Umeå University's logo

Background: Chronic airway obstruction (CAO) and restrictive spirometry pattern (RSP) are associated with mortality, but sex-specific patterns of all-cause and specific causes of death have hardly been evaluated.

Objectives: To study the possible sex-dependent differences of all-cause mortality and patterns of cause-specific mortality among men and women with CAO and RSP, respectively, to that of normal lung function (NLF).

Design: Population-based prospective cohort study.

Methods: Individuals with CAO [FEV1/vital capacity (VC) < 0.70], RSP [FEV₁/VC ⩾ 0.70 and forced vital capacity (FVC) < 80% predicted] and NLF (FEV₁/VC ⩾ 0.70 and FVC ⩾ 80% predicted) were identified within the Obstructive Lung Disease in Northern Sweden (OLIN) studies in 2002–2004. Mortality data were collected through April 2016, totally covering 19,000 patient-years. Cox regression and Fine–Gray regression accounting for competing risks were utilized to estimate hazard ratios (HRs) with 95% confidence intervals (CIs) adjusted for age, body mass index, sex, smoking habits and pack-years.

Results: The adjusted hazard for all-cause mortality was higher in CAO and RSP than in NLF (HR, 95% CI; 1.69, 1.31–2.02 and 1.24, 1.06–1.71), and the higher hazards were driven by males. CAO had a higher hazard of respiratory and cardiovascular death than NLF (2.68, 1.05–6.82 and 1.40, 1.04–1.90). The hazard of respiratory death was significant in women (3.41, 1.05–11.07) while the hazard of cardiovascular death was significant in men (1.49, 1.01–2.22). In RSP, the higher hazard for respiratory death remained after adjustment (2.68, 1.05–6.82) but not for cardiovascular death (1.11, 0.74–1.66), with a similar pattern in both sexes.

Conclusion: The higher hazard for all-cause mortality in CAO and RSP than in NLF was male driven. CAO was associated with respiratory death in women and cardiovascular death in men, while RSP is associated with respiratory death, similarly in both sexes.

Rationale: Spirometry is essential for diagnosis and assessment of prognosis in COPD.

Objectives: To identify FEV1 trajectories and their determinants, based on annual spirometry measurements among individuals with and without airway obstruction. Furthermore, to assess mortality in relation to trajectories.

Methods: In 2002-04, individuals with airway obstruction (AO) (FEV1/VC<0.70, n=993) and age- and sex-matched non-obstructive (NO) referents were recruited from population-based cohorts. Annual spirometries until 2014 were utilized in joint-survival Latent Class Mixed Models to identify lung function trajectories. Mortality data were collected during 15 years of follow-up.

Results: Three trajectories were identified among the AO-cases and two among the NO referents. Trajectory membership was driven by baseline FEV1%predicted (%pred) in both groups and additionaly, pack-years in AO and current smoking in NO. Longitudinal FEV1%pred level depended on baseline FEV1%pred, pack-years and obesity. The trajectories were distributed: 79.6% T1AO FEV1-high with normal decline, 12.8% T2AO FEV1-high with rapid decline, and 7.7% T3AO FEV1-low with normal decline (mean 27, 72 and 26 mL/year) among AO-individuals, and 96.7% T1NO FEV1-high with normal decline and 3.3% T2NO FEV1-high with rapid decline (mean 34 and 173 mL/year) among referents. Hazard for death was increased for T2AO (HR1.56) and T3AO (HR3.45) vs. T1AO, and for T2NO (HR2.99) vs. T1NO.

Conclusions: Three different FEV1 trajectories were identified among those with airway obstruction and two among the referents, with different outcomes in terms of FEV1-decline and mortality. The FEV1 trajectories among airway obstructive and the relationship between low FVC and trajectory outcome are of particular clinical interest.

Background: In 2019, WHO estimated COPD to be the third leading cause of death in the world. However, COPD is probably underestimated as cause of death due to the well-known under-diagnosis.

Aim: To evaluate the proportion of and factors associated with COPD recorded as cause of death in a long-term follow-up of a population-based COPD cohort.

Methods: The study population includes all individuals (n = 551) with COPD defined as chronic airway obstruction (post-bronchodilator FEV1/FVC<0.70) + respiratory symptoms identified after re-examinations of four population-based cohorts. Mortality and underlying or contributing cause of death following ICD-10 classification were collected from the Swedish National Board of Health and Welfares register from date of examination in 2002–04 until 2016.

Results: The study sample consisted of 32.3% GOLD 1, 55.9% GOLD 2, and 11.8% GOLD 3–4. The mean follow-up time was 10.3 (SD3.77) years and the cumulative mortality 45.0%. COPD (ICD-10 J43-J44) was recorded on 28.2% (n = 70) of the death certificates (11.1%, 25.7% and 57.1% by GOLD stage), whereof n = 35 had COPD recorded as underlying and n = 35 as contributing cause of death. To have COPD recorded as cause of death was independently associated with ex- and current smoking and a self-reported physician diagnosis of COPD, while male sex, overweight/obesity and higher FEV1% of predicted associated with the absence.

Conclusions: COPD was largely underreported cause of death. Even among those with severe/very severe disease, COPD was only mentioned on 57.1% of the death certificates.

Background: COPD has increased in prevalence worldwide over several decades until the first decade after the millennium shift. Evidence from a few recent population studies indicate that the prevalence may be levelling or even decreasing in some areas in Europe. Since the 1970s, a substantial and ongoing decrease in smoking prevalence has been observed in several European countries including Sweden. The aim of the current study was to estimate the prevalence, characteristics and risk factors for COPD in the Swedish general population. A further aim was to estimate the prevalence trend of COPD in Northern Sweden from 1994 to 2009.

Methods: Two large random population samples were invited to spirometry with bronchodilator testing and structured interviews in 2009-2012, one in south-western and one in northern Sweden, n = 1839 participants in total. The results from northern Sweden were compared to a study performed 15 years earlier in the same area and age-span. The diagnosis of COPD required both chronic airway obstruction (CAO) and the presence of respiratory symptoms, in line with the GOLD documents since 2017. CAO was defined as post-bronchodilator FEV1/FVC < 0.70, with sensitivity analyses based on the FEV1/FVC < lower limit of normal (LLN) criterion.

Results: Based on the fixed ratio definition, the prevalence of COPD was 7.0% (men 8.3%; women 5.8%) in 2009-2012. The prevalence of moderate to severe (GOLD >= 2) COPD was 3.5%. The LLN based results were about 30% lower. Smoking, occupational exposures, and older age were risk factors for COPD, whereof smoking was the most dominating risk factor. In northern Sweden the prevalence of COPD, particularly moderate to severe COPD, decreased significantly from 1994 to 2009, and the decrease followed a decrease in smoking.

Conclusions: The prevalence of COPD has decreased in Sweden, and the prevalence of moderate to severe COPD was particularly low. The decrease follows a major decrease in smoking prevalence over several decades, but smoking remained the dominating risk factor for COPD.

Background: Chronic obstructive pulmonary disease (COPD) is common. The estimated prevalence is about 10% among adults, but varies largely dependent on the major risk factors age and smoking. Under-diagnosis of COPD is substantial and is related to disease severity. Thus, subjects with mild to moderate COPD are underrepresented in medical registers among health care providers as well as in national registers. Post- bronchodilator (BD) spirometry is mandatory for the diagnosis of COPD, but not sufficient to assess and manage COPD. Phenotyping based on spirometry and clinical manifestations can make it easier to apply individual assessment of subjects with COPD. COPD is a systemic disease with pulmonary and extra-pulmonary manifestations and comorbidities are common. Comorbidities most probably contribute to the observed increased mortality among subjects with COPD, however, the impact of comorbidities on mortality and causes of death among subjects with mild to moderate COPD is unclear. Furthermore, there seems to be sex-dependent differences with regard to susceptibility to risk factors, clinical manifestation and outcomes.

Aim: The overall aim of this thesis was to identify and characterize clinical relevant COPD phenotypes in population-based studies, using spirometry together with clinical characteristics such as respiratory symptoms, exacerbations, and comorbidities, and their impact on mortality and further, also cause of death.

Methods: This thesis is based on data from the Obstructive Lung Disease in Northern Sweden (OLIN) COPD study. The study population was recruited in the years 2002-2004, when all 993 individuals with (FEV1/VC<0.70) were identified after examinations of population-based cohorts, together with age- and sex-matched non-obstructive referents (n=in total 1,986). In this thesis, cross-sectional data from recruitment were used together with mortality data from the Swedish Tax Agency from the date of recruitment in 2002-2004 and onwards. Data on cause of death was collected from the Swedish National Board for Health and Welfare register for all deaths until 31 December 2015. Spirometry was used to identify the following spirometric groups, in paper I: Non-COPD (FEV1/VC≥0.70); COPD (pre- BD FEV1/VC<0.70); in paper II: Non- obstructive (FEV1/VC≥0.70), Pre- not post-BD obstructive (pre- not post-BD FEV1/VC<0.70); COPD (post-BD FEV1/VC<0.70); In paper III: Normal Lung Function (NLF, FEV1/VC≥0.7 & FVC≥80% predicted), COPD (post BD FEV1/VC<0.70) and Lower Limit of Normal COPD (LLN-COPD, the LLN criterion applied among those with COPD); in paper IV: NLF and COPD defined as in paper III, and Restrictive Spirometric pattern (RSP, FEV1/VC≥0.70 & FVC<80% predicted). The OLIN-COPD study and collection of data on causes of death were approved by the regional ethical committee at Umeå University.

Results: Paper I: Subjects with COPD had more productive cough than non-COPD, and men more than women. Productive cough increased the risk for exacerbations in COPD and non-COPD and productive cough was associated with worse survival in both groups. In adjusted models (HR;95%CI) the increased risk for death associated with productive cough among those with COPD persisted (1.48;1.13-1.94) when compared with non-COPD without productive cough, significantly so also among men with COPD (1.63;1.17-2.26), but not among women (1.23;0.76-1.99).

Paper-II: Pre-BD spirometry misclassified every fourth subject as having COPD. Subjects with pre- but not post-BD obstruction were similar to subjects with COPD regarding reported ‘any respiratory symptoms’, asthma before the age of 40, exacerbations, and comorbidities. The cumulative mortality among subjects with pre- not post-BD obstruction was similar to among subjects in the non-obstructive group, still, the survival was better than among those with COPD. The increased risk for death for COPD persisted also in an adjusted model (1.24; 1.04-1.49) when compared with the non-obstructive group, and the pattern was similar among men and women (1.27; 1.00-1.60 and1.24; 0.92-1.13).

Paper III: Men with COPD had more CVD and DM compared to women, while anxiety/depression (A/D) was more common among women than men in all spirometric groups. Men had a higher cumulative mortality than women in all groups. However, CVD seemed to have a greater impact on mortality among women than men, while anxiety/depression increased the risk for death similarly in both sexes. The use of the LLN criterion did not change the observed pattern.

Paper IV: CVD was the most common cause of death in all spirometric groups, NLF, RSP and COPD, followed by cancer. Those with COPD and RSP had a similar and higher cumulative mortality than those with NLF. RSP and COPD had an increased risk for CVD death and respiratory death, independent of age, sex, smoking habits and BMI-category, however, the increased risk for CVD death did not reach statistical significance in RSP. In all the groups, the risk for deaths due to cancer was similar, however, lung cancer was more common in COPD than in NLF and RSP. The pattern was fairly similar among men and women. 

Conclusions: Simple diagnostic procedures like history of respiratory symptoms, exacerbations, and comorbidity can, together with spirometry, contribute with important clinical classification of prognostic importance. Productive cough increased the risk for exacerbations in both COPD and non-COPD. The highest risk for exacerbations and death was observed among subjects with COPD and productive cough. It was impossible to distinguish COPD from those with pre- not post-BD obstruction based on the history of respiratory symptoms, asthma, exacerbations and comorbidities. Still, COPD was associated with an increased risk for death while pre- not post-BD obstruction had better survival than COPD but similar as non-obstructive. There were sex-dependent differences regarding comorbidities and mortality. CVD was less common among women but had a greater impact on mortality compared to among men while A/D, less common among men, increased the risk for death similarly in both sexes. CVD and cancer were the most common causes of death in all spirometric groups. RSP had a similar and higher mortality as COPD when compared with NLF. The risk for cancer-related death was similar in all groups, while the results indicated that COPD and RSP had an increased risk for CVD and respiratory death.

Introduktion: Kronisk obstruktiv lungsjukdom (KOL) är en folksjukdom som förekommer hos cirka 10% i den vuxna befolkningen. Förekomsten (prevalensen) varierar dock och är framförallt beroende av ålder och rökning. Underdiagnostiken av KOL är betydande och är relaterad till sjukdomens svårighetsgrad. Upp till 70 % av dem med KOL är inte identifierade inom hälso- och sjukvården. Även feldiagnostik är vanligt, nästan lika vanlig som underdiagnostik. Personer med lindrig till måttligt svår KOL är särskild sårbara för under- och feldiagnostik. Det förefaller finnas betydelsefulla könsskillnader vad gäller känslighet för riskfaktorer som rökning samt kliniska sjukdomsuttryck och prognos vid KOL. Det krävs lungfunktionsmätning (spirometri) efter luftrörsvidgande läkemedel (bronkdilatation (BD)) för att påvisa kvarstående luftrörsförträngning (luftvägsobstruktion) vid diagnostik av KOL. Spirometri görs i allt för liten omfattning hos personer med misstänkt KOL, och vid undersökning med spirometri, det är inte ovanlig att den genomförs utan luftrörsvidgande mediciner. Det krävs mer än enbart spirometri för korrekt handläggning (utredning och behandling) av personer med KOL. Kliniska sjukdomsuttryck (fenotyper), baserade på spirometri och kliniska uppgifter om exempelvis luftvägssymtom och uppgift om försämringsperioder (exacerbationer) kan underlätta bedömning av KOL och tillämpning av individualiserad handläggning vid KOL. Enkel klinisk fenotypning skulle vara tillämpningsbar i det daglig kliniska arbetet. Samsjuklighet (komorbiditeter) är vanliga vid KOL och bland dem är hjärt- och kärlsjukdomar vanligast. Komorbiditet bidrar förmodligen till den ökade dödligheten som har observerats hos personer med KOL, men det är oklart vilken betydelse komorbiditeter har för mortalitet framförallt bland personer med lindrig och måttligt svår KOL samt vilka som är de vanligaste dödsorsakerna.

Syfte: Det övergripande syftet med denna avhandling var att identifiera och karakterisera kliniska fenotyper av KOL i en befolkningsbaserad studie med hjälp av spirometri och enkla kliniska uppgifter såsom förekomst av luftvägssymtom, exacerbationer och komorbiditet, samt värdera deras betydelse i relation till mortalitet och dödsorsaker. 

Metod: Avhandlingen utgår från en befolkningsstudie i norra Sverige där en långtidsuppföljning av personer med och utan KOL pågår inom ramen för det epidemiologiska forskningsprogrammet Obstruktiv Lungsjukdom i Norrbotten  (OLIN) studierna; OLINs KOL-studie. Efter kliniska undersökningar av populationsbaserade kohorter under åren 2002-2004 identifierades alla (n=993) individer med luftvägsobstruktion (FEV₁/VC <0.70) tillsammans med lika många ålders- och könsmatchade referenspersoner utan luftvägsobstruktion, totalt n=1986. Tvärsnittsdata från undersökningstillfället vid rekryteringen under åren 2002-04 har använts i samtliga fyra delarbeten, tillsammans med mortalitets data från Skatteverkets folkbokföringsregister. Dödsorsakuppgifter har erhållits från Socialstyrelsens dödsorsaksregister för alla som avlidit fram till den 31 december 2015. Följande spirometriska grupper har identifierats; delarbete I: Ej KOL (FEV₁/VC≥0,70); KOL (pre-BD FEV₁/VC<0,70), delarbete II: Icke obstruktiv; obstruktiv före men inte efter BD (pre- not post BD obstruktiv); KOL (post-BD FEV₁/VC<0.70), delarbete III: Normal lungfunktion (NLF, FEV₁/VC≥ 0,70 & FVC≥80% av förväntat värde), KOL (post-BD FEV₁/VC<0,70) och KOL enligt Lower limit of normal (KOL enligt LLN; femte percentilen av normalvärde för post-BD FEV₁/VC<0,70), delarbete IV: NLF och KOL definierat som i delarbete III, Restriktivt spirometriskt mönster (RSP, FEV₁/VC≥0,70 & FVC<80% av förväntat värde). KOL-studien och inhämtande av uppgifter om dödsorsaker från dödsorsaksregistret är godkända av den Etiska kommittén vid Umeå universitet. 

Resultat: Delarbete I: Produktiv hosta var mer vanlig förekommande bland personer med KOL än bland dem som inte hade KOL, och framförallt bland män. Produktiv hosta ökade risken för exacerbationer hos både personer med och utan KOL och i båda könen.  Produktiv hosta var även förknippat med sämre överlevnad bland både dem med och utan KOL.

Den ökade risken för död (HR;95%CI) bland dem med KOL och produktiv hosta jämfört med dem utan KOL och utan produktiv hosta, kvarstod oberoende av kön, ålder, rökvanor och förekomst av hjärtsjukdom (1,48;1,13-1,94). Den förhöjda risken var signifikant bland män med KOL (1,63;1,17–2,26), men inte bland kvinnor med KOL (1,23;0,76–1,99).  

Delarbete II: Spirometri utan bronkodilatation medför att varje fjärde person fel-klassificeras som obstruktiv. Personer med luftvägsobstruktion före men inte efter BD kunde inte skiljas från personer med KOL, vad gäller förekomsten av några luftvägssymtom, astmadiagnos före 40 års ålder, exacerbationer och förekomsten av komorbiditeter. Dock hade de hade bättre överlevnad i jämförelse med dem med KOL och överlevnaden var på samma nivå som hos dem utan KOL. De med KOL hade 24 % ökad risk för död, jämfört med dem utan KOL, oberoende av ålder, kön, rökvanor, BMI och komorbiditet och detta riskmönster var likartat bland män och kvinnor.  

Delarbete III: Hjärtkärlsjukdomar var en större riskfaktor för död bland kvinnor, trots att de var mer vanligt förekommande bland män. Ångest/depression var vanligare hos kvinnor än hos män i alla spirometriska grupper, men det var en lika stark riskfaktor för död bland män och kvinnor. Män hade dock genomgående sämre överlevnad än kvinnor i alla spirometriska grupper. Mönstret var likartat också när KOL definierades enlig LLN-kriterierna.  

Delarbete IV: Hjärtkärlsjukdomar var den vanligaste dödsorsaken i alla spirometriska grupper, NLF, RSP och KOL, följt av cancersjukdomar. Personer med KOL och RSP hade en likartad och ökad mortalitet, jämfört med dem i NLF-gruppen. Personer med RSP och KOL hade en ökad risk för att dö av sjukdomar i andningsorganen jämfört med dem i NLF-gruppen, oberoende av ålder, kön, rökvanor och BMI. De med KOL och RSP hade också en ökad risk för död i hjärtkärlsjukdomar där dock riskökningen för dem i RSP inte nådde statistisk signifikans. Risk att dö till följd av cancersjukdomar var lika stor i alla spirometriska grupper. I analyser genomförda separat bland män och kvinnor var mönstren likartade. 

Slutsatser: Enkla kliniska uppgifter om luftvägssymptom, exacerbationer och komorbiditet kan tillsammans med spirometri användas vid klinisk fenotypning av prognostisk betydelse hos män och kvinnor med KOL.

Produktiv hosta ökade risken för exacerbationer hos män och kvinnor, både med och utan KOL.  Den största risken för exacerbationer och död observerades hos personer med samtidig KOL och produktiv hosta.

Spirometri utan bronkodilatation orsakade en betydande felklassificering av KOL. Uppgift om luftvägssymtom, exacerbationer och komorbiditet kan inte ersätta spirometri efter bronkodilatation för att ställa rätt diagnos. Dessutom har rätt diagnos prognostiskt betydelse, då endast de med KOL hade en ökad risk för död jämfört med dem utan KOL.

Trots att hjärtkärlsjukdomar var mindre vanliga bland kvinnor än bland män, föreföll de vara en starkare riskfaktor för död bland kvinnor. Ångest/depression var vanligare bland kvinnor än bland män, men ökade risken för död likvärdigt i båda könen. Hjärtkärlsjukdomar och cancer var de vanligaste dödsorsakerna i alla spirometriska grupper. Personer med RSP och KOL hade en likvärdig och högre mortalitet än dem med NLF. Resultaten indikerar att de med KOL och RSP hade en ökad risk för död av hjärtkärlsjukdomar och sjukdomar av andningsorganen jämfört med NLF, medan risken att dö av cancer var densamma i alla grupperna.

Background: Comorbidities probably contribute to the increased mortality observed among subjects with chronic obstructive pulmonary disease (COPD), but sex differences in the prognostic impact of comorbidities have rarely been evaluated in population-based studies. The aim of this study was to evaluate the impact of common comorbidities, cardiovascular disease (CVD), diabetes mellitus (DM), and anxiety/depression (A/D), on mortality among men and women with and without airway obstruction in a population-based study.

Methods: All subjects with airway obstruction [forced expiratory volume in 1 second (FEV1)/(forced) vital capacity ((F)VC) <0.70, n = 993] were, together with age- and sex-matched referents, identified after examinations of population-based cohorts in 2002-2004.

Spirometric groups: normal lung function (NLF) and COPD (post-bronchodilator FEV1/(F)VC <0.70) and additionally, LLN-COPD (FEV1/(F)VC <lower limit of normal). Mortality data was collected until December 2015. Results: In COPD, the prevalence of CVD and DM was higher in men, whereas the prevalence of A/D was higher in women. The cumulative mortality was significantly higher in COPD than NLF, and higher in men than women in both groups. Among women with COPD, CVD and A/D but not DM increased the risk of death independent of age, body mass index, smoking habits, and disease severity, whereas among men DM and A/D but not CVD increased the risk for death. When the LLN criterion was applied, the pattern was similar.

Conclusion: There were sex-dependent differences regarding the impact of comorbidities on prognosis in COPD. Even though the prevalence of CVD was higher in men, the impact of CVD on mortality was higher in women, and despite higher prevalence of A/D in women, the impact on mortality was similar in both sexes. The reviews of this paper are available via the supplemental material section.

Comorbidities contribute to the increased mortality observed among subjects with COPD, but the prognostic impact and possible sex differences have rarely been evaluated in population-based studies.

Aim: To evaluate the impact of common comorbidities; cardiovascular disease (CVD), diabetes mellitus (DM) and anxiety/depression (A/D), on mortality in a population-based study of subjects with (COPD) and without airway obstruction.

Methods: All subjects with airway obstruction (FEV1/(F)VC<0.70, n=993), were, together with age- and sex matched referents, identified after examinations of population-based cohorts in 2002-04. Spirometric groups: Normal Lung Function (NLF), COPD; post- bronchodilator fixed ratio (GOLD) and lower limit of normal (LLN). Mortality data were collected until December 2015.

Results: The cumulative mortality was significantly higher in GOLD-COPD than NLF, and higher in men than women in both groups. CVD, DM and A/D independently increased the risk for death (Hazard Ratio; 95% CI, 1.50-1.59; 1.07-2.11) in GOLD-COPD when adjusted for age, sex, smoking habits, BMI and FEV1% predicted, while in NLF A/D (1.54; 1.03-2.30) but not CVD (1.20; 0.87-1.65) or DM (1.46; 0.95-2.26). Among women with GOLD-COPD, CVD and A/D but not DM increased the risk for death, while among men DM and A/D, but not CVD. When the LLN-criterion was applied, the significantly increased risk for death associated with comorbidities remained among men, but not among women.

Conclusion: CVD, DM and A/D increased the risk for death in GOLD-COPD, but there seems to be sex-dependent differences in prognosis associated with comorbidities, also in relation to different spirometric criteria for COPD.