Sex Differences in Survival of Oxygen-dependentPatients with Chronic Obstructive Pulmonary DiseaseMaria-Christina L. Machado, Jerry A. Krishnan, Sonia A. Buist, Andrew L. Bilderback, Guilherme P. Fazolo,Michelle G. Santarosa, Fernando Queiroga, Jr., and William M. Vollmer

State Public Hospital of Sao Paulo; Respiratory Division of Federal University of Sao Paulo, Sao Paulo, Brazil; Johns Hopkins University,Baltimore, Maryland; Oregon Health and Science University; and Center for Health Research, Portland, Oregon

Rationale: Chronic obstructive pulmonary disease (COPD) is a lead-ing cause of death worldwide. The prevalence of COPD is risingamong women and is approaching that of men, but it is not knownif sex affects survival.Objectives: To measure the survival differences between men andwomen with oxygen-dependent COPD.Methods: We conducted a 7-yr prospective cohort study of 435outpatients with COPD (184 women, 251 men) referred for long-term oxygen therapy (LTOT) at two respiratory clinics in Sao Paulo,Brazil. Baseline data were collected on enrollment into oxygen ther-apy, when patients were clinically stable.Measurements: We examined the effect of sex on survival usingKaplan-Meier survival curves, and then used Cox proportional haz-ards models to control for potential confounders.Main Results: In unadjusted analyses, we observed a nonsignificanttrend toward increased mortality for women (hazard ratio, 1.28;95% confidence interval, 0.98–1.68; p � 0.07). After accounting forpotential confounders (age, pack-years smoked, PaO2, FEV1, bodymass index), females were at a significantly higher risk of death(hazard ratio, 1.54; 95% confidence interval, 1.15–2.07; p � 0.004).Other independent predictors of death were lower PaO2 (p � 0.001)and lower body mass index (p � 0.05).Conclusions: Among patients with COPD on LTOT, women weremore likely to die than men.

Keywords: sex differences; chronic obstructive pulmonary disease,hypoxemic; survival

Chronic obstructive pulmonary disease (COPD) is one of themost important causes of morbidity and mortality worldwide,being the fourth leading cause of death in the United States andthe fifth in Brazil (1–3). Mortality is especially high in patientswith severe disease. Although COPD has been historically morecommon in men, there are data to suggest that women may beat increased risk of developing COPD (4). Other studies suggestthat women may have more severe COPD and greater COPD-associated mortality (5–8).

In light of this hypothesized increased susceptibility forwomen, the recent data suggesting that the sex gap in COPDprevalence is decreasing is of particular concern (5–8). We reportresults of a 7-yr follow-up of outpatients with COPD referredfor long-term oxygen therapy (LTOT). We hypothesized that,in patients with oxygen-dependent COPD, women have a highermortality rate than men. Some of the results of this study havebeen previously reported in the form of an abstract (9).

(Received in original form July 8, 2005; accepted in final form June 14, 2006 )

Correspondence and requests for reprints should be addressed to Maria-ChristinaLombardi Machado, Rua do Simbolo, 16/apto 21, Morumbi. Sao Paulo–SP,05713-570 Brazil. E-mail: mchrismachado@pneumo.epm.br

Am J Respir Crit Care Med Vol 174. pp 524–529, 2006Originally Published in Press as DOI: 10.1164/rccm.200507-1057OC on June 15, 2006Internet address: www.atsjournals.org

METHODS

Study Subjects

We conducted a prospective cohort study of 435 outpatients with COPD(184 women, 251 men) enrolled in the LTOT program at two respiratoryclinics affiliated with two hospitals in Sao Paulo, Brazil, between January1996 and January 2003. Criteria recommended by the Global Initiativefor Chronic Obstructive Lung Diseases (GOLD) (1) and the BrazilianThoracic Society (BTS) (2) were used to define COPD. Patients wereeligible for this study if they had a post-bronchodilator ratio of FEV1

to FVC less than 0.70 and were enrolled in the LTOT program for aminimum of 6 mo. Current smokers (based on patient self-report andphysician judgment) were not eligible for the LTOT offered in therespiratory clinics, and hence are not included in the study cohort. Inaddition, patients with a diagnosis of lung cancer before initiating LTOTwere excluded from this study.

Participants were managed according to the GOLD/BTS recom-mendations, including LTOT therapy. Due to their disease severity,none of the study participants was enrolled in a pulmonary rehabilita-tion program. Baseline data were collected when patients were clinicallystable (i.e., no exacerbation in the past 30 d). All study participantswere followed until January 2003 or death. The average observationtime was 27 mo (range, 6–80 mo). No patients were lost to follow-up. Thestudy was reviewed and approved by the institutional review boards ofboth hospitals.

Measurements

On enrollment, the following data were collected: sex, age, lungfunction, PaO2 and PaCO2 in mm Hg measured on room air by thepH and blood gas analyzer (instrument used: ABL330; Radiometer,Copenhagen, Denmark), body mass index (BMI) in kg/m2, the numberof pack-years of cigarette smoking, the number of hospitalizations dueto respiratory causes in the previous 12 mo, and the Charlson comorbid-ity index (10). Post-bronchodilator spirometry was performed accordingto American Thoracic Society criteria (11) and used to measure FEV1

and FVC, and their ratio. American Thoracic Society recommendationswere followed to calculate the post-bronchodilator percent of predictedFEV1 (FEV1, % predicted). The Charlson comorbidity index was calcu-lated using information on comorbid conditions, as assessed by a reviewof clinical records (hospital data and clinic charts); as it includes COPD,the minimum index score for everyone was 1.

Statistical Analysis

We used Student’s t test and the Pearson �2 statistic to compare differ-ences in baseline characteristics between men and women. In theseanalyses, BMI (� 18.5, 18.5–24.9, 25–29.9, � 30 kg/m2), the number ofhospitalizations (0, 1, 2, � 3), and the number of comorbid conditionsweighted by the Charlson comorbidity index (1, 2, � 3) were analyzedas ordinal, categorical variables to facilitate interpretation. Kaplan-Meier survival curves were compared using the log rank test. We devel-oped a multivariable Cox proportional hazards model to comparesurvival between men and women after adjusting for potential con-founders. Those variables significant at the p level less than 0.20 inunivariate analyses were included in the multivariate model. We alsoevaluated the possibility of interactions between sex and other indepen-dent predictors of mortality in the multivariable model. All p valuesare two-sided. The term significant refers to a p value less than 0.05.Computations were performed using STATA, version 8.2 (Stata Corpo-ration, College Station, TX [12]).

Machado, Krishnan, Buist, et al.: Sex Differences in COPD Survival 525

RESULTS

At the baseline visit, the study cohort ranged in age from 35 to85 yr (mean � 66.6 yr), the average FEV1 was 31.4% predicted,and the average PaO2 was 51.7 mm Hg (Table 1). Women weresignificantly younger and had less lifetime exposure to cigarettesmoking than men, despite having similar levels of FEV1, PaO2,and PaCO2. Over 75% of both men and women reported one ormore hospitalizations in the 12 mo before enrollment.

Over two-thirds of participants (68.1%) had at least oneclinically significant comorbid condition other than COPD (i.e.,Charlson comorbidity index � 2). Men had a higher Charlsoncomorbidity index than women, although differences were notsignificant.

Results of the Kaplan-Meier analysis showed a nonsignificanttrend toward increased mortality among women compared withmen (Figure 1). By 48 mo, there were relatively few patients inthe cohort, so the survival curve estimates begin to grow unstablepast this point. In the univariate Cox proportional hazards mod-els (Table 2), a lower PaO2, lower FEV1, greater number of pack-years smoked, and lower BMI were all significantly associatedwith higher mortality. In the multivariable (i.e., adjusted) analysis,women had a significantly higher risk of death compared withmen (hazard ratio, 1.54; 95% confidence interval, 1.15–2.07) afteradjusting for age, pack-years smoked, PaO2, FEV1, and BMI. Inthis model, lower PaO2 and lower BMI were also significantindependent predictors of mortality (Table 2). None of the base-line patient characteristics significantly modified the associationbetween sex and survival in the multivariable model (tests forinteraction with age [p � 0.28], pack-years smoked [p � 0.25],PaO2 [p � 0.74], FEV1 [p � 0.22], BMI categories [p � 0.64],and Charlson comorbidity index [p � 0.29]).

DISCUSSION

In this 7-yr prospective cohort study of patients with COPDreferred for LTOT, women were younger and reported fewerpack-years of cigarette smoking then men, yet had similar impair-ment in lung function and oxygenation. We found that womenhad a 54% increase in the risk of death after initiating LTOTcompared with men.

TABLE 1. BASELINE PATIENT CHARACTERISTICS

Total Male FemaleCharacteristics (n � 435) (n � 251) (n � 184) p Value

Age, yr 66.6 � 7.6 69.3 � 7.2 62.9 � 6.5 � 0.0001Pack-years smoked 69.6 � 30.1 72.5 � 30.1 65.7 � 29.8 � 0.01PaO2, mm Hg 51.7 � 5.5 51.7 � 5.3 51.8 � 5.8 0.90PaCO2, mm Hg 47.0 � 5.7 47.0 � 5.9 46.9 � 5.4 0.80FEV1% predicted 31.4 � 8.0 31.4 � 8.4 31.4 � 7.5 0.96BMI, kg/m2

� 18.5 36 (8.3) 19 (7.6) 17 (9.2)18.5–24.9 222 (51.0) 135 (53.8) 87 (47.3) 0.3625–29.9 125 (28.7) 65 (25.9) 60 (32.6)� 30.0 52 (12.0) 32 (12.8) 20 (10.9)

No. of hospitalizations0 73 (16.8) 41 (16.3) 32 (17.4)1 130 (29.9) 72 (28.7) 58 (31.5) 0.872 129 (29.7) 76 (30.3) 53 (28.8)�3 103 (23.7) 62 (24.7) 41 (22.3)

Charlson comorbidity index1 139 (32.0%) 87 (34.7) 52 (28.3)2 183 (42.1%) 94 (37.4) 89 (48.4) 0.07� 3 113 (26.0%) 70 (27.9) 43 (23.4)

Definition of abbreviation: BMI � body mass index.Data for continuous variables expressed as mean (SD); other data expressed as n (%). p value refers to comparisons of patient

characteristics by sex.

We know that the life expectancy is poor in patients withadvanced COPD, particularly when FEV1 is less than 1 L, PaO2is less than 55 mm Hg, and hypercapnia or pulmonary hyperten-sion is present (13–18). The only therapeutic regimen that hasbeen shown to improve the life expectancy in these patientsis oxygen therapy. Two well-known studies, Medical ResearchCouncil (MRC) (19) and Nocturnal Oxygen Therapy Trial(NOTT) (20), found that oxygen therapy improved survival inpatients with COPD who were markedly hypoxemic.

Although the differences in survival between men and womenin our unadjusted analysis failed to reach statistical significance(p � 0.07), after adjusting for potentially confounding factors(age, severity of lung disease, FEV1, PaO2, lifetime history ofpack-years smoked, and BMI), women on LTOT were morelikely to die than men. Interestingly, we found that men andwomen exhibited similar survival rates during the initial follow-up period; differences in survival became more apparent onlyafter approximately 3 yr of follow-up. The clinical managementfor COPD was similar for both groups, and was based on GOLD/BTS guidelines. We do not know why survival differences oc-curred largely during the second half of the follow-up period;further research is needed to confirm our findings and identifypotential explanations.

Our findings that women with COPD on LTOT fare worsethan men are consistent with the results of some previous studies(26, 27), but not with those of others (19–25, 28–31). We specu-late that differences in patient populations and analytic ap-proaches may help explain the discrepancy between our resultsand those of previous studies. For example, we used GOLD/BTS criteria to define COPD (including the requirement forairflow obstruction based on spirometry) and to identify patientswith indications for LTOT. By contrast, the definition of COPDwas based on clinical criteria alone in one study (21), two otherstudies did not require evidence of airflow obstruction (22, 23),and a fourth study (25) included patients with respiratory condi-tions other than COPD (e.g., tuberculosis, pulmonary fibrosis)who were eligible for LTOT. We excluded current smokers andpatients with preexisting lung cancer; it is unclear if these patientswere systematically excluded in previous studies as well. TheMRC clinical trial (19) excluded patients who were older than

526 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 174 2006

Figure 1. Differences in survival, men versus women (p �

0.07 [log-rank]). The number of males and females in thefollow-up period at Months 0 (baseline), 24, 48, and 72are shown below graph. Kaplan-Meier survival curvesfor males and females after initiating long-term oxygentherapy (LTOT).

70 yr, presented with another respiratory disorder, or had sys-temic hypertension and/or coronary arterial disease. The NOTT(20) study excluded patients if they were considered to be “toosick,” lived far from the hospital, or (as in the study by Cooperand colleagues [24]) were considered high risk for nonadherenceto LTOT. Finally, in contrast to previous analyses, we accountedfor several potential confounders, including pack-years of ciga-rettes smoked and BMI.

Studies have documented the prognostic value of low bodyweight in patients with COPD and in the general population(32–34). Celli and coworkers (35) studied a multidimensionalgrading system, BODE (B � body mass index, O � airwayobstruction, D � dyspnea, E � exercise capacity) that can predictrisk of death in patients with COPD. The BODE index, basedon four variables that can mark the degree of disease severity,

TABLE 2. RESULTS OF COX REGRESSION ANALYSES

Univariate Multivariate†

Characteristics HR (95% CI) p Value HR (95% CI) p Value

Female (vs. male) 1.28 (0.98–1.68) 0.07 1.54 (1.15–2.07) 0.004Age, per 10 yr 1.13 (0.94–1.37) 0.19 1.17 (0.95–1.45) 0.14Smoking, per 10 pack-years 1.06 (1.02–1.10) 0.005 1.04 (0.99–1.08) 0.09PaO2, per 10 mm Hg 0.51 (0.41–0.64) � 0.001 0.57 (0.45–0.72) � 0.001PaCO2, per 10 mm Hg 1.03 (0.81–1.30) 0.81 — —FEV1, per 10% predicted 0.82 (0.69–0.97) 0.02 0.91 (0.75–1.09) 0.33BMI, kg/m2

� 18.5 3.06 (1.63–5.73) � 0.001 2.22 (1.15–4.28) 0.0118.5–24.9 2.34 (1.36–3.99) 0.002 1.75 (1.01–3.06) 0.0425.0–29.9 1.39 (0.78–2.49) 0.26 1.23 (0.68–2.23) 0.49� 30.0 1.00 (—) —

No. of hospitalizations0 1.00 (—) —1 1.04 (0.68–1.59) 0.862 1.21 (0.80–1.83) 0.37� 3 or more 1.22 (0.79–1.88) 0.36

Charlson comorbidity index*1 1.00 (—) —2 1.09 (0.79–1.50) 0.61� 3 1.60 (0.70–1.44) 0.99

Definition of abbreviations: BMI � body mass index; CI � confidence interval; HR � hazard ratio.* By definition, all patients have a score of at least 1 due to their chronic obstructive pulmonary disease.† Limited to those variables with p � 0.20 in univariate analyses.

is better than FEV1 alone at predicting the risk of death fromany cause and from a respiratory cause among patients withCOPD. Gray-Donald and coworkers (36) studied the role of BMIin the prognosis of patients with severe COPD in a cohort ofCanadian patients, including those with hypoxemia, recruitedfor a clinical trial of negative-pressure ventilation. In the totalcohort, lower BMI and use of home oxygen were independentlyassociated with reduced survival. These reports are in agreementwith our results, which found that the worse hypoxemia andlower BMI (� 24.9 kg/m2) were independently associated withhigher mortality. Schools and coworkers (37) showed that,despite a BMI greater than 25 kg/m2 being an independent pre-dictor of mortality in COPD, the negative effect of low bodyweight can be reversed by appropriate therapy in some patientswith COPD. We recommend additional cohort studies using

Machado, Krishnan, Buist, et al.: Sex Differences in COPD Survival 527

standardized approaches to diagnose and manage patients withoxygen-dependent COPD to confirm our findings.

Results of previous studies suggest than women who smokehave a greater decrease in FEV1 compared with men who smoke,suggesting a possible increased susceptibility to developmentof COPD in women (38, 39). There may also be an increasedpropensity for women to develop bronchial hyperresponsivenesscompared with men (40–43), although data from the U.S. LungHealth Study (43) would suggest that the increased bronchialhyperresponsiveness in women smokers, compared with menwho smoke, may be largely due to smaller airways. These obser-vations are consistent with findings in the current study, wherewomen had similar levels of impaired lung function (measuredby FEV1, PaO2, and PaCO2) despite fewer pack-years of cigarettesmoking.

COPD affects many organ systems in addition to the lungs(44, 45). For example, individuals who smoke and developCOPD further increase their risk of cardiovascular disease (46).The concurrence of COPD with cardiovascular disease repre-sents more than the simultaneous presence of relatively commonconditions (47). Individuals with reduced FEV1 are at increasedrisk for atherosclerosis. Zureik and coworkers (48) studied 194healthy men free of coronary heart disease to determine therelationship between FEV1 and pulse wave velocity, a surrogatemeasurement for central arterial stiffness, endothelial dysfunc-tion, and atherosclerosis. They showed that reduced FEV1 wasassociated with increased pulse wave velocity, suggesting thatairways disease was an independent risk factor for central arterialstiffness. Several groups have reported on the relationship be-tween FEV1 and cardiovascular mortality (49–51). All of thesestudies showed that reduced FEV1 among female and maleadults, independent of established risk factors, such as cigarettesmoking or hypertension, is an important risk factor for cardio-vascular mortality and a reasonable marker for COPD in popula-tion-based studies.

There is substantial evidence linking oxidative stress (52–57)and airway inflammation (58–62) with COPD disease progres-sion and severity (63–65). For example, hydrogen peroxide,8-isoprostane, and lipid peroxides are elevated in the breath orserum of individuals with COPD (64). Similarly, an increasednumber of neutrophils, macrophages, and natural killer lympho-cytes in the airway wall are associated with more symptomaticCOPD and lower FEV1 (66). There is now considerable evidenceof both local and systemic oxidative stress in patients withCOPD. However, we found no studies comparing oxidativestress and airway inflammation in women and men with COPDand evaluating whether such differences help to explain theincreased susceptibility and severity of COPD among women.We speculate that the sex differences in mortality found in ourstudy will be better understood after further studies of systemicinflammation and oxidative stress in patients with severe COPD,with attention to possible sex differences.

This study has limitations that should be considered wheninterpreting our findings. Whereas our analyses accounted fora number of clinical factors on enrollment, we did not collectdata on adherence to LTOT during the follow-up period. Also,although current smokers were ineligible for enrolling in ourLTOT program, we did not collect objective data on smokingstatus (e.g., serum cotinine) on initiating LTOT therapy or duringthe follow-up period. It is therefore possible that confoundingdue to these factors (e.g., sex differences in adherence to LTOTor smoking status after initiating LTOT) may have contributedto our results. Another limitation is that our study had a limitednumber of participants with follow-up times greater than 48 mo(n � 67 patients—15% of the initial study cohort). If the follow-up is censored at 48 mo, then the adjusted hazard ratio for death

in women compared with men is 1.35 (95% CI, 1.00–1.83; p �0.050). These observations suggest that larger studies, in whichadditional patients are followed for more than 4 yr, are neededto obtain more precise comparisons of the risk of death betweenmen and women with COPD treated with LTOT.

CONCLUSIONS

Our results suggest that, when properly adjusted for potentialconfounding factors, women with severe COPD using LTOThave a greater risk of death than do men. It is not clear whythis is the case; we recommend additional studies that addressthe role of inflammation, bronchial hyperresponsiveness, oxida-tive stress, and other variables that may influence women’s lowersurvival rate.

Conflict of Interest Statement : M.-C.L.M. does not have a financial relationshipwith a commercial entity that has an interest in the subject of the manuscript.J.A.K. does not have a financial relationship with a commercial entity that has aninterest in the subject of the manuscript. S.A.B. has served on advisory boardsfor GlaxoSmithKline, ALTANA, Schering Plough, and Merck, and donates mostof her honoraria to the American Thoracic Society. She has participated in COPDworkshops funded by AstraZeneca and GlaxoSmithKline and is Scientific Directorfor the Burden of Obstructive Lung Disease (BOLD) Initiative that receives un-restricted educational grants to the Kaiser Permanente Center for Health Researchfrom Boehringer Ingelheim, Pfizer, GlaxoSmithKline, AstraZeneca, Novartis,Cheesy, and Merck. A.L.B. does not have a financial relationship with a commercialentity that has an interest in the subject of the manuscript. G.P.F. does not havea financial relationship with a commercial entity that has an interest in the subjectof the manuscript. M.G.S. does not have a financial relationship with a commercialentity that has an interest in the subject of the manuscript. F.Q. does not have afinancial relationship with a commercial entity that has an interest in the subjectof the manuscript. W.M.V. does not have a financial relationship with a commercialentity that has an interest in the subject of the manuscript.

Acknowledgment : The lead author acknowledges the contribution of the Ameri-can Thoracic Society’s Methods in Epidemiologic, Clinical, and OperationsResearch course in the development of this article.

References

1. Pauwels RA, Buist AS, Calverly PMA, Jenkins CR, Hurd SS. Globalstrategy for the diagnosis, management, and prevention of chronicobstructive pulmonary disease. NHLBI/WHO Global Initiative forChronic Obstructive Lung Disease (GOLD) Workshop summary. AmJ Respir Crit Care Med 2001;163:1256–1276.

2. Jardim JR, Oliveira JA, Nascimento O. II consenso Brasileiro de DPOC.J Brasileiro de Pneumologia 2004;30:S1–S42. Available at: www.scielo.br/jbpneumo

3. Celli B, MacNee W. Standards for the diagnosis and treatment of patientswith COPD: a summary of ATS/ERS position paper. Eur Respir J 2004;23:1–15.

4. Davis RM, Novotny TE. Changes in risk factors: the epidemiology ofcigarette smoking on chronic obstructive pulmonary disease. Am RevRespir Dis 1989;140:S82–S84.

5. Wise RA. Changing smoking patterns and mortality from chronic ob-structive pulmonary disease. Prev Med 1997;26:418–421.

6. National Center for Health Statistics. Vital Statistics of the United States,Vol. II, Pt. A. Washington, DC: U.S. Government Printing Office,Public Health Service; 1991.

7. Klebba AJ. Mortality from diseases associated with smoking, UnitedStates, 1960–1977. Washington, DC: National Center for Health Statis-tics; 1982. DHHS Publication No. (PHS)82–1854 (Vital and HealthStatistics Series 20, No.17).

8. Feinleib M, Rosemberg HM, Collins JG, Delozier JE, Pokras R, ChevarleyFM. Trends in COPD morbidity and mortality in the United States.Am Rev Respir Dis 1989;140(Suppl.):S9–S18.

9. Machado MCL, Buist SA, Curtis JR, Fazolo GP, Santarosa M. Theinfluence of gender on survival for patients with oxygen-dependentCOPD [abstract]. Am J Respir Crit Care Med 2004;169:A220.

10. Charlson ME, Pompei P, Ales KL, Mackenzie CR. A new method ofclassifying prognostic comorbidity in longitudinal studies: develop-ment and validation. J Chron Dis 1987;40:373–383.

11. American Thoracic Society Statement. Standards for the diagnosisand care of patients with Chronic Obstructive Pulmonary Disease.Am J Respir Crit Care Med 1995;152:77S–120S.

12. StataCorp. Stata Statistical Software, release 8.2. College Station, TX:Stata Corporation; 2005

528 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 174 2006

13. Burrows B, Earle RH. Course and prognosis of chronic obstructive lungdisease: a prospective study of 200 patients. N Engl J Med 1969;280:397–404.

14. Boushy SF, Thompson HK Jr, North LB, Beale AR, Snow TR. Prognosticin chronic obstructive pulmonary disease. Am Rev Respir Dis 1973;108: 1373–1383.

15. Postma DS, Burema J, Gimeno F, May JF, Smit JM, Steenhuis EJ, WeeleLT, Sluiter HJ. Prognosis in severe chronic obstructive pulmonarydisease. Am Rev Respir Dis 1979;119:357–367.

16. Traver GA, Cline MG, Burrows B. Predictors of mortality in chronicobstructive pulmonary disease: a 15-year follow-up study. Am RevResp Dis 1979;119:895–902.

17. Bishop JM, Cross KW. Physiological variables and mortality in patientswith various categories of chronic respiratory disease. Bull Eur Physio-pathol Respir 1984;20:495–500.

18. Weitzenblum E, Hirth C, Ducolone A, Mirhom R, RasaholinjanaharyJ, Ehrhart M. Prognostic value of pulmonary artery pressure in chronicobstructive pulmonary disease. Thorax 1981;36:752–758.

19. Medical Research Council Working Party. Report of long-term domicili-ary oxygen therapy in chronic hypoxic cor pulmonale complicatingchronic bronchitis and emphysema. Lancet 1981;1:681–685.

20. Nocturnal Oxygen Therapy Trial Group. Continuous or nocturnal oxygentherapy in hypoxemic chronic obstructive lung diseases. Ann InternMed 1980;93:391–398.

21. Chailleux E, Fauroux B, Binet F, Dautzenberg B, Polu JM, for theObservatory Group of ANTADIR. Predictors of survival in patientsreceiving domiciliary oxygen therapy or mechanical ventilation: a 10-year analysis of ANTADIR observatory. Chest 1996;109:741–749.

22. Aida A, Miyamoto K, Nishimura M, Aiba M, Kira S, Kawakami Y, andthe Respiratory Failure Research Group in Japan. Prognostic value ofhypercapnia in patients with chronic respiratory failure during long-term oxygen therapy. Am J Respir Crit Care Med 1998;158:188–193.

23. Miyamoto K, Aida A, Nishimura M, Aiba M, Kira S, Kawakami Y.Gender effect on prognosis of patients receiving long-term home oxy-gen therapy. Am J Respir Crit Care Med 1995;152:972–976.

24. Cooper CB, Waterhouse J, Howard P. Twelve year clinical study ofpatients with hypoxic cor pulmonale given long term domiciliary oxy-gen therapy. Thorax 1987;42:105–110.

25. Strom K, Boe J, Boman G, Midgren B, Rosenhall L. Long-term domicili-ary oxygen therapy: experiences acquired from the Swedish oxygenregister. Monaldi Arch Chest Dis 1993;48:473–478.

26. Strom K, Boe J. The Swedish Society of Chest Medicine. Quality assess-ment and predictors of survival in long-term domiciliary oxygen ther-apy. Eur Respir J 1991:4:50–58.

27. Strom K. Survival of patients with chronic obstructive pulmonary diseasereceiving long-term domiciliary oxygen therapy. Am Rev Respir Dis1993;147:585–591.

28. Chailleux E, Laaban J-P, Veale D. Prognostic value of nutritional deple-tion in patients with COPD treated by long-term oxygen therapy: datafrom ANTADIR observatory. Chest 2003;123:1460–1466.

29. Hjalmarsen A, Melbye H, Wilsgaard T, Holmboe JH, Opdahl R, ViitanenM. Prognostic for chronic obstructive pulmonary disease patients whoreceive long-term oxygen therapy. Int J Tuberc Lung Dis 1999;3:1120–1126.

30. Dallari R, Barozzi G, Pinelli G, Merighi V, Grandi P, Mazotti M, TartoniPL. Predictors of survival in subjects with chronic pulmonary diseasetreated with long-term oxygen therapy. Respiration 1994;61:8–13.

31. Dubois P, Jamart J, Machiels J, Smeets F, Lulling J. Prognosis of severelyhypoxemic patients receiving long-term oxygen therapy. Chest 1994;105:469–474.

32. Landbo C, Prescott E, Lange P, Vestbo J, Almdal TP. Prognostic valueof nutritional status in chronic obstructive pulmonary disease. Am JRespir Crit Care Med 1999;160:1856–1861.

33. Vandenbergh E, Van de Woestijne K, Gyselen A. Weight changes inthe terminal stages of chronic obstructive lung disease. Am Rev RespirDis 1967;95:556–566.

34. Wilson, DO, Rogers RM, Wright E, Anthonisen NR. Body weight inchronic obstructive pulmonary disease. Am Rev Respir Dis 1989;139:1435–1438.

35. Celli BR, Cote CG, Marin JM, Casanova C, Montes de Oca M, MendezRA, Plata VP, Cabral HJ. Body-mass index, airflow obstruction,dyspnea and exercise capacity index in chronic obstructive pulmonarydisease. 2004. N Engl J Med; 350: 1005–1012.

36. Gray-Donald K., Gibbons L, Shapiro SH, Macklem PT, Martin JG.Nutritional status and mortality in chronic obstructive pulmonary dis-ease. Am J Respir Crit Care Med 1996;153:961–966.

37. Schools AMWJ, Slangen J, Volovics L, Wouters EFM. Weight loss is areversible factor in the prognosis of chronic obstructive pulmonarydisease. Am J Respir Crit Care Med 1998;157:1791–1797.

38. Chen Y, Horne SL, Dosman JA. Increased susceptibility to lung disfunc-tion in female smokers. Am Rev Respir Dis 1991;143:1224–1230.

39. Prescott E, Bjerg AM, Andersen PK, Lange P, Vestbo J. Gender differ-ence in smoking effects on lung function and risk of hospitalizationfor COPD: results from a Danish longitudinal population study. EurRespir J 1997;10:822–827.

40. Xu X, Weiss ST, Rijcken B, Schouten JP. Smoking, changes in smokinghabits, and rate of decline in FEV1: new insight into gender differences.Eur Respir J 1994;7:1056–1061.

41. Leynaert B. Is bronchial hyperresponsiveness more frequent in womenthan in men? Am J Respir Crit Care Med 1997;156:1413–1420.

42. Kanner RE, Connett JE, Altose MD, Buist AS, Lee WW, Tashkin DP,Wise RA. Gender difference in airway hyperresponsiveness in smokerswith mild COPD. Am J Respir Crit Care Med 1994;150:956–961.

43. Paoletti P, Carrozzi L, Viegi G, Modena P, Ballerin L, Di Pede F, GradoL, Baldacci S, Pedreschi M, Vellutini M, et al. Distribution of bronchialresponsiveness in a general population: effect of Sex, age, smoking,and level of pulmonary function. Am J Respir Crit Care Med 1995;151:1770–1777.

44. Andreassen H, Vestbo J. Chronic obstructive pulmonary disease as asystemic disease: an epidemiological perspective. Eur Respir J 2003;22:2S–4S.

45. Agusti AGN. Systemic effects of chronic obstructive pulmonary disease.Proc Am Thorac Soc 2005;2:367–370.

46. Gross CP, Anderson GF, Powe NR. The relation between funding bythe National Institutes of Health and the burden of disease. N EnglJ Med 1999;340:1881–1887.

47. Rennard SI. Clinical approach to patients with chronic obstructive pulmo-nary disease and cardiovascular disease. Proc Am Thorac Soc 2005;2:94–100.

48. Zureik M, Benetos A, Neukirch C, Courbon D, Bean K, Thomas F,Ducimetiere P. Reduced pulmonary function is associated with centralarterial stiffness in men. Am J Respir Crit Care Med 2001;164:2181–2185.

49. Hole DJ, Watt GC, Davey-Smith G, Hart CL, Gillis CR, HawthorneVM. Impaired lung function and mortality risk in men and women:findings from Renfrew and Paisley prospective population study. BMJ1996;313:711–715.

50. Schunemann HJ, Dorn J, Grant BJ, Winkelstein W Jr, Trevisan M.Pulmonary function is a long-term predictor of mortality in the generalpopulation: 29-year follow-up of the Buffalo Health Study. Chest2000;118:656–664.

51. Hospers JJ, Postma DS, Rijcken B, Weiss ST, Schouten JP. Histamineairway hyper-responsiveness and mortality from chronic obstructivepulmonary disease: a cohort study. Lancet 2000;356:1313–1317.

52. MacNee W, Rahman I. Oxidants and antioxidants as therapeutic targetsin chronic obstructive pulmonary disease. Am J Respir Crit Care Med1999;160(Suppl):S58–S65.

53. Taylor JC; Madison R; Koninska D. Is antioxidant deficiency related tochronic obstructive pulmonary disease? Am Rev Respir Dis 1986;134:285–289.

54. MacNee W. Chronic obstructive pulmonary disease from science to theclinic: the role of glutathione in oxidant-antioxidant balance. MonaldiArch Chest Dis 1997;52:479–485.

55. Rahman I; MacNee W. Role of oxidants/antioxidants in smoking-inducedlung diseases. Free Radic Biol Med 1996;21:669–681.

56. Rahman I, Morrison D, Donaldson K, MacNee W. Systemic oxidativestress in asthma, COPD, and smokers. Am J Respir Crit Care Med1996;154:1055–1060.

57. Rahman I, Van-Schadewijk AAM, Crowther AJL, Hiemstra PS, StolkJ, MacNee W, De Boer WI. 4-Hydroxy-2-nonenal, a specific lipidperoxidation product, is elevated in lungs of patients with chronicobstructive pulmonary disease. Am J Respir Crit Care Med 2002;166:490–495.

58. Saetta M, Turato G, Maestrelli P, Mapp CE, Fabbri LM. Cellular andstructural bases of chronic obstructive pulmonary disease. Am J RespirCrit Care Med 2001;163:1304–1309.

59. Saetta M, Baraldo S, Corbino L, Turato G, Braccioni F, Rea F, CavallescoG, Tropeano G, Mapp CE, Maestrelli P et al. CD8�ve cells in thelungs of smokers with chronic obstructive pulmonary disease. Am JRespir Crit Care Med 1999;160:711–717.

60. Di Stefano A, Capelli A, Lusuardi M, Balbo P, Vecchio C, MaestrelliP, Mapp CE, Fabbri LM, Donner CF, Saetta M. Severity of airflow

Machado, Krishnan, Buist, et al.: Sex Differences in COPD Survival 529

limitation is associated with severity of airway inflammation in smok-ers. Am J Respir Crit Care Med 1998;158:1277–1285.

61. Wouters EFM. Local and systemic inflammation in chronic obstructivepulmonary disease. Proc Am Thorac Soc 2005;2:26–33.

62. Balzano G, Stefanelli F, Iorio C, De Felice A, Melillo EM, Martucci M,Melillo G. Eosinophilic inflammation in stable chronic obstructivepulmonary disease: relationship with neutrophils and airway function.Am J Respir Crit Care Med 1999;160:1486–1492.

63. MacNee W. Oxidant/antioxidants and COPD. Chest 2000;117:303S–317S.

64. Senior RM. Mechanisms of COPD: conference summary. Chest 2000;117(5 Suppl 1):320S–323S.

65. Barreiro E, De la Puente B, Minguella J, Corominas JM, Serrano S,Hussain SNA, Gea J. Oxidative stress and respiratory muscle dys-function in severe chronic obstructive pulmonary disease. Am J RespirCrit Care Med 2005;171:1116–1124.

66. Saetta M, Stefano A, Turato G, Maestrelli P, Turato G, Ruggieri MP,Calgagni P, Mapp CE, Ciaccia A, Fabbri LM. Airway eosinophilia inchronic bronchitis during exacerbations. Am J Respir Crit Care Med1994;150:1646–1652.