Why Epidemic Waves Keep Coming Back Without New Variants

Epidemic waves don’t need viral mutations or seasonal changes to keep rolling through populations.

New research reveals that simple human psychology, specifically how we respond to delayed information about disease spread, can create multiple waves of infection all by itself. The finding challenges assumptions about what drives recurring outbreaks and suggests public health officials may need to rethink intervention timing.

Scientists at Iowa State University built mathematical models showing how information delays between actual infection rates and public awareness create a feedback loop. When disease spreads rapidly, people don’t know to take precautions yet. By the time they learn about rising cases and start social distancing or masking, infections have already peaked and begun falling.

The Delayed Response Cycle

Here’s where it gets interesting: as case numbers drop, people gradually relax their guard. But there’s another delay—this time between falling case reports and behavioral changes. This creates what researchers call “epidemic rebounds” without any external factors like new variants or seasonal weather patterns.

The study found that delays of about 5-7 days between disease prevalence and behavioral response produced the most epidemic waves. Shorter delays led to single, prolonged outbreaks. Much longer delays mimicked standard epidemic patterns with one major peak.

What’s particularly striking is how the disease generation time—the average period between infection and onward transmission—directly correlates with optimal wave-generating delays. Faster-spreading diseases like COVID-19 created maximum waves with shorter information delays, while slower diseases required longer delays to produce the same oscillating pattern.

When Fewer Waves Mean Fewer Infections

Counter-intuitively, the research revealed that scenarios producing fewer epidemic waves sometimes resulted in lower total infection rates. The models showed final epidemic sizes ranging from 52% to 71% of the population, compared to nearly 80% without behavioral responses.

This occurs because larger initial waves can deplete the susceptible population enough to prevent subsequent waves entirely. Parameter choices at the threshold where wave numbers change produced the lowest final epidemic sizes—suggesting that a few damped oscillations might be optimal for minimizing total infections.

Beyond Simple Behavioral Models

The researchers tested their findings using both Hill functions and logistic functions to model behavioral responses, with qualitatively similar results. They also validated the core findings using SEIR models (which include an exposed/latency period) rather than just standard SIR models.

Key factors influencing wave patterns included:

• Behavioral response sensitivity—how quickly people react to prevalence changes
• Response midpoint—the infection level triggering 50% contact reduction
• Disease transmission and recovery rates
• Information delay duration

Real-World Implications

The model helps explain patterns observed during early COVID-19 waves in the United States, when behavioral responses rather than policy mandates primarily shaped transmission. The research suggests that public health interventions face an inherent challenge: successful short-term disease reduction can inadvertently set up conditions for subsequent waves.

As the authors note in their conclusion, “policymakers should take into account the adaptive nature of human behavior when designing epidemic intervention strategies.” The work emphasizes integrating social and operational factors into infectious disease models rather than focusing solely on biological mechanisms.

The study appears in PNAS Nexus and represents a significant step toward understanding how human psychology shapes epidemic patterns independent of viral evolution or environmental factors. For public health officials, it offers both sobering insights about the challenges of maintaining vigilance and potential strategies for optimizing intervention timing.

Looking Forward

Future research could explore more complex behavioral responses, including varying compliance levels within population subgroups and factors like “epidemic fatigue” or economic constraints. The current model assumes behavioral responses depend solely on disease prevalence, but real-world decision-making involves additional psychological and social factors.

The findings underscore a fundamental tension in epidemic management: the very success of public health measures in reducing immediate disease spread can create conditions for future outbreaks through relaxed behavioral vigilance.