changes differed so markedly across populations, and what this tells us about how societies respond to crisis. One conclusion is that COVID-19 did not create a global mental health crisis, it actually exposed how fragile our mental health systems and social protections already were, and how quickly inequalities widen when those systems fail.

Across countries, the most severe and persistent mental health challenges did not emerge out of infection or fear of the virus itself. They were driven by the loss of livelihoods, food security, education, safety, social connection and access to care. This distinction positions mental health in crises as a fundamental social and economic challenge, rather than merely a clinical one.

A large body of research on COVID-19 and mental health emerged early in the pandemic, much of it focused on short-term psychological distress. This new review builds on that foundation in several important ways.

First, it examines patterns over time, showing how initial spikes in distress evolved and where they persisted. It then brings clinical service disruption and public policy responses into the same analytical frame, allowing us to see how health systems and social protection interact to shape outcomes. More importantly, it presents key evidence from Africa and other low- and middle-income settings that shapes understanding of how mental health systems function under strain, allowing a more integrated understanding of mental health in crises.

What the evidence shows
The overall review highlights that multiple consistent patterns have since emerged to describe global mental health collapse across regions.

In many countries, symptoms of anxiety and depression rose sharply early in the pandemic and then stabilised. However, these overall trends masked substantial inequalities. Poorer communities, women, children and young people experienced deeper and more persistent distress. Eating disorders increased, particularly among young women. People with pre-existing mental health conditions often lost access to care just as their needs intensified.

Mental health services were widely disrupted. Lockdowns, workforce shortages and the suspension of face-to-face care reduced access across both high-income and low- and middle-income countries. Digital and phone-based services expanded rapidly and provided an essential lifeline for some, but they also excluded many people without reliable internet, private space, stable housing or digital literacy.

Across contexts, mental health outcomes tracked social and economic disruption more closely than viral transmission. Where people lost jobs, income, housing or access to education, mental health deteriorated fastest. Where governments acted to buffer these shocks, populations were better protected.

The central lesson
One of the clearest lessons from this body of evidence is that social protection functions as mental health protection. Countries that implemented policies to protect jobs and incomes, prevent hunger and evictions, and keep children connected to education consistently saw better mental health outcomes at the population level.

Measures such as cash transfers, wage subsidies, school-based mental health support and services for survivors of violence did not eliminate distress, but they reduced its scale and duration. These findings challenge the tendency to treat mental health as a downstream consequence of crises, something to be addressed only after economic and social damage has occurred.

COVID-19 showed that when livelihoods collapse, mental health suffers rapidly at scale, and that waiting until people present to clinics is not prevention. It is delayed response.

Lessons in crisis preparedness
The pandemic also exposed the limits of health-system-only approaches to mental health in emergencies. Even well-resourced services struggled when demand rose suddenly and staff were redeployed or unavailable. In many settings, mental health systems simply did not have the capacity to absorb shock.

Community-based and task-shared approaches showed particular promise, especially where they were embedded in existing social and primary care structures. These models are often more flexible, culturally appropriate and scalable than specialist-only services. Yet they remain under-evaluated and underfunded, particularly during crises.

Digital care will continue to play an important role, but COVID-19 demonstrated that digital-first strategies can deepen exclusion if they are not deliberately designed to reach those most at risk. Crisis preparedness therefore requires the deliberate integration of mental health into social protection, primary care, education and community support, rather than reliance on any single delivery channel.

Key lessons from African contexts
Evidence from African countries and other low-resource settings offers critical insights into how mental health systems function under constraint. In these contexts, the links between livelihoods, social cohesion and mental wellbeing are often more visible, and the consequences of social disruption are felt more quickly.

The review shows that mental health deteriorates fastest where informal economies collapse, food insecurity rises and social safety nets are weak. At the same time, it highlights the potential of community-based and task-shared care to reach large populations when properly supported.

Importantly, African evidence in this work does not simply illustrate vulnerability. It demonstrates innovation. Community networks, non-specialist providers and locally adapted interventions played a vital role during the pandemic, even when formal services were disrupted. These experiences have global relevance as countries everywhere grapple with workforce shortages and rising demand.

Looking ahead
This work is not the final word. On top of the first and second article, a third paper, currently in development, will examine the longer-term mental, cognitive and neurological consequences of COVID-19 and long COVID, issues that will shape mental health systems for years to come.

Future crises, whether pandemics, climate shocks or economic disruptions, are inevitable. Repeating the mental health harms of COVID-19 is not. The evidence now makes clear that mental health is shaped by the conditions in which people live, work and learn. Whether societies act on that knowledge will determine how resilient we are when the next crisis arrives.