Dyspnoea as a predictor of cause-specific heart/lung disease mortality in Bangladesh: a prospective cohort study

• Published in: Journal of epidemiology and community health (1979) Vol. 70; no. 7; pp. 689 - 695
• Main Authors: Pesola, Gene R, Argos, Maria, Chinchilli, Vernon M, Chen, Yu, Parvez, Faruque, Islam, Tariqul, Ahmed, Alauddin, Hasan, Rabiul, Rakibuz-Zaman, Muhammad, Ahsan, Habibul
• Format: Journal Article
• Language: English
• Published: England BMJ Publishing Group Ltd 01.07.2016; BMJ Publishing Group LTD
• Subjects: Aged; Aged, 80 and over; Arsenic; Bangladesh - epidemiology; Cardiovascular disease; Cardiovascular diseases; Causes of death; Chronic illnesses; Chronic obstructive pulmonary disease; Data collection; Developing countries; Dyspnea; Epidemiology; Female; Health risk assessment; Heart diseases; Heart Diseases - mortality; Heart failure; Humans; Industrialized nations; Kaplan-Meier Estimate; Laboratories; LDCs; Lung diseases; Lung Diseases - mortality; Lungs; Male; Middle Aged; Mortality; Mortality rates; Mycobacterium; Other topics; Prospective Studies; Public health; Pulmonary heart disease; Pulmonary tuberculosis; Questionnaires; Stroke; Studies; Tobacco smoking; Tuberculosis; Bangladesh; ISW, Bangladesh; LONGITUDINAL STUDIES; Cardiovascular disease; Epidemiology of chronic diseases; AVOIDABLE DEATHS; Clinical epidemiology
• ISSN: 0143-005X; 1470-2738
• DOI: 10.1136/jech-2015-206199

BackgroundThe spectrum of mortality outcomes by cause in populations with/without dyspnoea has not been determined. The study aimed to evaluate whether dyspnoea, a symptom, predicts cause-specific mortality differences between groups. The hypothesis was that diseases that result in chronic dyspnoea, those originating from the heart and lungs, would preferentially result in heart and lung disease mortality in those with baseline dyspnoea (relative to no dyspnoea) when followed over time.MethodsA population-based sample of 11 533 Bangladeshis was recruited and followed for 11–12 years and cause-specific mortality evaluated in those with and without baseline dyspnoea. Dyspnoea was ascertained by trained physicians. The cause of death was determined by verbal autopsy. Kaplan-Meier survival curves, the Fine-Gray competing risk hazards model and logistic regression models were used to determine group differences in cause-specific mortality.ResultsCompared to those not reporting dyspnoea at baseline, the adjusted HRs were 6.4 (3.8 to 10.7), 9.3 (3.9 to 22.3), 1.8 (1.2 to 2.8), 2.2 (1.0 to 5.1) and 2.8 (1.3 to 6.2) for greater risk of dying from chronic obstructive pulmonary disease (COPD), asthma, heart disease, tuberculosis and lung cancer, respectively. In contrast, there was a similar risk of dying from stroke, cancer (excluding lung), liver disease, accidents and other (miscellaneous causes) between the dyspnoeic and non-dyspnoeic groups. In addition, the HR was 2.1 (1.7 to 2.5) for greater all-cause mortality in those with baseline dyspnoea versus no dyspnoea.ConclusionsDyspnoea, ascertained by a single question with binary response, predicts heart and lung disease mortality. Individuals reporting dyspnoea were twofold to ninefold more likely to die of diseases that involve the heart and/or lungs relative to the non-dyspnoeic individuals. Therefore, in those with chronic dyspnoea, workup to look for the five common dyspnoeic diseases resulting in increased mortality (COPD, asthma, heart disease, tuberculosis and lung cancer), all treatable, should reduce mortality and improve the public health.