Dyspnea (breathlessness) is a multidimensional subjective sensation of respiratory discomfort (5). In healthy subjects, dyspnea increases with the intensity of exercise, and a feeling of dyspnea is normal during heavy exertion, whereas patients with advanced cardiorespiratory or neuromuscular disorders might experience dyspnea even at rest (1). Dyspnea can be evaluated with different tools (9). In this article, dyspnea is used to describe shortness of breath that is caused by different levels of daily activities and therefore only the following measurement tool is presented here. A widely used version in epidemiological research is the four-question modified British Medical Research Council (mMRC) scale, which is an activity-based unidimensional scale (26,29). Increased dyspnea measured by mMRC scale has been shown to associate with higher age, socioeconomic disadvantages, smoking history, obesity, and physical inactivity (4). High dyspnea score measured with the mMRC scale can be considered as poor cardiorespiratory fitness and exercise intolerance because dyspnea score has been found to correlate moderately well with the 6-min walking distance, at least in obese (20) and chronic obstructive pulmonary disease patients (r = −0.51 and −0 to 53, P < 0.01) (5,7). There are no studies examining this relation among normal weight healthy subjects, and therefore, we can only hypothesize that the correlation would be similar among healthy subjects. A recent update on dyspnea issued by the American Thoracic Society stated that, especially among the sedentary population, dyspnea can be viewed as a manifestation of poor cardiovascular fitness (23), which is what dyspnea is considered here as.
There are few studies examining the association between dyspnea and all-cause mortality in healthy subjects. Stavem et al. (28) found that dyspnea, measured with the mMRC scale, was associated with all-cause mortality among healthy men even after adjusting for physical fitness and other known risk factors. Only one study has investigated how changes in dyspnea predict all-cause mortality; persistent dyspnea and the development of dyspnea predicted all-cause mortality, whereas remission of dyspnea was not associated with mortality (8). However, the baseline health status of the participants was not clear in that study. Neither of these studies took into account the level of physical activity.
Systematic reviews and meta-analyses of observational studies have consistently demonstrated the protective effect of physical activity (24,31) and cardiorespiratory fitness (16) on all-cause mortality. It is believed that the associations between physical activity/fitness and mortality could be affected by genetic factors predisposing to sedentariness (25,30), which also may affect the lifespan (17). Both physical fitness (3) and physical activity participation (11,30) contain a genetic component. Observational studies of unrelated individuals are not able to take these genetic components into account because the individual genes influencing the fitness-related parameters are mostly unknown. Therefore, if one wishes to investigate these problems with an observational study design, then twin studies can be a useful way of studying the net effect of genes and environment, especially for complex disorders and behavioral traits (2). Monozygotic (MZ) twins have the same genomic sequence and share many childhood experiences. If a difference is detected in the degree of dyspnea within MZ pair and this is associated with a systematic difference in future risk of death, then one must consider that the association cannot be accounted for by genetic confounding.
Among healthy people, the use of dyspnea experienced during specified daily activities as an indicator of fitness has not been widely used even though it correlates moderately well with fitness measurements. Therefore, the aim of this study was to test the causal nature of the association between baseline fitness measured as dyspnea or changes in the level of dyspnea occurring during the 6-yr baseline period, and premature mortality assessed as all-cause mortality during the 28-yr follow-up. The analyses are adjusted for genes and childhood environment by using a twin design; i.e., Finnish Twin Cohort data from years 1975 and 1981 were investigated to predict all-cause mortality. We hypothesized that persistent dyspnea or occasional symptoms (low fitness), even when adjusted for genes and childhood environment, would be associated with increased all-cause mortality risk.
The Finnish Twin Cohort is a nationwide sample of all same-sex twin pairs born before 1958 and with both cotwins alive in 1967 (13). A baseline questionnaire was sent out to all twin candidates in 1975. Among those whose addresses could be identified (93.5% of subjects) in 1975, the response rate for twins was 87.6% in this age group. The total number of subjects who answered the questionnaire was 26,225, including 12,069 complete twin pairs with both members alive at the beginning of the prospective follow-up. A subsequent questionnaire was mailed in 1981 to the verified twins. The corresponding response rate among those responding in 1975 and alive in 1981 was 90.7%. A total of 21,379 twin individuals (10,159 male and 11,220 female) answered the dyspnea questions in both questionnaires (1975 and 1981). This included 8672 complete twin pairs (2671 MZ, 5417 dizygotic (DZ), and the rest with unknown zygosity). Determination of zygosity was based on an accurate and validated questionnaire method (27). The participants were informed about the purposes of the overall cohort study when given the baseline questionnaire in 1975. By responding to the questionnaire, participants also gave their informed consent for future register follow-ups. The record linkages were approved by the appropriate authorities responsible for the registers and the Ethics Committee of the Department of Public Health, University of Helsinki.
To remove the confounding factors due to health status, we excluded those subjects who had reported experiencing either persistent cough or any lung disease because these could be assumed to have a major effect on dyspnea, leaving 17,937 individuals for analysis in 1981. In addition, we studied a subgroup of 14,375 apparently healthy individuals. Subjects with chronic diseases (such as angina pectoris, myocardial infarction, stroke, diabetes, cardiovascular disease, malignant cancer, and lung disease) possibly affecting the ability to engage in leisure physical activity before 1982 had been identified by a questionnaire in 1981 and from the medical records as described in detail elsewhere (18).
Baseline dyspnea and covariate assessment.
The postal questionnaires in 1975 and 1981 were very similar, including identical questions on physical activity, dyspnea, occupation, alcohol use, smoking, and physician diagnosed diseases. Dyspnea was assessed with a four-question mMRC dyspnea scale, and the questions were concerning whether the subject became breathless during walking and performing daily tasks (9). The questions are as follows: 1) Do you usually get short of breath when you walk uphill, climb stairs, or hurry on level ground? 2) Do you usually get short of breath when walking on level ground at an ordinary pace with people of your own age? 3) Do you have to stop to breathe because of shortness of breath when walking at your own pace for 150 meters? 4) Do you usually get short of breath when standing still, for example, when dressing or washing? (12). This scale has five response categories (range: 0, no yes responses, asymptomatic, to 4, all answers yes, breathless during daily tasks).
In the first set of analyses, when baseline 1975 and 1981 dyspnea was analyzed separately, dyspnea was used as a five-category variable. For the second set of analyses, dyspnea change variable was created. For that, five-level dyspnea was first divided into two categories, asymptomatic (0) versus others (1–4), as has been done in other studies (4). Then dyspnea change variable was constructed; subjects, who had no dyspnea during either of the baseline years, were “asymptomatic.” Subjects who had dyspnea during both baseline years belonged to the “persistent dyspnea” group. If subjects increased from no dyspnea in 1975 to any level of dyspnea in 1981, the group was called “development of dyspnea,” and if the level of dyspnea decreased from 1975 to no dyspnea in 1981, the group was called “remission.”
Age, sex, physical activity, BMI, social class, smoking, and alcohol use in 1975 and 1981 were used as covariates in the study. Physical activity was a continuous variable quantified as MET-hours per day. Assessment of leisure activity volume (MET index) was based on a series of structured questions (18) on leisure physical activity (monthly frequency, mean duration, and mean intensity) and physical activity during journeys to and from work. BMI was assessed by using baseline self-reported weight and height. The correlation between self-reported and measured BMI has been found to be very high in another study of Finnish twins (22). Smoking was used as continuous variable describing pack years smoked during life (14). Alcohol was used as a continuous variable expressed as grams consumed daily (15). Six categories were used to describe social class, on the basis of self-reported job titles using the classification criteria of the Central Statistical Office of Finland (6).
All-cause mortality is used as the follow-up outcome measure. Exact dates of death and emigration from Finland were available from the Population Register Centre of Finland.
All data were analyzed by using Stata IC 12 version. First, the Cox proportional hazard model was used to estimate how five-level baseline dyspnea in 1975 and 1981 predict all-cause mortality by the end of 2010. The analyses were started from May 1, 1976, for first analyses (because this was the time point when the 1975 questionnaire survey was completed for all subjects) and from the date of response of the individual participant to the 1981 questionnaire. Follow-up was until date of death, date of emigration, or end of follow-up, whichever came first. The follow-up time was computed in days using age as the timescale in the analyses. Second, the Cox proportional hazard model was used to estimate how persistence and changes in dyspnea (dyspnea change variable) predict mortality. Follow-up was started from the date of response of the individual participant to the 1981 questionnaire and continued as explained previously.
All analyses were first done as individual-based analyses. Because the observations obtained from twin pairs may be correlated, robust estimators of variance (the cluster option in Stata) were used when estimating SE. The individual analyses were adjusted for age and sex, and then for smoking because this seemed to be the most influential covariate, and then the full model with all confounding factors in the analyses (age, sex, smoking, physical activity, BMI, alcohol, and social class) was created. Only the 1981 variables were included as covariates in the model because further adjustments for these variables from 1975 did not change the HR. Because of missing data, inclusion of covariates decreased the number of participants in some models. All individual-based analyses were first done for all subjects and then for different subgroups: subjects with no respiratory disease or persistent cough in 1981 and healthy subjects in 1981, men and women. The pairwise analyses (using the strata option in Stata) were also conducted separately for dizygotic and MZ twin pairs, such that the baseline hazard was twin pair specific, and the comparison of mortality was within twin pairs. This yields a within-pair HR when summed over all pairs in which one has died and one is still alive.
For additional analysis, dyspnea was also divided into three categories (asymptomatic (0), mild dyspnea (grade 1 in mMRC scale), and severe dyspnea (grades 2–4)) to analyze persistence and change between these categories. Results for these are shown in the Supplemental Digital Content (SDC, http://links.lww.com/MSS/A353, http://links.lww.com/MSS/A354, http://links.lww.com/MSS/A355) material.
Table 1 shows the baseline characteristics of the cohort subdivided into the primary dyspnea categories (dyspnea change variable). Altogether, 55.2% of the subjects were asymptomatic during both baseline years, whereas 44.8% of subjects reported having had dyspnea symptoms at least on one of the years. Subjects who were asymptomatic during both baseline years were younger, had higher physical activity level, smoked less, and had lower BMI.
First, baseline data from years 1975 and 1981 were used to analyze mortality risk. Individual analyses showed that HR increased as severity of dyspnea increased (P for trend <0.001). Age- and sex-adjusted HR increased linearly with dyspnea level from 1.26 to 2.42 for level 4 dyspnea compared with being asymptomatic in 1975 (Fig. 1A, see also table in SDC 1, http://links.lww.com/MSS/A353 HR for mortality according to baseline 1975 dyspnea level). Pairwise analyses showed a similar trend (Fig. 1B), and a significant difference was also seen among MZ pairs (see table in SDC 1, http://links.lww.com/MSS/A353). Similar results were seen when mortality was predicted by 1981 dyspnea level (Fig. 2A and Table 2) and also seen when dyspnea in 1981 was additionally adjusted for dyspnea in 1975.
Second, persistence and change of dyspnea as predictors of mortality were analyzed. During the follow-up, 4828 deaths occurred among the 21,076 subjects, with 545,060 person-years follow-up. The Cox regression analyses showed higher HR for mortality in all three dyspnea groups. The age- and sex-adjusted HR for death was 1.69 (95% confidence interval (CI), 1.57–1.81) for subjects with persistent dyspnea, 1.22 (95% CI, 1.11–1.34) in dyspnea remission, and 1.32 (95% CI, 1.20–1.45) for onset of dyspnea compared with asymptomatic individuals (Fig. 2B and Table 3). The excess mortality decreased when the analysis was adjusted with all covariates, but it remained significant among dyspnea developers (HR, 1.16; 95% CI, 1.05–1.28) and in the persistent dyspnea group (HR, 1.41; 95% CI, 1.30–1.52). Similar results were also seen when men and women were analyzed separately, with men having slightly higher HR than women (Table 3). The risk remained significant even when baseline healthy subjects were analyzed separately, as fully adjusted HR for subjects with persistent dyspnea was 1.34 (95% CI, 1.16–1.55). The risk is clearly higher if baseline healthy subjects had persistent severe dyspnea (levels 2–4), i.e., their fully adjusted HR was 1.99 (95% CI, 1.32–3.00) compared with asymptomatic healthy subjects (see table in SDC 2, http://links.lww.com/MSS/A354 HR for mortality according severity of dyspnea persistency). Significant results were also seen in those with persistent mild (level 1) dyspnea because their fully adjusted HR was 1.25 (95% CI, 1.09–1.47).
Altogether, 4296 pairs were discordant for baseline dyspnea change and 1576 pairs were discordant for death during follow-up, and out of these, 863 (210 MZ and 588 DZ, rest with unknown zygosity) pairs were discordant for both baseline dyspnea and death during follow-up. Among the 420 death discordant MZ twin pairs, 53 pairs had extreme discordance with one twin being asymptomatic and his/her cotwin being dyspneic during both years; of these, in 35 pairs, the persistent dyspnea sufferer had died although his/her symptom-free cotwin was still alive at the end of follow-up. The converse was true among 18 pairs, giving a pairwise odds ratio of 1.94 (95% CI, 1.07–3.55; McNemar test P = 0.02). The numbers were 18 and 5, respectively, among health discordant MZ pairs, equivalent to an odds ratio of 3.6 (95% CI, 1.29–12.4, P = 0.0067).
Pairwise analyses of survival time showed similar HR as observed in the individual analyses. Thus, among all subjects, HR was 1.56 (95% CI, 1.33–1.83) for persistent dyspnea, 1.17 (95% CI, 0.97–1.42) for dyspnea development, and 1.27 (95% CI, 1.05–1.54) for remission (Fig. 2C), whereas HR in the fully adjusted model was 1.47 (95% CI, 1.23–1.77) for persistent dyspnea. Significant results were also seen among healthy MZ pairs with the persistent dyspnea group because the HR in basic analysis was 2.40 (95% CI, 1.25–4.59), and in smoking adjusted analyses, 2.47 (95% CI, 1.22–5.01) and 2.64 (95% CI, 1.21–5.74) in the fully adjusted model compared with asymptomatic subjects (Table 4). The results for three category variable show that healthy MZ pairs had clearly increased HR even with mild persistent dyspnea because HR was 2.50 (1.13–5.52) in the fully adjusted model (see table in SDC 3, http://links.lww.com/MSS/A355 HR for mortality according to severity of dyspnea persistency in pairwise setting).
The study shows that any level of dyspnea increases the mortality risk during three decades of follow-up. This was also seen in apparently healthy subjects as well as separately in men and women. However, remission from dyspnea did seem to reverse the risk at least in fully adjusted models. The highest HR was recorded in the pairwise analyses among subjects with persistent severe dyspnea during the 6-yr baseline period. Pairwise analyses showed that the risk was also persistent within baseline healthy dizygotic and MZ pairs.
The present study showed that dyspnea/fitness is associated with all-cause mortality even after adjustment for various known risk factors and in healthy population. Stavem et al. (28) found similar HR (1.47; 95% CI, 1.08–2.00) as in the present study. Only one other study (8) has examined how changes in dyspnea predict mortality, and in that, the development of dyspnea and persistent dyspnea was associated with increased all-cause mortality. In that study, remission of dyspnea was not associated with mortality, which is also consistent with our overall results. Decreasing the dyspnea level without becoming asymptomatic did not seem to reverse the association completely (see tables in SDC 2, http://links.lww.com/MSS/A354 and SDC 3, http://links.lww.com/MSS/A355), indicating that being completely asymptomatic is important. The only other study (8) that investigated the associations between change in dyspnea and mortality did not perform a subgroup analysis of only healthy subjects. The present study was moreover able to control unmeasured familial and genetic factors by using a discordant twin pair study design. In addition, we were able to adjust for physical activity level. It is of note that most of the studies on dyspnea have been conducted in patients experiencing lung disease, such as chronic obstructive pulmonary disease, and therefore, these subjects cannot be compared with the healthy population.
Interestingly, one of the highest HR was seen in fully adjusted model among healthy MZ pairs (HR, 2.64; P = 0.014) when examining the differences between subjects with persistent dyspnea (any level of dyspnea) and those with no symptoms. No significant differences were seen if subjects were asymptomatic during only one of the baseline years (developers or remission) in the fully adjusted model among healthy DZ or MZ pairs. These results might indicate that occasional dyspnea might not be such a major risk for all-cause mortality, especially if lifestyle changes are undertaken or illnesses causing dyspnea are treated. Physical activity could be one of the important modifiable risk factors, as can be seen in the baseline characteristics table. Those subjects who developed dyspnea during the 6-yr baseline period decreased their physical activity levels (mean PA was 2.1 MET·h·d−1 in 1975 and 1.9 MET·h·d−1 in 1981, P = 0.002); in contrast, those subjects who became asymptomatic during the 6-yr baseline had increased their physical activity level (from 1.8 MET·h·d−1 in 1975 to 2.4 MET·h·d−1 in 1981, P < 0.001). This indicates that increasing the physical activity level might reduce future dyspnea, but this needs to be confirmed. However, this relation could become like a vicious cycle as early signs and fear of dyspnea might cause patients (or any individual) reduce their daily activities, which then results in reduced cardiovascular fitness and muscle strength, and again that increases the feeling of dyspnea (10). Alternatively, some other cause of increased dyspnea may also result in less physical activity.
An increased risk of mortality was seen among MZ pairs even when they displayed persistent but only mild dyspnea. This is clear evidence that dyspnea is a causal contributor of all-cause mortality even among middle-age healthy people, and the association is not due to genetic confounding. Because dyspnea can be seen as a manifestation of poor cardiovascular fitness in the sedentary population (23), it could be assumed that only increases in fitness status could decrease the levels of dyspnea and all-cause mortality. Although both fitness and physical activity are well-known predictors of mortality (16,31), according to Lee et al. (21), fitness seems to be an independent predictor of mortality even when adjusted for physical activity, but the association is not totally clear the other way around. However, this could be due to less accurate measurement of physical activity (self-reported) versus fitness. In our study, poor fitness (indicated as increased dyspnea) was found to be a predictor of mortality, as even mild persistent, and once measured, dyspnea adjusted for physical activity increased the mortality HR. Asymptomatic subjects answered that they “do not usually get short of breath when they walk uphill, climb stairs, or hurry on level ground”, which indicates a good fitness level. A simple question on dyspnea in common, everyday situations seems to be able to identify those individuals at increased risk. Therefore, this easily and quickly used mMRC scale has a potential to be used as a screening tool for finding those individuals who are unfit and therefore who are at risk of premature all-cause mortality. However, the relation between dyspnea and fitness among healthy individuals should be investigated to enable stronger conclusions.
Strengths and limitations.
The present study has many strengths: long follow-up, a population-based twin study including all same-sex twin pairs born in Finland before 1958, a large number of subjects (n > 21,000) with a sample size greater than that in any earlier study, twin study design controlling for the genetic predisposition and childhood family environment, as well as recording changes in dyspnea during the baseline period. Because health status was recorded at baseline, we were able to study healthy subjects separately, revealing that dyspnea not only is a symptom of illness but also can be considered as a valid indicator of poor fitness. Even though it has been discussed whether the results from twin studies are generalizable, it is noteworthy that twins are considered to be similar as the other people of the same birth cohort in the same nation (19). There is no clear evidence why twin studies could not be generalized (19), especially no literature is available in relation to lung function or mortality.
The study has a few limitations. The main baseline predictors were collected with questionnaires, and no objective measurements of respiratory functions or fitness were performed. Therefore, we were not able to adjust the models to include all known risk factors, for example, lung function, which is associated with dyspnea. One of the weaknesses is inherent in the dyspnea questionnaire because the mMRC dyspnea scale only relates to specific walking activities, and therefore, the scale does not provide a multidimensional evaluation of dyspnea (5). However, as a clinical or public health tool, a short questionnaire could be easily used, and therefore, it is important to investigate the predictive value of these kinds of short and easily used tools. Also, this type of tool measures the activity-related dyspnea (breathlessness) that can be considered to express physical fitness, which was the aim of this article.
The results show that persistent dyspnea or dyspnea development predict an increase in all-cause mortality during a 28-yr follow-up even among healthy subjects and within MZ twin pairs. An easily obtained dyspnea score, which correlates with fitness outcomes, could have the potential for use as a screening tool for identifying individuals with low levels of fitness and an increased mortality risk.
This work was supported by the Academy of Finland (grants 265240, 263278) and META-PREDICT (within the European Union Seventh Framework Programme, HEALTH-F2-2012-277936). The funding sources had no role in the design and conduct of the study; the collection, analysis, and interpretation of data; or the preparation, review, or approval of the manuscript.
None of the authors acknowledge any relations/conditions/circumstances that presents a potential conflict of interest, and the results of the present study do not constitute endorsement by the American College of Sports Medicine.
1. American Thoracic Society. Dyspnea
. Mechanisms, assessment, and management: a consensus statement. American Thoracic Society. Am J Respir Crit Care Med
. 1999; 159 (1): 321–40.
2. Boomsma D, Busjahn A, Peltonen L. Classical twin studies and beyond. Nat Rev Genet
. 2002; 3 (11): 872–82.
3. Bouchard C, Dionne FT, Simoneau JA, Boulay MR. Genetics
of aerobic and anaerobic performances. Exerc Sport Sci Rev
. 1992; 20: 27–58.
4. Bowden JA, To TH, Abernethy AP, Currow DC. Predictors of chronic breathlessness: a large population study. BMC Public Health
. 2011; 11: 33.
5. Camargo LA, Pereira CA. Dyspnea
in COPD: beyond the modified Medical Research Council scale. J Bras Pneumol
. 2010; 36 (5): 571–8.
6. Central Statistical Office of Finland. Alphabetical list of occupations and classification of social class. Helsinki, Finland: Central Statistical Office, 1972.
7. Eakin EG, Sassi-Dambron D, Ries AL, Kaplan RM. Reliability and validity of dyspnea
measures in patients with obstructive lung disease. Int J Behav Med
. 1995; 2 (2): 118.
8. Figarska SM, Boezen HM, Vonk JM. Dyspnea
severity, changes in dyspnea
status and mortality
in the general population: the Vlagtwedde/Vlaardingen study. Eur J Epidemiol
. 2012; 27 (11): 867–76.
9. Gerlach Y, Williams MT, Coates AM. Weighing up the evidence—a systematic review of measures used for the sensation of breathlessness in obesity. Int J Obes (Lond)
. 2013; 37 (3): 341–9.
10. Hayen A, Herigstad M, Pattinson KTS. Understanding dyspnea
as a complex individual experience. Maturitas
. 2013; 76 (1): 45–50.
11. Kaprio J, Koskenvuo M, Sarna S. Cigarette smoking, use of alcohol, and leisure-time physical activity among same-sexed adult male twins
. Prog Clin Biol Res
. 1981; 69 (Pt C): 37–46.
12. Kaprio J, Sarna S, Koskenvuo M, Rantasalo I. The Finnish Twin Registry: Baseline Characteristics. Section II. History of symptoms and illnesses, use of drugs, physical characteristics, smoking, alcohol and physical activity. Helsinki: Public Health Publication M 37; 1978. Available From: Department of Public Health, University of Helsinki.
13. Kaprio J, Koskenvuo M. Genetic and environmental factors in complex diseases: the older Finnish Twin Cohort. Twin Res
. 2002; 5 (5): 358–65.
14. Kaprio J, Koskenvuo M. A prospective study of psychological and socioeconomic characteristics, health behavior and morbidity in cigarette smokers prior to quitting compared to persistent smokers and non-smokers. J Clin Epidemiol
. 1988; 41 (2): 139–50.
15. Kaprio J, Koskenvuo M, Langinvainio H, Romanov K, Sarna S, Rose RJ. Genetic influences on use and abuse of alcohol: a study of 5638 adult Finnish twin brothers. Alcohol Clin Exp Res
. 1987; 11 (4): 349–56.
16. Kodama S, Saito K, Tanaka S, et al. Cardiorespiratory fitness as a quantitative predictor of all-cause mortality
and cardiovascular events in healthy men and women: a meta-analysis. JAMA
. 2009; 301 (19): 2024–35.
17. Kujala UM. Physical activity, genes, and lifetime predisposition to chronic diseases. Eur Rev Aging Phys Act
. 2011; 8: 31–6.
18. Kujala UM, Kaprio J, Sarna S, Koskenvuo M. Relationship of leisure-time physical activity and mortality
: the Finnish twin cohort. JAMA
. 1998; 279 (6): 440–4.
19. Kyvik KO. Generalisability and assumptions of twin studies. In: Spector TD, Snieder H, MacGregor AJ, editors. Advances in Twin and Sib-Pair Analysis
. London: Greenwich Medical Media; 2000. pp. 53–66.
20. Launois C, Barbe C, Bertin E, et al. The modified Medical Research Council scale for the assessment of dyspnea
in daily living in obesity: a pilot study. BMC Pulm Med
. 2012; 12 (1): 61.
21. Lee DC, Sui X, Ortega FB, et al. Comparisons of leisure-time physical activity and cardiorespiratory fitness as predictors of all-cause mortality
in men and women. Br J Sports Med
. 2011; 45 (6): 504–10.
22. Mustelin L, Silventoinen K, Pietilainen K, Rissanen A, Kaprio J. Physical activity reduces the influence of genetic effects on BMI and waist circumference: a study in young adult twins
. Int J Obes (Lond)
. 2009; 33 (1): 29–36.
23. Parshall MB, Schwartzstein RM, Adams L, et al.American Thoracic Society Committee on Dyspnea
. An official American Thoracic Society statement: update on the mechanisms, assessment, and management of dyspnea
. Am J Respir Crit Care Med
. 2012; 185 (4): 435–52.
24. Physical Activity Guidelines Advisory Committee. Physical Activity Guidelines Advisory Committee Report, 2008
. Washington (DC): Available From: U.S. Department of Health and Human Services; 2008.
25. Rankinen T, Bouchard C. Invited commentary: physical activity, mortality
, and genetics
. Am J Epidemiol
. 2007; 166 (3): 260–2.
26. Rose GA, Blackburn H. Cardiovascular survey methods. Monogr Ser World Health Organ
. 1968; 56: 1–188.
27. Sarna S, Kaprio J, Sistonen P, Koskenvuo M. Diagnosis of twin zygosity by mailed questionnaire. Hum Hered
. 1978; 28 (4): 241–54.
28. Stavem K, Sandvik L, Erikssen J. Breathlessness, phlegm and mortality
: 26 years of follow-up in healthy middle-aged Norwegian men. J Intern Med
. 2006; 260 (4): 332–42.
29. Stenton C. The MRC breathlessness scale. Occup Med (Lond)
. 2008; 58 (3): 226–7.
31. Woodcock J, Franco OH, Orsini N, Roberts I. Non-vigorous physical activity and all-cause mortality
: systematic review and meta-analysis of cohort studies. Int J Epidemiol
. 2011; 40 (1): 121–38.