Few studies from the general population have investigated the role of anxiety in reporting dyspnoea. We examined the independent and combined association of lung function and anxiety symptoms with the prevalence of dyspnoea in different situations.

The study included 5627 women and 5066 men who participated in the Lung study of the Nord-Trøndelag Health Study second survey in 1995–97. In a cross-sectional design we used logistic regression to calculate adjusted odds ratios (ORs) for reporting dyspnoea associated with levels of percent predicted FEV₁ (ppFEV₁) and anxiety (Hospital Anxiety and Depression Scale).

Overall, there was a linear inverse association between ppFEV₁ and dyspnoea (all P_trend < 0.001), and a positive association between anxiety symptoms and dyspnoea (all P_trend < 0.001). In combined analysis, using people with ppFEV₁ ≥100 without anxiety as reference, the OR (95% confidence interval) for reporting dyspnoea when walking on flat ground was 6.23 (3.45–11.28) in women with ppFEV₁ <80 without anxiety and 15.14 (7.13–32.12) in women with ppFEV₁ <80 with anxiety. The corresponding ORs among men were 5.75 (2.23–14.18) and 15.19 (4.74–48.64), respectively. Similar patterns were seen for dyspnoea when sitting still and woken at night by dyspnoea.

Impaired lung function and anxiety symptoms were independently associated with reporting dyspnoea. Within lung function levels, reporting dyspnoea was more common among people with anxiety symptoms than among people without. This suggests that, in addition to its relation to reduced lung function, the subjective experience of breathing discomfort may also influence or be influenced by anxiety.