Respiratory disease is the third main cause of mortality worldwide and the incidence keeps increasing on both disability adjusted life years (DALYs) and mortality.1–3 The most common respiratory diseases that are the leading cause of mortality are chronic respiratory diseases (CRD), including Chronic Obstructive Pulmonary Disease (COPD) as the most prevalent, followed by asthma.2,3 For instance, in 2017, the estimated deaths due to CRD were 3.91 million, which composed 7% of all deaths worldwide.3 Therefore, respiratory diseases are a major issue in public health that needs to be tackled.

Skeletal muscle dysfunction appears with aging4 and is common and exacerbated in CRD.5,6 This dysfunction has a great negative impact on physical performance, leading to disability,7 and importantly, is related to increased mortality risk.5,8 However, among health care providers, a lack of awareness regarding the critical role of skeletal muscle dysfunction still prevails.5,9 An easy-to-use and objective tool to evaluate whether skeletal muscle dysfunction occurs, is measuring handgrip strength. Apart from being an indicator of overall muscle strength,10–12 handgrip strength has been widely used as a biomarker of health status as well as to predict mortality.12,13

While the association between handgrip strength and all-cause mortality is more deeply explored,13–15 no previous studies have been specifically focused on handgrip strength and respiratory disease mortality, and thus its relationship remains unclear. For instance, a cohort study among 502 293 participants from the UK investigating all-cause mortality as well as disease specific mortality reported that lower handgrip strength was related to higher all-cause mortality as well as to cause-specific mortality from respiratory diseases.14 However, previous data from the Prospective Urban Rural Epidemiology (PURE) study,15 where an association between handgrip strength and all-cause mortality was found, no significant association between handgrip strength and hospital admission due to respiratory disease was revealed. In addition, a recent meta-analysis could not determine causality between disease-specific mortality, including respiratory diseases, and handgrip strength.13 Notably, none of the previous studies mentioned above included a sample involving a large number of European countries. Since differences in cultural, environmental and lifestyle factors might influence the results,14 new studies with more representative data from this specific continent are needed. Furthermore, understanding the potential causal and confounding pathways between handgrip strength and respiratory disease mortality in a large representative sample of subjects could help to clarify the potential role of this measure as a clinical tool for mortality risk estimation of respiratory disease. Because higher levels of handgrip strength might relate to both chronic respiratory diseases such as chronic obstructive pulmonary disease (COPD) and acute respiratory diseases such as pneumonia,16,17 a potential joint effect of handgrip strength could be considered. The main objective of this study was to investigate the association between handgrip strength and respiratory disease mortality using only adults free from prior or current known respiratory disease. We hypothesised that higher handgrip strength was prospectively associated with reduced respiratory disease mortality risk.

In this large-scale study including 135 787 subjects, each one kg increase of handgrip strength was prospectively associated with a substantial respiratory disease mortality reduction.

Our result endorses previous findings, where higher handgrip strength was inversely related to mortality.13–15 However, despite the association found between all-cause mortality and handgrip strength, in a large-scale longitudinal study conducted in 17 countries, no association between handgrip strength and hospital admission due to respiratory disease was found.15 Nevertheless, hospital admission is the result of acute respiratory deterioration, and since the underlying mechanism of CRD differs from that of an acute exacerbation5,25 a different effect on skeletal muscle dysfunction could be expected, which could have caused the difference in findings. In addition, the median follow-up time at the aforementioned study was merely four years whereas in our study, the median follow up was 7.4 years. Because skeletal muscle dysfunction is more extensive as the disease progresses,5,26 this might explain the controversial results. The meta-analysis by Lee showing an inverse association of handgrip strength with all-cause mortality,13 included one study that did specifically investigate the association between handgrip strength and respiratory disease.27 This is a 24-year follow-up study from the UK, where a negative association was found between handgrip strength and respiratory disease mortality. However, after adjusting for potential confounders (height, smoking, social class, physical activity, diagnosed disease at baseline, calorie intake, reported weight loss, and the measures of body composition) the association was not statistically significant. Since that study included only participants from the UK and as there are differences in cultural and environmental factors other than adjusted for, like the use of alcohol or diet preferences,28,29 this might explain the difference in the results. Interestingly, in a more recent meta-analysis investigating the association between handgrip strength and mortality, morbidity and health related quality of life in patients with COPD, a small to moderate association between handgrip strength and mortality was found,30 however within the included studies contrasting evidence was present. For instance, authors of a longitudinal cohort study claimed that they could not confirm associations between handgrip strength and respiratory disease mortality that are previously found in cross-sectional studies.31 However, this longitudinal study only included 194 patients with COPD that were followed-up and it has been shown that the disease progression and the procedure of skeletal muscle dysfunction within COPD patients are very variable.5,32 A possible mechanism explaining the variability of skeletal muscle dysfunction in patients with COPD is that there are three different domains in which muscle dysfunction is present: a clinical/functional domain, a metabolic domain, and an anatomical domain, and all of these domains can get dysfunctional in an independent way and therefore cause a variable and unique manifestation of skeletal muscle dysfunction.5,32 The most common CRD that is the leading cause of mortality is COPD,2,3 and therefore causing a high variability of skeletal muscle dysfunction in our group of participants. Nevertheless, our longitudinal study includes a large heterogeneous population and may therefore capture the high variability of muscle dysfunction within the participants.

We found a linear association for respiratory disease mortality risk reduction and an increase of handgrip strength, indicating higher muscle strength levels are consistently associated with lower respiratory disease mortality risk. In fact, we found that an increase of 1 kilogram of handgrip strength was associated with a respiratory disease mortality reduction of 5%. It is likely that handgrip strength is simply a proxy measures of overall muscle strength. Thus, a training program aiming to improve overall strength levels might be beneficial for patients with respiratory disease. We suggest that by training and increasing handgrip strength we can intercept the clinical/functional domain as a mechanisms accounting for skeletal muscle dysfunction seen for patients with respiratory disease.5 The clinical/functional domain is about the endurance and strength of the muscle, and this can be improved by training.33 However, this should be tested in large-scale randomized trials before any firm recommendations can be provided.

Although several validated tests assess the systemic involvement of CRDs and therefore obtain a reasonable measurement of muscle dysfunction,34 the handgrip measurement offers some advantages, particularly because it is easy to perform, is low cost, and requires very little equipment. In addition, it can be used in contexts in which patients cannot walk or make significant physical effort, such as in the 6 min walk test or have difficulty performing strength manoeuvres (e.g., knee flexors or extensors).35

The key strengths of the present study are the use of a large and representative sample from 28 countries with an objective measure of handgrip strength. Another critical strength is the use of time-varying handgrip strength and covariates in our modelling strategy, which reduces the possibility of bias of our estimates, which also accounted for competing risks of mortality due to other causes). Furthermore, repetitive measures of handgrip strength served to model the continuous dose-response association of handgrip strength with respiratory disease mortality, which is critical to identifying both thresholds and shape of the dose-response association. Furthermore, we also took measures to minimise the chance of reverse causation (i.e., lower handgrip strength as result of the course of the disease) by removing all the events that occurred within the first 2 years of follow up. On the other hand, there are several limitations to be considered in the present study. First, due to the number of participants with missing values concerning time-varying covariates, we imputed a substantial number of values, which might lead to some measuring bias, although complete case analyses scarcely differed from the main analyses. Second, due to the low number of events, the estimates for the fully adjusted model (Model B) are considerably unstable for sensitivity analyses regarding dose-response associations. Third, the participation rate at baseline was moderate (56%), which possibly increased the risk of selection bias. However, such losses are compensated through the use of refresher samples.19 Also, there is the chance of some attrition bias which might hamper the accuracy of our estimations, but the average retention rate in SHARE (81%) importantly reduces such possibility.19 For this reason, we included a weight variable in the analyses to compensate both non-response and attrition. Fourth, the use of a proxy relative for assessing respiratory disease mortality might lead to a certain degree of misclassification in that variable. Although prior research has observed that a relative proxy is a robust substitute to identify death status in adult populations when information on death is not available,36 information on specific mortality cause may be subject to a higher risk of misclassification bias. Fifth, because individuals with other prior or current chronic respiratory disease conditions other than COPD could not be identified, several of the mortality outcomes of the present studies might be more related to other prior critical chronic respiratory conditions such as asthma instead of lower handgrip strength.2 Lastly, because the prior or current COPD condition is self-reported, certain degree of measurement bias concerning this variable is still plausible

Lower handgrip strength is associated with higher mortality due to respiratory disease. Improving handgrip strength by one single kilogram in patients with respiratory disease, could potentially reduce their mortality by 5%.

None declared.

RLB is supported by the European Union - Next Generation EU.