I have compared my data published in J Appl Physiol 108:1440-1446, 2010 with results showed by Elizabeth T Thomas et al. in BMJ Open 2019;9:e028150. doi:10.1136/bmjopen-2018-028150. Could anybody explain why the rate of FEV1 decline with ageing analyzed in longitudinal studies is twice higher that the rate of FEV1 decline with age analyzed in my cross-section study (where ‘age’ means both ageing and differences between generations). The only explanation that I might propose is that older generations have/had better lung functions, and thus their FEV1 decline with ageing is partly compensated by better starting lung functions. In consequence, we observe seemingly lower decline with ageing in cross-sectional studies. This seems, however, to be an odd explanation since, at least in my opinion (e.g., see my last article ‘Spirometry: A Need for Periodic Updates of National Reference Values’ in Advs Exp. Medicine, Biology), better living conditions during maturation should mean better starting lung functions. That my study is not correct or the Polish population is ‘strange’ might be another explanation; however data for the Swedish population presented 7 year later by Jonas Brisman et al. in Clin Physiol Funct Imaging. 2017 Nov;37(6):640-645. doi: 10.1111/cpf.12349 are very similar. Thus, could anybody explain this paradox?
It could be due to the fact that those who lose the most FEV1 with age are most likely to not be sampled in the cross sectional study, due to death or severe illness
Hence the remainder will have less loss and the rate of decline will be less, compared to a longitudinal study of the same individuals with good follow up. .
Dear Tomasz Golczewski , I've gone through yours' and Thomas et al's papers. Following are my observations those may be the causes of so wide variations in findings: 1) Your one was cross-sectional, later one is longitudinal data; 2) your one: prospective, later one: a cohort study (having some scopes of deviations); 3) possibility of volunteer bias in the recent one; 4) inaccessibility to unpublished individual participant data; 5) availability and awareness of improved healthcare system; 6) changes of lifestyle in later one because populations had comorbidities like HTN, DM (dieting, exercise, physiotherapy, counseling, etc); 7) environmental changes. These are my views, quite nominal; but may have some impact on the fact findings. Regards- Rabiul
@James Leigh :
The cross-sectional analyses I know were performed to determine reference spirometric values. Therefore, only asymptomatic subjects were taken into account in each age groups. Thus, you may be right; however only partly right because in the article by Thomas et al. (that compares data from other papers) also only asymptomatic subjects were taken into account, which is mentioned in several places of this paper, e.g., in the abstract: ‘[…] asymptomatic adults without known respiratory disease.’.
I’ll admit: my question has been a kind of provocation: I would like to start a discussion on differences in lung functions between generations; according to my knowledge few authors discuss this problem (in fact, I know only me :). I suppose that older generations were more physically active, and lung functions depend on this activity. Present young generations are active but mostly in Internet (and previously, it was ‘TV activity’). For example, when me and my friends were children, we spent free time solely playing various physical games, riding on bicycles, etc. My grandchildren and their friend … don't get me started.
However, maybe another explanation of the difference between the FEV1 decline rates is more correct.
@Rabiul Alam
Dear Rabiul Alam, all items you listed are right (except this stated my analysis as prospective; it was retrospective) but they do not explain the difference that I mentioned. See, please, my response to James Leigh. What you think about differences between generations?