Epidemiology of panic attacks, panic disorder and the moderating role of age: Results from a population-based study.

BACKGROUND

The aim of this study was to investigate the prevalence of panic attacks (PA) and panic disorder (PD) over the lifespan, determine the main correlates and analyze the potential moderating role of age.

METHOD

We analyzed cross-sectional data from a nationally-representative sample of 4,569 non-institutionalized adults. Three panic groups were created according to results in the CIDI interview: non-panic, PA (without PD) and PD. Panic groups were used as outcomes in adjusted multinomial regression models where several correlates were investigated. Interactions between each covariate and age were explored.

RESULTS

The highest prevalence rates of PA (9.5%) and PD (3.3%) were found in people aged 30-39 and 40-49, respectively. Respondents aged 80 + presented the lowest rates. In the adjusted multinomial model, younger ages, having depression, and poorer levels of quality of life were significant correlates for both PA and PD, whereas being female and having 2 or more chronic conditions were only significant for PA (not for PD) and being a frequent drinker (last 12-months) only for PD. Age significantly interacted with frequent alcohol consumption in the last 12-months for PD. The predicted probability of having PD decreased with age for life-abstainer or occasional drinkers, whereas the probability increased with older ages for frequent drinkers.

LIMITATIONS

Low prevalence of PA and PD resulted in large confidence intervals whereas self-reports could be affected by recall bias.

CONCLUSIONS

Overall, our results suggest that the probability of having PA and PD decreases as people age and that the significant risk factors are independent of age. However, older adults with a frequent drink pattern seem to be at higher risk of PD. Future longitudinal studies are needed to determine the trajectories of PD and alcohol consumption over the lifespan.