Skip to main content
NCBI home page
Respiratory Research logoLink to Respiratory Research
. 2026 Jul 8;27:398. doi: 10.1186/s12931-026-03803-2

Airway wall thickness and PRISm are associated with cognitive impairment in individuals with cigarette smoking exposure

Karin F Hoth 1,2,✉, Carinda Linkenmeyer 1, Gregory L Kinney 3, John D Newell Jr 4, Elizabeth A Regan 5, Barry Make 5, Katherine A Pratte 6, Alejandro P Comellas 7, Eric A Hoffman 4, Spyridon Fortis 7, Jess G Fiedorowicz 8, Dawn L DeMeo 9,10, Abebaw M Yohannes 11,12, Anand S Iyer 12, Marilyn G Foreman 13, Surya P Bhatt 12, Ken M Kunisaki 14, Sharon M Lutz 15,16, Edwin K Silverman 9,10, James D Crapo 5, Erin Austin 17
PMCID: PMC13663183  PMID: 42420978

Abstract

Background

Individuals with chronic obstructive pulmonary disease (COPD) and cigarette smoking exposure are at increased risk of cognitive impairment; however, the clinical characteristics of those at risk are incompletely understood.

Methods

We conducted a secondary analysis of COPDGene cohort data to identify clinical characteristics, particularly pulmonary function and lung CT metrics, associated with cognitive impairment. Cognitive items available from Phase 3 included: self-report of a cognitive disorder diagnosis and cognitive difficulties, and probable cognitive impairment (pCI) defined by Mini-Cog ≤ 3. Thirty-seven variables were considered for inclusion in a mixed effects logistic regression with pCI as the dependent variable including demographics, smoking history, medical history, symptom severity, pulmonary function testing, and COPD-related lung CT variables.

Results

Among 2,079 participants (mean age = 68.7[SD ± 8.6] years, 51.2% female, 28.7% African American), the frequency of pCI differed by smoking/COPD severity group (9.4% never smokers, 18.7% smoking history-normal spirometry, 30.0% preserved ratio impaired spirometry [PRISm], 26.3% GOLD 1–4; p < 0.001). Older age, male sex, lower education, African American race, lower body mass index, higher anxiety symptoms, and CT metrics (higher Perc15 and Pi10) were associated with pCI in the final model.

Conclusions

Just under thirty percent of participants with smoking exposure and evidence of lung disease on spirometry (PRISm or GOLD 1–4) had pCI, with the highest frequency in PRISm and advanced COPD. Several demographic and clinical characteristics were associated with pCI including airway disease measured via Pi10 on lung CT pointing to a potential link between specific COPD-related physiology and cognitive impairment.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12931-026-03803-2.

Keywords: Mini-Cog, Cognition, Airway wall thickness, Cigarette smoking

Background

Individuals with chronic obstructive pulmonary disease (COPD) are at increased risk for cognitive impairment [1] with meta-analysis finding that approximately 30% of patients have evidence of cognitive impairment [2]. Prevalence is higher when using more sensitive domain specific neuropsychological tests or when targeting patients during or soon after hospitalization for an exacerbation [3]. The presence of COPD has been prospectively associated with new onset mild cognitive impairment (MCI) during the subsequent 5 years, with a dose-response relationship between COPD duration and risk [4]. Cognitive impairment is impactful as it has been associated with reduced quality of life [5], limited independence in daily functioning [6, 7], difficulty with medication adherence [8], errors with respiratory inhaler medication technique [9, 10], risk for exacerbations [11], and mortality during hospitalization in COPD [12].

Identifying which individuals are at risk for COPD-related cognitive impairment is important to guide clinical monitoring and inform future mechanistic studies. The presence of advanced COPD (classified using traditional GOLD spirometric criteria) and hypoxemia are well known risk factors for cognitive impairment [3, 13–16]; however, some patients experience cognitive impairment in the absence of hypoxemia earlier in disease progression [17, 18]. In response to calls to examine a wider range of clinical measures [19, 20], studies over the past decade have examined comorbidities including cardiovascular disease [21] and COPD/sleep apnea overlap syndrome [22–24] and their association with cognition. There is also increasing evidence that many people with smoking exposure show lung changes and experience clinical symptoms before they meet criteria for COPD diagnosis based on spirometry [25–27]. Preserved ratio impaired spirometry (PRISm; defined as reduced forced expiratory volume in 1 s (FEV1) in the context of preserved FEV1/forced vital capacity (FVC)) has been associated with adverse clinical outcomes including mortality and hospitalizations [28–30]. PRISm has also been associated with systemic inflammation and increased prevalence of metabolic disease including diabetes and cardiovascular comorbidities [31–34]. Collectively, these findings suggest that it is important to include a range of individuals, including those at risk for COPD due to prior smoking exposure and those with advanced COPD, in research focused on cognition in order to capture the full spectrum of disease.

The goals of the current analysis were two-fold. First, we described the prevalence of self-reported cognitive difficulties and objectively assessed probable cognitive impairment (pCI) in a cohort of older adults enriched for a history of smoking exposure and spanning the spectrum of COPD severity (defined by Global Initiative for Chronic Obstructive Lung Disease [GOLD] [1] and Preserved Ratio Impaired Spirometry [PRISm] [35]). Second, we identified demographic and clinical characteristics associated with pCI. We hypothesized that participants with PRISm and GOLD 1–4 COPD would have higher rates of pCI than participants without spirometric evidence of lung disease. In addition to demographics (e.g., older age, lower education) and medical comorbidities (e.g., cardiovascular disease and sleep apnea), we hypothesized that structural abnormalities in the lung assessed using CT would be associated with pCI. We reasoned that since structural abnormalities on CT have been associated with inflammation and other mechanisms [36] likely to have systemic effects, lung CT measures may be more strongly associated with pCI than spirometric impairment.

Methods

Cohort

COPDGene is a longitudinal, observational cohort study that enrolled Non-Hispanic white and African American individuals with ≥ 10 pack-years of cigarette smoking exposure across 21 US centers [37]. A smaller group of never smoking controls were also enrolled. Data collection has occurred across three phases with visits approximately every five years. Data reported here were collected at in person Phase 3 (10-year follow-up) visits from 2018-2020, prior to the COVID-19 pandemic.

Measures

Cognition

COPDGene Phase 3 included two self-report questions about cognition and the Mini-Cog. Self-report included: (1) prior diagnosis by a physician with a cognitive disorder, and (2) perceived confusion or memory loss “that is happening more often or is getting worse” over the previous 12 months [38, 39].

The Mini-Cog [40, 41] is a standardized, validated screening tool that takes 3–5-minutes to administer and involves two tasks: 3-word recall and clock drawing. Prior to Phase 3, research coordinators were provided training on administration and scoring of the Mini-Cog via webinar, with a recording of the training available throughout the study. Word list 1 (banana, sunrise, and chair) was employed [42]. Total Mini-Cog scores range from 0 to 5, with one point for each of the 3 words immediately correctly recalled and a score of either 0 or 2 for Clock Drawing based on standardized criteria [43]. Scores were entered by site coordinators into a REDCap database [44, 45] with an upload of Clock Drawing, which was re-scored by an adjudication team who coded Clock Drawing error types (Appendix B). The cohort was categorized into two groups using a cut point to maximize sensitivity based on prior research [38]:

  • “Probable cognitive impairment”(pCI) = Mini-Cog total ≤ 3; and

  • “Low cognitive impairment risk” = Mini-Cog total of 4–5

This cut point has been used in several large descriptive studies with evidence to support that it identifies individuals with milder cognitive impairment at earlier stages, as compared to the originally validated dementia cut point of ≤ 2 [42, 46].

Demographic and clinical characteristics

Demographics and smoking history were obtained via self-report. Clinical measures included post-bronchodilator spirometry, systolic blood pressure, blood oxygen saturation (SpO2), diffusion capacity for carbon monoxide (DLCO), body mass index (BMI; weight [kg]/height2[m]), and 6-minute walk test. Questionnaires assessed functional capacity (Duke Activity Status Index[DASI] [47]), anxiety and depression (Hospital Anxiety and Depression Scale[HADS-A and HADS-D ] [48]), lifetime problematic alcohol use (CAGE Substance Abuse Screening Tool ≥ 2) [49], dyspnea (modified Medical Research Council[mMRC] Dyspnea Scale ≥ 2) and impact of COPD (COPD Assessment Test[CAT]). Severe COPD exacerbations in the prior year were defined via self-report of an emergency room visit or hospitalization due to flare-up of COPD. Medical history was self-reported at Phases 1–3 and items in the analysis reflect report at any Phase.

Participants were grouped into smoking status/COPD groups (at Phase 3) including: Never Smokers, Normal Spirometry (smoking history with FEV1/FVC≥0.70 and FEV1%predicted ≥ 80), preserved ratio impaired spirometry (PRISm; smoking history with FEV1/FVC≥0.70 and FEV1%predicted < 80) and GOLD spirometry stage (GOLD 1-4 [50]).

Lung CT measures

COPDGene CT acquisition and quantification methods have been previously published [37]. Coached, full inspiratory and expiratory lung CT data were collected at Phase 3 with a low dose protocol (mean effective radiation of 1.5 mSv ± 0.7) [51]. We included three CT measures processed using Thirona software to reflect key components of COPD. First, we included % voxels≤-856 Hounsfield units (HU) on the expiratory scan referred to as low attenuation area (LAA-856E) to capture gas trapping, thought to reflect small airway dysfunction. Second, lung density was assessed on inspiratory CT using Perc15, the HU value at the 15th percentile of the lung CT histogram, often used as a marker of emphysema [52]. Finally, Pi10 was included as a measure of airway wall thickness relative to luminal area based on all segmented airways and defined as the square root of the wall area of a hypothetical airway with an internal perimeter of 10 mm [53]. Strong repeatability of these measures has been demonstrated [54].

Statistical analysis

Participant characteristics and probable cognitive impairment (pCI)

Candidate variable selection and pruning

Prior to analysis we identified 37 variables with potentially meaningful relationships with cognitive impairment selected via literature review and investigator expertise. The variables, presented in Table 1, fell into eight general domains: demographics, smoking history, COPD status and symptom severity, lung CT measures, physical measures (e.g., blood pressure), symptoms of depression and anxiety, functional capacity, and medical history. We used forward stepwise logistic regression and an alpha of 0.20 to prune from the list of 37 candidate variables those that had minimal evidence of an association with pCI. This approach prevented removing participants due to missing measurements on variables that were unlikely to meet the inclusion criteria in our final regression model, while maintaining a broader set of variables for exploratory network analysis.

Table 1.

Demographic and clinical characteristics for all participants and the two Mini-Cog defined groups: Low cognitive risk and Probable Cognitive Impairment (pCI) (N = 2079)

Variable
[n]1 		Characteristics for All Participants2 Low Cognitive Risk
(Mini-Cog 4–5)		 Probable Cognitive Impairment
(Mini-Cog ≤ 3)

Age

[n = 2079]

 68.7 (8.6) 68.5 (8.4) 69.4 (9.2)

Sex

[n = 2079]

 Female 1065 (51.2%) 849 (79.7%) 216 (20.3%)
Male 1014 (48.8%) 754 (74.4%) 260 (25.6%)

Race

[n = 2079]

 Non-Hispanic White (NHW) 1483 (71.3%) 1208 (81.5%) 275 (18.5%)
African American (AA) 596 (28.7%) 395 (66.3%) 201 (33.7%)

Education

[n = 2001]

 < High School (HS) 205 (10.2%) 122 (59.5%) 83 (40.5%)
HS Graduate or GED 434 (21.7%) 314 (72.4%) 120 (27.6%)
Some College and Above 1362 (68.1%) 1096 (80.5%) 266 (19.5%)

Income

[n = 2077]

 < $15,000 411 (19.8%) 287 (69.8%) 124 (30.2%)
$15,000-$35,000 432 (20.8%) 330 (76.4%) 102 (23.6%)
$35,000-$50,000 284 (13.7%) 230 (81.0%) 54 (19.0%)
$50,000-$75,000 276 (13.3%) 234 (84.8%) 42 (15.2%)
> $75,000 295 (14.2%) 256 (86.8%) 39 (13.2%)
Decline to answer 379 (18.2%) 265 (69.9%) 114 (30.1%)

Coupled

[n = 2077]

 Yes 903 (43.5%) 731 (81.0%) 172 (19.0%)
No 1174 (56.5%) 871 (74.2%) 303 (25.8%)

ATS Packyears

[n = 2077]

 41.9 (24.7) 41.0 (24.4) 44.7 (25.4)
Smoking Status [n = 2079] Never 153 (7.4%) 129 (84.3%) 24 (15.7%)
Former 1273 (61.2%) 1001 (78.6%) 272 (21.4%)
Current 653 (31.4%) 473 (72.4%) 180 (27.6%)
Airflow Limitation/COPD GOLD Group [n = 2079] Never Smokers 106 (5.1%) 96 (90.6%) 10 (9.4%)
Smoking History-Normal Spirometry 812 (39.1%) 660 (81.3%) 152 (18.7%)
PRISm 237 (11.4%) 166 (70.0%) 71 (30.0%)
GOLD 1 240 (11.5%) 175 (72.9%) 65 (27.1%)
GOLD 2 385 (18.5%) 286 (74.3%) 99 (25.7%)
GOLD 3 193 (9.3%) 146 (75.6%) 47 (24.4%)
GOLD 4 106 (5.1%) 74 (69.8%) 32 (30.2%)

Spirometry: FEV1% predicted

[n = 2079]

 79.8 (26.1) 80.9 (25.8) 75.9 (26.7)

Diffusion Capacity for Carbon Monoxide (DLCO)

[n = 1700]

 22.4 (4.0) 22.3 (4.0) 22.4 (4.0)

Resting SaO2

[n = 2068]

 96.1 (2.7) 96.1 (2.7) 96.0 (2.8)

COPD Assessment Test (CAT)

Total Score

[n = 2079]

 11.0 (8.3) 10.6 (8.1) 12.4 (8.7)

Severe exacerbation

in last year

[n = 2079]

 Yes 173 (8.3%) 115 (66.5%) 58 (33.5%)
No 1906 (91.7%) 1488 (78.1%) 418 (21.9%)

Dyspnea

(mMRC ≥ 2)

[n = 2079]

 Yes 759 (36.5%) 561 (73.9%) 198 (26.1%)
No 1320 (63.5%) 1042 (78.9%) 278 (21.1%)

6 min Walk Test

(distance in meters)

[n = 1954]

 1310.9 (413.2) 1340.9 (411.6) 1206.1 (402.1)
Total Lung Capacity: [n = 1961] 5.3 (1.4) 5.4 (1.4) 5.2 (1.4)

LAA-856E

[n = 1604]

 22.7 (19.2) 22.3 (19.0) 23.9 (20.0)

Perc15

[n = 1961]

 -927.7 (30.0) -928.7 (28.9) -924.6 (33.1)

Pi10

[n = 1928]

 2.3 (0.6) 2.3 (0.6) 2.4 (0.6)
Systolic Blood Pressure (in mmHg) [n = 2078] 130.2 (17.3) 130.0 (17.3) 130.8 (17.3)

Body Mass Index (BMI; kg/m2)

[n = 2079]

 28.6 (6.2) 28.7 (6.2) 28.2 (6.5)

Hospital Anxiety and Depression Scale-

Anxiety Symptoms (HADS-A)

[n = 2078]

 3.8 (3.5) 3.6 (3.4) 4.4 (3.9)

Hospital Anxiety and Depression Scale- Depressive Symptoms (HADS-D)

[n = 2078]

 3.2 (3.1) 3.0 (3.0) 3.7 (3.4)
Duke Activity Status Index (DASI) [n = 2079] 40.6 (16.1) 41.5 (15.9) 37.7 (16.6)

Depressive Disorder

[n = 2079]

 Yes 535 (25.7%) 410 (76.6%) 125 (23.4%)
No 1544 (74.3%) 1193 (77.3%) 351 (22.7%)
Anxiety Disorder [n = 2079] Yes 414 (19.9%) 312 (75.4%) 102 (24.6%)
No 1665 (80.1%) 1291 (77.5%) 374 (22.5%)

Stroke

[n = 2079]

 Yes 100 (4.8%) 74 (74.0%) 26 (26.0%)
No 1979 (95.2%) 1529 (77.3%) 450 (22.7%)

Diabetes

[n = 2079]

 Yes 427 (20.5%) 311 (72.8%) 116 (27.2%)
No 1652 (79.5%) 1292 (78.2%) 360 (21.8%)

Kidney Disease

[n = 2079]

 Yes 94 (4.5%) 65 (69.1%) 29 (30.9%)
No 1985 (95.5%) 1538 (77.5%) 447 (22.5%)

Liver Disease

[n = 2079]

 Yes 98 (4.7%) 72 (73.5%) 26 (26.5%)
No 1981 (95.3%) 1531 (77.3%) 450 (22.7%)

Obstructive Sleep Apnea

[n = 2079]

 Yes 560 (26.9%) 433 (77.3%) 127 (22.7%)
No 1519 (73.1%) 1170 (77.0%) 349 (23.0%)

Hypertension

[n = 2079]

 Yes 1183 (56.9%) 885 (74.8%) 298 (25.2%)
No 896 (43.1%) 718 (80.1%) 178 (19.9%)
Congestive Heart Failure (CHF) [n = 2079] Yes 93 (4.5%) 71 (76.3%) 22 (23.7%)
No 1986 (95.5%) 1532 (77.1%) 454 (22.9%)

Atrial Fibrillation (Afib)

[n = 2079]

 Yes 153 (7.4%) 109 (71.2%) 44 (28.8%)
No 1926 (92.6%) 1494 (77.6%) 432 (22.4%)
Coronary Artery Disease (CAD) [n = 2079] Yes 332 (16.0%) 254 (76.5%) 78 (23.5%)
No 1747 (84.0%) 1349 (77.2%) 398 (22.8%)
Peripheral Vascular Disease (PVD) [n = 2079] Yes 84 (4.0%) 56 (66.7%) 28 (33.3%)
No 1995 (96.0%) 1547 (77.5%) 448 (22.5%)
Open in a new tab

1: [n] was total non-missing values for the variables. Most variables had no missing values with only DLCO and LAA-856E on lung CT missing > 10%

2: mean (standard deviation) listed for continuous variables and total non-missing values (%) by level of categorical variables

Exploratory network analysis using a mixed graphical model

A mixed graphical model (MGM) was fit for the variables that survived pruning and pCI. Using participants with complete data for these variables, the MGM estimated the complete network of dependencies (strength of associations) between each variable pair after accounting for the other variables and their pairwise interactions. Ranking the dependencies provides network level insight into the strongest pairs and clusters of related variables. MGMs do not estimate variables’ effect on one outcome; therefore, the most accurate identification of candidates associated with pCI was through logistic regression.

Mixed effects logistic regression model

After we estimated the final logistic model as described above, we fit a mixed effects logistic model with pCI as the outcome via backwards selection applied to the pruned list of candidate variables with a 0.05 p-value threshold. Study site was included as a random effect and continuous variables were standardized to improve model convergence. All two-way interactions were evaluated for inclusion using forward selection.

R version 4.4.1 was used for all analyses. The MGM was generated with the mgm package [55], with results visualized using the ggplot2, qgraph, and wesanderson packages [56–58]. Mixed effect logistic model estimation and evaluation used the lmer and influence.ME packages [57]. A 5% significance threshold was used in hypothesis testing, and 95% confidence intervals are two-sided.

Prevalence of self-reported cognitive difficulties and probable cognitive impairment (pCI)

Frequencies of the two self-reported measures of cognitive difficulties (prior diagnosis and worsening confusion/memory loss) were calculated and compared to pCI via Mini-Cog using McNemar’s Chi-squared test. Prevalences of pCI via Mini-Cog were compared across smoking/COPD groups using Chi-squared tests. Frequencies of Clock Drawing error types were calculated to identify common errors and quantify differences in error type between smoking/COPD groups using Chi-squared tests.

Results

Participants

The sample included COPDGene participants who completed in-person Phase 3 study visits from 2018-2020 prior to the COVID-19 pandemic with complete Mini-Cog data. Of the 2,079 in the study sample, 1,991 has a history of cigarette smoking (i.e., were current or former smokers with at least a 10 pack-year history) and 106 were never smokers who did not meet spirometric criteria for COPD. More specifically, there were 924 individuals with a history of smoking who had GOLD 1–4 COPD, 237 individuals with a history of smoking who had a PRISm pattern on spirometry, 812 individuals with a history of smoking who had normal spirometry, and 106 never smokers who did not meet spirometric criteria for COPD. The sample was approximately 69 years old and approximately half were female. Just under one third of the sample were African American, with two thirds non-Hispanic Whites. Table 1 summarizes the demographic and clinical characteristics of the sample.

Frequencies of probable cognitive impairment and self-reported cognitive difficulties

Frequencies of pCI and self-reported cognitive items are shown in Table 1. 27.0% (314/1,161) of participants with a history of smoking exposure and evidence of lung disease on spirometry (PRISm or GOLD 1–4) had pCI, as compared to 17.6% (162/918) for Smoking Exposure- Normal Spirometry (p < 0.001). The highest frequencies of pCI were in PRISm (30.0%) and GOLD 4 (30.2%). Both Word Recall and Clock Drawing scores differed across smoking/COPD groups (Supplementary Table 1). The most common Clock Drawing error was time set incorrectly (at 9%), which was statistically significantly different across smoking/COPD groups (p = 0.02; Supplemental Table 2).

Only 3.8% (79/2,060) of the sample reported a previous clinical diagnosis of a cognitive disorder (Table 2). 44% of those individuals were classified as having pCI (35/79 44.3%) via Mini-Cog, indicating inconsistency between prior diagnosis and current objective assessment (p < 0.001).

Table 2.

Agreement between self-report of cognitive difficulties and probable cognitive impairment assessed via Mini-Cog

Self-Report of Confusion or Memory Loss in Previous 12 Months Self-Report Cognitive Disorder
No Yes No Yes
1625 (78.9%) 435 (21.1%) 1981 (96.2%) 79 (3.8%)

Low Cognitive Risk

(Mini-Cog 4–5)

 1587 (77.0%) 1285 (62.4%) 302 (14.7%) 1543 (74.9.%) 44 (2.1%)

Probable Cognitive Impairment

(Mini-Cog ≤ 3)

 473 (23.0%) 340 (16.5%) 133 (6.5%) 438 (21.3%) 35 (1.7%)
Agreement between Mini-Cog and Self-Report 68.8% 76.6%
Disagreement between Mini-Cog and Self-Report 31.2% 23.4%
Adjusted McNemar’sInline graphicTest p = 0.27 p < 0.001
Open in a new tab

Mixed graphical modeling links to probable cognitive impairment

The application of forward stepwise logistic regression pruned the original 37 candidate demographic and clinical variables to 14 (see Supplemental Fig. 1’s legend). A mixed graphical model (MGM) was fit for the 14 variables and pCI, which were in mixed formats (categorical [binary, multilevel] and continuous). In Supplemental Fig. 1 variables are depicted as nodes and colored by domain. pCI is featured with a diamond and the top 50% of dependencies are shown.

The MGM depicts the strongest candidates associated with pCI and their inter-connectedness in a network structure. The variables with strong dependencies included demographics (age, gender, race, education, and income), clinical variables (Pi10, severe exacerbations, and a subset of medical history variables (peripheral vascular disease[PVD], coronary artery disease [CAD], and kidney disease). The MGM suggests that some candidate variables may impact PCI in a complicated way; for example, Perc15 (node #8) had a strong association with some demographic variables that were linked to pCI.

Logistic regression modeling probable cognitive impairment

The final mixed effects logistic regression model (Table 3) included 8 variables and one interaction term with area under the curve of 0.69 (95% bootstrap confidence interval [0.66,0.72]). The model had no collinearity, influential observation, or logit assumption issues. Race was the strongest predictor with African American participants having 2.36 times the odds of pCI than Non-Hispanic Whites (95% CI: [1.74,3.21]). Older age, higher Perc15, higher Pi10, and higher HADS-A score were associated with increased odds of pCI. The final model included an interaction between race and HADS-A (Supplementary Table 3). African American participants had on average higher HADS-A scores than Non-Hispanic White participants. HADS-A was not associated with pCI in African Americans whereas higher anxiety was associated with pCI in Non-Hispanic White participants. Being female, Non-Hispanic White, having completed more education, and increased BMI were associated with decreased odds of pCI.

Table 3.

Final mixed effects logistic regression model with pCI as the dependent variable

Variable Logistic Coefficient
[95% CI]		 Odds Ratio
[95% CI]		 p-value
Age*

0.41

[0.27,0.55]

1.51

[1.31,1.73]

 < 0.001
Female compared to Male

-0.37

[-0.60,-0.13]

0.69

[0.55,0.88]

 0.002

African American (AA) compared to

Non-Hispanic White (NHW)

0.86

[0.55,1.17]

2.36

[1.74,3.21]

 < 0.001

High School Graduate or GED compared to

Not Completing High School

-0.37

[-0.76,0.02]

0.69

[0.47,1.03]

 0.07

Some College or College Degree compared to

Not Completing High School

-0.69

[-1.07,-0.31]

0.50

[0.34,0.73]

 < 0.001
Body Mass Index (BMI)*

-0.13

[-0.26,-0.01]

0.88

[0.77,0.99]

 0.04
Anxiety Symptoms (HADS-A)*

0.26

[0.11,0.41]

1.30

[1.12,1.51]

 0.001
Interaction of Race (AA) and HADS-A

-0.27

[-0.49,-0.05]

0.76

[0.61,0.95]

 0.02
Lung CT: Perc15*

0.14

[0.00,0.27]

1.14

[1.00,1.31]

 0.046
Lung CT: Pi10*

0.22

[0.10,0.33]

1.24

[1.11,1.40]

 < 0.001
Open in a new tab

Shown are variables significant at the p<0.05 level; Other candidate variables entered in the model that were removed by backward selection included: severe exacerbations in prior year, income, kidney disease, sleep apnea, coronary artery disease, peripheral vascular disease

Confidence Intervals (CI) were calculated using Wald Method

*Age, Perc15, Pi10, BMI, and HADS-A were standardized (centered on mean and divided by standard deviation)

Post-hoc examination of lung CT measures by smoking/COPD group

Given that the frequency of pCI was elevated in PRISm and that higher (not lower as expected) Perc 15 was associated with pCI in our regression modeling, we conducted several post-hoc analyses to better understand Perc15 in relation to other variables in our sample and determine the potential impact of total lung volume and current smoking. First, we examined the pattern of standardized values for FEV1%predicted, Perc15 and Pi10 by smoking/COPD group (Supplemental Fig. 2 provides boxplots by group). The notable observation is that while Pi10 shows the anticipated pattern of increasing values from Never Smokers toward GOLD 4 (with PRISm similar to GOLD 2), the pattern for Perc15 differs. Perc15 is notably higher for PRISm than all other groups. Second, we re-analyzed the data utilizing adjusted lung density, which incorporates total lung volume, in place of Perc15 to determine whether differences in lung volume could account for the findings. The results were unchanged from the original model when using adjusted lung density (Inline graphic=0.07, p = 0.05). Third, we considered the possibility that collinearity could be impacting our model; however, examining variance inflation factors showed that the model had no collinearity (max VIF=1.88 < 5) indicating that the inclusion of the two lung CT measures does not explain the observation. Finally, we examined the potential impact of current smoking as a main effect and interaction, and it did not change the direction of effect or statistical significance of Perc15.

Discussion

The goals of the study were to examine the prevalence of cognitive difficulties and identify clinical factors associated with probable cognitive impairment (pCI) in a large sample of older adults enriched for individuals with cigarette smoking exposure and COPD. 27% of participants with smoking exposure (≥ 10 pack-years) and evidence of lung disease on spirometry (PRISm or GOLD 1–4) had pCI based on the Mini-Cog, prevalence comparable to prior literature utilizing other dementia screening tools [2]. The highest frequencies of pCI were observed among individuals with PRISm physiology (reduced FEV1 in the setting of a preserved FEV1/FVC ratio) and very severe COPD (GOLD 4; both 30%). In contrast, 9% of never smokers and 19% of those with smoking exposure, but normal spirometry, had pCI. Regression modeling revealed that older age, male sex, African American race, lower education level, lower BMI, elevated anxiety, higher lung density (Perc15) and more airway disease (Pi10) were associated with pCI. Evidence of a link between lung CT measures and pCI is an important contribution consistent with the idea that COPD-related lung pathophysiology may uniquely increase risk for cognitive impairment.

That individuals with COPD, classified using traditional GOLD criteria, are at higher risk for cognitive impairment than healthy age-matched controls has been demonstrated in multiple prior studies [59]; however, the observation that individuals with PRISm are at risk for cognitive impairment is not as well established. Individuals with PRISm have proportionate decreases in FEV1 and FVC, a pattern associated with increased risk for mortality [60]. Participants in the Rotterdam study with PRISm or GOLD 2–4 were shown to have worse cognitive performance and a higher prevalence of vascular brain lesions than those with normal spirometry [61]. Kawatoko and colleagues [62] observed that community dwelling older adults with PRISm had a higher risk of new onset dementia in the subsequent 5 years than those with normal spirometry. PRISm may be an important clinical entity for identifying risk for cognitive impairment. Potential mechanisms by which PRISm may be linked to cognition include comorbid cardiovascular disease and systemic inflammation, which are common in PRISm [29, 63, 64] and known risk factors for cognitive impairment [65, 66].

The finding that COPD-related changes on lung CT improved predictions of pCI is important. Two previous studies have examined lung CT and cognition. One observed a link between the presence of emphysema, visually detected on CT, and worse cognition in mid-life using data from community dwelling participants in the Coronary Artery Risk Development in Young Adults (CARDIA) study [67]. Another examined a relatively smaller cohort of individuals diagnosed with COPD and concluded that emphysema was associated with cognitive impairment [68] In the COPDGene cohort higher Pi10, a measure of airway wall thickness was strongly associated with pCI. This suggests that airway disease may be an important aspect of COPD physiology increasing risk for cognitive impairment, perhaps through systemic inflammation and vascular mechanisms [69]. We anticipated that higher LAA-856E, reflecting greater gas trapping, would be associated with pCI, but this association was not observed. Finally, contrary to our expectation, a significant association was observed between higher Perc15 and pCI. The presence of emphysema drives Perc15 lower and thus, we anticipated that lower, not higher, Perc15 would be associated with cognitive impairment. It is important to recognize that inflammation in the lung increases Perc15. Thus, it is possible that the association between higher Perc15 and pCI that we observed was due to lung inflammation. While we do not have direct measures of lung inflammation, in post-hoc analyses we observed that participants with PRISm showed increased Pi10 and also had higher Perc15 than any other group (Supplemental Fig. 2). The combined findings suggest that individuals with airway disease are at elevated risk for cognitive impairment. Airway disease could impact cognition via systemic inflammation and cardiovascular changes having downstream impact on the cerebrovasculature. It may be that the association between lung density measures and cognition differs across COPD progression with inflammation driving the association early. Taken together, the analyses suggest that altered lung physiology is associated with cognition beyond demographic and other clinical factors. Future studies should include lung CT measures and examine potential mechanisms driving the link with cognition, which may differ at different points in the progression of COPD.

In addition to pulmonary measures, several demographic and clinical characteristics were associated with pCI. Advanced age and lower education are known risk factors for cognitive impairment. The observation that male sex was associated with pCI could be due to higher rates of cardiovascular comorbidities and diabetes and is an area for future study [70]. pCI was more frequent among African Americans than non-Hispanic Whites in our sample. Previous studies in COPD have also observed elevated prevalence of cognitive impairment among African Americans [71]. Racial differences in risk are likely the result of health disparities such as environmental exposures, lifestyle factors, and access to medical treatment. Hence, clinicians should be aware of this risk and the importance of referral where appropriate.

The interaction between race and anxiety was statistically significant in the regression suggesting that anxiety was more strongly associated with pCI among non-Hispanic White participants than African American participants. African American participants overall (with or without PCI) had higher anxiety than non-Hispanic White participants (Supplementary Table 3). Given the smaller sample size for African American participants it is possible that we had limited power to detect an association between anxiety and pCI for this subset of participants. Clinically, it is essential to select screening tools that minimize bias from cultural differences. There is some evidence that the Mini-Cog shows less bias than some other screening tools [72], but careful consideration of individual clinical circumstances is essential including the possibility of a false positive result on screening measures associated with anxiety. Finally, lower BMI was associated with pCI. Low BMI may be a proxy for advanced COPD with associated cachexia [73] that may impact the brain and cognition.

The current analysis examined history of a cognitive disorder diagnosis and worsening confusion or memory loss. The contrast between the low number of participants who reported having a prior cognitive disorder diagnosis (4%) and the 27% identified as having pCI via the Mini-Cog may, in part, reflect the limitation of retrospective self-report, but is likely also driven by under-recognition of cognitive impairment [74, 75]. Recommendations for dementia screening include self-report and objective screening, as each provides unique information [38]. In our data, only 28% of the individuals who had pCI on the Mini-Cog reported worsening confusion or memory loss, suggesting that many participants that might benefit from further cognitive assessment would be missed using self-report alone. The discrepancy underscores the importance of a combination of a clinical interview and objective screening when identifying possible impairment.

Several study limitations are important to consider. First, the Mini-Cog is a screening tool and does not provide a diagnosis of cognitive impairment. Comprehensive cognitive disorder diagnosis was not conducted as a part of COPDGene. Nonetheless, the Mini-Cog has been validated in relation to gold-standard diagnosis [41] and is utilized clinically. Future work that includes comprehensive cognitive assessment is needed to confirm the results. Second, the assessment of medical comorbidities relied on self-report. Medical record review or diagnostic assessment would have yielded more precise information regarding comorbidities. Third, although we included 37 variables potentially relevant for cognition, there are other factors that may impact cognition that were not included. For example, limited data were available regarding developmental history and neuropsychiatric comorbidities. Additionally, other aspects of lung structure and function that can be identified using lung CT could be important such as interstitial lung abnormalities (ILA). Finally, the way in which participants were recruited for COPDGene impacts the population of people that our results are likely to generalize to. COPDGene included only a small number of individuals who were never smokers and thus, it is not optimally designed for examining the effect of cigarette smoking exposure compared to never smoking on cognition. While we did include initial descriptive comparisons that included never smokers, the primary modeling and conclusions are not focused on presence or absence of smoking per se, but rather characteristics associated with cognitive impairment among individuals enriched for those with a significant cigarette smoking history. Related to the structure of the cohort, COPDGene did not recruit individuals with COPD in the absence of cigarette smoking history and thus, we cannot be sure that our conclusions generalize to individuals with COPD who have not previously smoked cigarettes. Despite these limitations, the present study included a large sample of individuals well-characterized regarding pulmonary function and, importantly, included people with smoking exposure regardless of COPD diagnosis capturing the full spectrum of COPD including those with PRISm, representing a strength.

Conclusions

In summary, older adults with a history of smoking exposure had elevated prevalence of pCI assessed via the Mini-Cog, particularly those with PRISm and very severe COPD (GOLD 4). Higher Perc15 and Pi10 were associated with pCI after accounting for a broad set of demographic and clinical variables. Lung CT may identify underlying pathophysiology that has downstream impacts on the brain. Further research replicating these findings and examining the possibility that the predominant mechanisms impacting cognition may differ among people with different COPD-related phenotypes will be important to build upon the findings. The current analyses highlight the importance of focusing on individuals with airway predominant disease and PRISm. Given heterogeneity within PRISm it is likely that there are certain individuals who are at particularly enhanced risk for cognitive impairment.

Supplementary Information

12931_2026_3803_MOESM1_ESM.docx (132.8KB, docx)

Supplementary Material 1: Supplementary Table 1. Frequencies of history of cognitive concerns and probable cognitive impairment based on the Mini-Cog. Supplementary Table 2. Frequencies of Mini-Cog Clock Drawing Error Types by Smoking/COPD Group. Supplementary Table 3. Interaction of Race x HADS-A. Supplementary Figure 1. Mixed graphical model for demographic and clinical variables and probable cognitive impairment (pCI) defined by Mini-Cog ≤3 (diamond node labeled ‘1’). Lines or edges connecting nodes show a dependency between two variables, conditional because it is the relationship after adjusting for all other variables and their pairwise interactions. The strongest 50% of the pair-wise relationships in the network are plotted with edges connecting the nodes. Thicker lines correspond to stronger dependencies. Dark grey edges connect pCI to the candidate variables and the lighter orange color depict edges between two non-Mini-Cog variables. Supplementary Figure 2. Side-by-side boxplots of smoking/COPD group by three standardized clinical variables: FEV1pp (left panel), Perc15 (center panel), and Pi10 (right panel). Values were standardized to facilitate comparison across measures.

12931_2026_3803_MOESM2_ESM.docx (19.8KB, docx)

Supplementary Material 2: Appendix A. COPDGene Sites and Investigators. Appendix B. COPDGene Phase 3 Mini-Cog Clock Drawing Error Classification.

Acknowledgements

We would like to acknowledge Soo Borson, MD for permission to use the Mini-Cog for this research and consulting regarding initial Mini-Cog scoring procedures.

Abbreviations

COPD

Chronic Obstructive Pulmonary Disease

pCI

Probable Cognitive Impairment

CT

Computed Tomography

GOLD

Global Initiative for Chronic Obstructive Lung Disease

PRISm

Preserved Ratio Impaired Spirometry

COPDGene

COPD Genetic Epidemiology study

Perc15

Lung density at the 15th percentile of the lung density histogram

Pi10

Average wall thickness of a theoretical airway with a lumen perimeter of 10 mm

MCI

Mild Cognitive Impairment

SpO2

Blood oxygen saturation

DLCO

Diffusion capacity for carbon monoxide

Kg

Kilograms

M

Meters

BMI

Body Mass Index

DASI

Duke Activity Status Index

HADS-A

Hospital Anxiety and Depression Scale- Anxiety Subscale

HADS-D

Hospital Anxiety and Depression Scale- Depression Subscale

mMRC

Modified Medical Research Council dyspnea scale

CAT

COPD Assessment Test

MSv

Millisieverts

HU

Hounsfield Units

LAA

Low Attenuation Area

FEV1

Forced Expiratory Volume in 1 s

FVC

Forced Vital Capacity

mm

Millimeters

MGM

Mixed Graphical Model

PVD

Peripheral Vascular Disease

CAD

Coronary Artery Disease

VIF

Variance Inflation Factor

Authors’ contributions

JDN, EAR, BM, APC, EAH, SF, DLD, ASI, MGF, SPB, KMK, EKS, and JDC contributed to study data collection. KFH, CL, EAR, and BM developed cognitive data processing procedures. KFH conceived of the current work with GLK and EA contributing to study design. EA implemented the statistical analysis. CL, GLK, JDN, KAP, APC, JGF, AMY, KFP, and SML contributed to the interpretation of data. KFH, CL, and EA drafted the initial version of the manuscript, and all authors provided input on subsequent revisions and approved the submitted version.

Funding

The COPDGene study (NCT00608764, posted January 8, 2008) is supported by grants from the NHLBI (U01HL089897 and U01HL089856), by NIH contract 75N92023D00011, and by the COPD Foundation through contributions made to an Industry Advisory Committee that has included AstraZeneca, Bayer Pharmaceuticals, Boehringer-Ingelheim, Genentech, GlaxoSmithKline, Novartis, Pfizer and Sunovion. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Heart, Lung, and Blood Institute or the National Institutes of Health. KFH reports funding from NIH R01 HL134822. SML was supported by NIH R01MH129337.

Data availability

Data for the COPDGene study are deposited in dbGAP. The specific combination of data analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study was conducted in accordance with the Declaration of Helsinki. The study was approved by each center’s Institutional Review Board (see list of centers in Appendix A) and written documentation of the informed consent process was provided by all participants.

Consent for publication

Not applicable.

Competing interests

JDN is a paid consultant, Medical Advisor, holder of stock shares and stock options with VIDA Diagnostics. JDN shares patents with VIDA Diagnostics and the University of Iowa. JDN receives Royalties from Elsevier and is a paid advisor to Sanofi. BM reports royalties from Wolters Kluwer Health, consulting fees from AstraZeneca, honoraria from Regeneron, and participation on advisory boards for Sanofi, Verona, Regeneron, GSK, Mt. Sinai, Baystate Medical Center and AstraZeneca. EAH reports consulting fees from Sanofi and that he is a founder and shareholder for VIDA Diagnostics. SF owns stock in ROMTech. DLD reports grant support from Bayer and the Alpha-1 Foundation and reports funds from Astra Zeneca as a member of a Medical Advisory Board. AMY reports consulting fees from Cheisi Bio Pharma and travel fees from Theravance Bio Pharma Limited. AMY is Associate Editor of Respiratory Medicine; his participation on the project complies with Respiratory Medicine requirements for recusal from review and decisions for submitted manuscript. ASI reports consulting fees from consulting fees from Verona, AstraZeneca and Medscape. MGF reports grant support from Novartis Beacon of Hope Center of Excellence. SPB reports grant funding from AstraZeneca, Connect Biopharma, Apreo, Sanofi, Genentech, Nuyaira, and the COPD Foundation, royalties from Springer, consulting fees from Sanofi, Regeneron, Chiesi, GSK, Genentech, Connect Biopharma, Apreo, Merck, AstraZeneca, Kymera, Verona, Boehringer Ingelheim, Uniquity, Polarean, and payments or honoraria from Medscape, Horizon CME, Integrity CE, Illuminate Health, and Integritas Communications. KMK reports payment from Nuvaira for participation on a data and safety monitoring board and travel support from Gilead Sciences. EKS reports grant support from Bayer and Northpond Laboratories. JC and DLD are on the COPD Foundation Board of Directors. KFH, BM, and DLD are members of the COPD Foundation Medical and Scientific Advisory Board. CL, APC, EAR, JGF, GLK, SML, KAP, and EA declare that they have no competing interests.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  • 1.Members GSC. The Global Strategy for Diagnosis, Management and Prevention of COPD 2023 Report. https://goldcopd.org/2023.
  • 2.Yohannes AM, Chen W, Moga AM, Leroi I, Connolly MJ. Cognitive Impairment in Chronic Obstructive Pulmonary Disease and Chronic Heart Failure: A Systematic Review and Meta-analysis of Observational Studies. J Am Med Dir Assoc. 2017;18(5):451 e1- e11. Epub 20170311. 10.1016/j.jamda.2017.01.014. PubMed PMID: 28292570. [DOI] [PubMed]
  • 3.Grant I, Heaton RK, McSweeny AJ, Adams KM, Timms RM. Neuropsychologic findings in hypoxemic chronic obstructive pulmonary disease. Arch Intern Med. 1982;142(8):1470–6. PubMed PMID: 7103628. [PubMed] [Google Scholar]
  • 4.Singh B, Mielke MM, Parsaik AK, Cha RH, Roberts RO, Scanlon PD, Geda YE, Christianson TJ, Pankratz VS, Petersen RC. A prospective study of chronic obstructive pulmonary disease and the risk for mild cognitive impairment. JAMA Neurol. 2014;71(5):581–8. 10.1001/jamaneurol.2014.94 . PubMed PMID: 24637951; PMCID: PMC4020948. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Schure MB, Borson S, Nguyen HQ, Trittschuh EH, Thielke SM, Pike KC, Adams SG, Fan VS. Associations of cognition with physical functioning and health-related quality of life among COPD patients. Respir Med. 2016;114:46–52. 10.1016/j.rmed.2016.03.005. Epub 20160310. [DOI] [PubMed] [Google Scholar]
  • 6.Baird C, Lovell J, Johnson M, Shiell K, Ibrahim JE. The impact of cognitive impairment on self-management in chronic obstructive pulmonary disease: A systematic review. Respir Med. 2017;129:130–9. 10.1016/j.rmed.2017.06.006. Epub 20170615. [DOI] [PubMed] [Google Scholar]
  • 7.Brunette AM, Warner K, Holm KE, Meschede K, Wamboldt FS, Kozora E, Moser DJ, Make BJ, Crapo JD, Moreau KL, Weinberger HD, Bowler R, Hoth KF. Daily Activities: The Impact of COPD and Cognitive Dysfunction. Arch Clin Neuropsychol. 2021;36(5):acaa090 767 79. 10.1093/arclin/acaa090. PubMed PMID: 33103191; PMCID: PMC8500183. [DOI] [PMC free article] [PubMed]
  • 8.O’Conor R, Muellers K, Arvanitis M, Vicencio DP, Wolf MS, Wisnivesky JP, Federman AD. Effects of health literacy and cognitive abilities on COPD self-management behaviors: A prospective cohort study. Respir Med. 2019;160:105630. PubMed PMID: 30850277; PMCID: PMC6693986. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Henkle BE, Freese RL, Dahlheimer M, Kane C, Hoth KF, Kunisaki KM. Cognitive function and inhaler technique following recovery from exacerbations of COPD. ERJ Open Res. 2023;9(3). Epub 20230502. 10.1183/23120541.00627-2022. PubMed PMID: 37143839; PMCID: PMC10152261. [DOI] [PMC free article] [PubMed]
  • 10.Iamthanaporn C, Wisitsartkul A, Chuaychoo B. Cognitive impairment according to Montreal Cognitive Assessment independently predicts the ability of chronic obstructive pulmonary disease patients to maintain proper inhaler technique. BMC Pulm Med. 2023;23(1):144. 10.1186/s12890-023-02448-x. Epub 20230426. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Dodd JW, Charlton RA, van den Broek MD, Jones PW. Cognitive dysfunction in patients hospitalized with acute exacerbation of COPD. Chest. 2013;144(1):119–27. 10.1378/chest.12-2099 . PubMed PMID: 23349026. [DOI] [PubMed] [Google Scholar]
  • 12.Spannella F, Giulietti F, Cocci G, Landi L, Lombardi FE, Borioni E, Cenci A, Giordano P, Sarzani R. Acute Exacerbation of Chronic Obstructive Pulmonary Disease in Oldest Adults: Predictors of In-Hospital Mortality and Need for Post-acute Care. J Am Med Dir Assoc. 2019;20(7):893–8. 10.1016/j.jamda.2019.01.125. Epub 20190228. [DOI] [PubMed] [Google Scholar]
  • 13.Borak J, Sliwinski P, Tobiasz M, Gorecka D, Zielinski J. Psychological status of COPD patients before and after one year of long-term oxygen therapy. Monaldi Arch Chest Dis. 1996;51(1):7–11. PubMed PMID: 8901313. [PubMed] [Google Scholar]
  • 14.Wilson DK, Kaplan RM, Timms RM, Dawson A. Acute effects of oxygen treatment upon information processing in hypoxemic COPD patients. Chest. 1985;88(2):239–43. 10.1378/chest.88.2.239 . PubMed PMID: 4017679. [DOI] [PubMed] [Google Scholar]
  • 15.Krop HD, Block AJ, Cohen E. Neuropsychologic effects of continuous oxygen therapy in chronic obstructive pulmonary disease. Chest. 1973;64(3):317–22. 10.1378/chest.64.3.317 . PubMed PMID: 4749376. [DOI] [PubMed] [Google Scholar]
  • 16.Yin M, Wang H, Hu X, Li X, Fei G, Yu Y. Patterns of brain structural alteration in COPD with different levels of pulmonary function impairment and its association with cognitive deficits. BMC Pulm Med. 2019;19(1):203. 10.1186/s12890-019-0955-y. Epub 20191107. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Kozora E, Emery C, Kaplan RM, Wamboldt FS, Zhang L, Make BJ. Cognitive and psychological issues in emphysema. Proc Am Thorac Soc. 2008;5(4):556–60. 10.1513/pats.200708-123ET . PubMed PMID: 18453371; PMCID: PMC2645335. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Liesker JJ, Postma DS, Beukema RJ, ten Hacken NH, van der Molen T, Riemersma RA, van Zomeren EH, Kerstjens HA. Cognitive performance in patients with COPD. Respir Med. 2004;98(4):351–6. 004. PubMed PMID: 15072176. [DOI] [PubMed] [Google Scholar]
  • 19.Borson S, Scanlan J, Friedman S, Zuhr E, Fields J, Aylward E, Mahurin R, Richards T, Anzai Y, Yukawa M, Yeh S. Modeling the impact of COPD on the brain. Int J Chron Obstruct Pulmon Dis. 2008;3(3):429 – 34. 10.2147/copd.s2066. PubMed PMID: 18990971; PMCID: PMC2629981. [DOI] [PMC free article] [PubMed]
  • 20.Lahousse L, Tiemeier H, Ikram MA, Brusselle GG. Chronic obstructive pulmonary disease and cerebrovascular disease: A comprehensive review. Respir Med. 2015;109(11):1371–80. 10.1016/j.rmed.2015.07. Epub 20150726. [DOI] [PubMed] [Google Scholar]
  • 21.Hoth KF, Moreau KL, Weinberger HD, Holm KE, Meschede K, Crapo JD, Make BJ, Moser DJ, Kozora E, Bowler RP, Pierce GL, Ten Eyck P, Wamboldt FS. Carotid Artery Stiffness is Associated With Cognitive Performance in Former Smokers With and Without Chronic Obstructive Pulmonary Disease. J Am Heart Assoc. 2020;9(9):e014862. 10.1161/JAHA.119.014862. Epub 20200426. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Luehrs RE, Moreau KL, Pierce GL, Wamboldt F, Aloia M, Weinberger HD, Make B, Bowler R, Crapo JD, Meschede K, Kozora E, Moser DJ, Hoth KF. Cognitive performance is lower among individuals with overlap syndrome than in individuals with COPD or obstructive sleep apnea alone: association with carotid artery stiffness. J Appl Physiol (1985). 2021;131(1):131–41. 10.1152/japplphysiol.00477.2020. Epub 20210513. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Alharbi AM, Alotaibi N, Uysal OF, Rakhit RD, Brill SE, Hurst JR, Mandal S. Cognitive outcomes in chronic obstructive pulmonary disease (COPD)/OSA overlap syndrome compared to obstructive sleep apnea (OSA) alone: a systematic review. Sleep Breath. 2025;29(5):275. 10.1007/s11325-025-03426-9. Epub 20250901. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Hansen KK, Lokke A, Jensen HI, Gantzhorn EK, Farver-Vestergaard I, Hilberg O. Examining the impact of obstructive sleep apnea on cognitive function in severe COPD. Sleep Breath. 2024;29(1):59. 10.1007/s11325-024-03228-5. Epub 20241214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Regan EA, Lynch DA, Curran-Everett D, Curtis JL, Austin JH, Grenier PA, Kauczor HU, Bailey WC, DeMeo DL, Casaburi RH, Friedman P, Van Beek EJ, Hokanson JE, Bowler RP, Beaty TH, Washko GR, Han MK, Kim V, Kim SS, Yagihashi K, Washington L, McEvoy CE, Tanner C, Mannino DM, Make BJ, Silverman EK, Crapo JD. Genetic Epidemiology of CI. Clinical and Radiologic Disease in Smokers With Normal Spirometry. JAMA Intern Med. 2015;175(9):1539–49. 10.1001/jamainternmed.2015.2735 . PubMed PMID: 26098755; PMCID: PMC4564354. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Woodruff PG, Barr RG, Bleecker E, Christenson SA, Couper D, Curtis JL, Gouskova NA, Hansel NN, Hoffman EA, Kanner RE, Kleerup E, Lazarus SC, Martinez FJ, Paine R 3rd, Rennard S, Tashkin DP, Han MK, Group SR. Clinical Significance of Symptoms in Smokers with Preserved Pulmonary Function. N Engl J Med. 2016;374(19):1811–21. 10.1056/NEJMoa1505971 . PubMed PMID: 27168432; PMCID: PMC4968204. [DOI] [PMC free article] [PubMed]
  • 27.Lambert AA, Bhatt SP. Respiratory symptoms in smokers with normal spirometry: clinical significance and management considerations. Curr Opin Pulm Med. 2019;25(2):138–43. .0000000000000550. PubMed PMID: 30461536. [DOI] [PubMed] [Google Scholar]
  • 28.Wan ES, Balte P, Schwartz JE, Bhatt SP, Cassano PA, Couper D, Daviglus ML, Dransfield MT, Gharib SA, Jacobs DR Jr., Kalhan R, London SJ, Navas-Acien A, O’Connor GT, Sanders JL, Smith BM, White W, Yende S, Oelsner EC. Association Between Preserved Ratio Impaired Spirometry and Clinical Outcomes in US Adults. JAMA. 2021;326(22):2287–98. 10.1001/jama.2021.20939 . PubMed PMID: 34905031; PMCID: PMC8672237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Higbee DH, Granell R, Davey Smith G, Dodd JW. Prevalence, risk factors, and clinical implications of preserved ratio impaired spirometry: a UK Biobank cohort analysis. Lancet Respir Med. 2022;10(2):149–57. 10.1016/S2213-2600(21)00369-6. Epub 20211102. [DOI] [PubMed] [Google Scholar]
  • 30.Washio Y, Sakata S, Fukuyama S, Honda T, Kan OK, Shibata M, Hata J, Inoue H, Kitazono T, Matsumoto K, Ninomiya T. Risks of Mortality and Airflow Limitation in Japanese Individuals with Preserved Ratio Impaired Spirometry. Am J Respir Crit Care Med. 2022;206(5):563 – 72. 10.1164/rccm.202110-2302OC. PubMed PMID: 35549659; PMCID: PMC9716906. [DOI] [PMC free article] [PubMed]
  • 31.Li G, Jankowich MD, Lu Y, Wu L, Shao L, Ke C. Preserved Ratio Impaired Spirometry, Metabolomics, and the Risk of Type 2 Diabetes. J Clin Endocrinol Metab. 2023;108(9):e769–78. 10.1210/clinem/dgad140 . PubMed PMID: 36897159. [DOI] [PubMed] [Google Scholar]
  • 32.Choi H, Oak CH, Jung MH, Jang TW, Nam SJ, Yoon T. Trend of prevalence and characteristics of preserved ratio impaired spirometry (PRISm): Nationwide population-based survey between 2010 and 2019. PLoS ONE. 2024;19(7):e0307302. 10.1371/journal.pone.0307302. Epub 20240723. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Li W, Luo C, Ying L, Zhou K. Association between preserved ratio impaired spirometry and cardiovascular disease subtypes: evidence from NHANES 2007–2012. BMC Cardiovasc Disord. 2025;25(1):696. 10.1186/s12872-025-05184-8. Epub 20250929. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Wang H, Yang R, Liu D, Li W, Prevalence. Risk Factors, Lung Function, and Associated Comorbidities of Adult Preserved Ratio Impaired Spirometry: A Meta-Analysis. MedComm (2020). 2025;6(6):e70235. Epub 20250531. 10.1002/mco2.70235. PubMed PMID: 40452817; PMCID: PMC12126599. [DOI] [PMC free article] [PubMed]
  • 35.Wan ES, Castaldi PJ, Cho MH, Hokanson JE, Regan EA, Make BJ, Beaty TH, Han MK, Curtis JL, Curran-Everett D, Lynch DA, DeMeo DL, Crapo JD, Silverman EK, Investigators CO. Epidemiology, genetics, and subtyping of preserved ratio impaired spirometry (PRISm) in COPDGene. Respir Res. 2014;15(1):89. 10.1186/s12931-014-0089-y. Epub 20140806. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Karimi R, Tornling G, Forsslund H, Mikko M, Wheelock A, Nyren S, Skold CM. Lung density on high resolution computer tomography (HRCT) reflects degree of inflammation in smokers. Respir Res. 2014;15(1):23. 10.1186/1465-9921-15-23. Epub 20140224. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Regan EA, Hokanson JE, Murphy JR, Make B, Lynch DA, Beaty TH, Curran-Everett D, Silverman EK, Crapo JD. Genetic epidemiology of COPD (COPDGene) study design. COPD. 2010;7(1):32–43. PubMed PMID: 20214461; PMCID: PMC2924193. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Cordell CB, Borson S, Boustani M, Chodosh J, Reuben D, Verghese J, Thies W, Fried LB. Medicare Detection of Cognitive Impairment W. Alzheimer’s Association recommendations for operationalizing the detection of cognitive impairment during the Medicare Annual Wellness Visit in a primary care setting. Alzheimers Dement. 2013;9(2):141–50. 10.1016/j.jalz.2012.09.011. Epub 20121220. [DOI] [PubMed] [Google Scholar]
  • 39.Goetzel RZ, Staley P, Ogden L, Strange P, Fox J, Spangler J, Tabrizi M, Beckowski M, Kowlessar N, Glasgoq RE, Taylor M. A framework for patient-centered health risk assessments- providing health promotion and disease prevention services to Medicare beneficiaries Atlanta. GA: Centers for Disease Control and Prevention; 2011. http://www.cdc.gov/policy/opth/hra/. Cited 2011 2019. [Google Scholar]
  • 40.Borson S, Scanlan J, Brush M, Vitaliano P, Dokmak A. The mini-cog: a cognitive ‘vital signs’ measure for dementia screening in multi-lingual elderly. Int J Geriatr Psychiatry. 2000;15(11):1021–7. 10.1002/1099-1166(200011)15:11<1021::aid-gps234>3. .0.co;2-6. PubMed PMID: 11113982. [DOI] [PubMed] [Google Scholar]
  • 41.Borson S, Scanlan JM, Chen P, Ganguli M. The Mini-Cog as a screen for dementia: validation in a population-based sample. J Am Geriatr Soc. 2003;51(10):1451–4. 10.1046/j.1532-5415.2003.51465. .x. PubMed PMID: 14511167. [DOI] [PubMed] [Google Scholar]
  • 42.McCarten JR, Anderson P, Kuskowski MA, McPherson SE, Borson S. Screening for cognitive impairment in an elderly veteran population: acceptability and results using different versions of the Mini-Cog. J Am Geriatr Soc. 2011;59(2):309 – 13. 10.1111/j.1532-5415.2010.03249.x. PubMed PMID: 21314650. [DOI] [PubMed]
  • 43.Lessig MC, Scanlan JM, Nazemi H, Borson S. Time that tells: critical clock-drawing errors for dementia screening. Int Psychogeriatr. 2008;20(3):459–70. Epub 20071001. doi: 10.1017/S1041610207006035. PubMed PMID: 17908348; PMCID: PMC2704110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Harris PA, Taylor R, Minor BL, Elliott V, Fernandez M, O’Neal L, McLeod L, Delacqua G, Delacqua F, Kirby J, Duda SN, Consortium RE. The REDCap consortium: Building an international community of software platform partners. J Biomed Inf. 2019;95:103208. PubMed PMID: 31078660; PMCID: PMC7254481. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inf. 2009;42(2):377–81. PubMed PMID: 18929686; PMCID: PMC2700030. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Rosenbloom M, Borson S, Barclay T, Hanson LR, Werner A, Stuck L, McCarten J. Routine cognitive screening in a neurology practice: Effect on physician behavior. Neurol Clin Pract. 2016;6(1):16–21. 0000000000000186. PubMed PMID: 26918200; PMCID: PMC4753831. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Hlatky MA, Boineau RE, Higginbotham MB, Lee KL, Mark DB, Califf RM, Cobb FR, Pryor DB. A brief self-administered questionnaire to determine functional capacity (the Duke Activity Status Index). Am J Cardiol. 1989;64(10):651–4. 10.1016/0002-9149(89)90496-7 . PubMed PMID: 2782256. [DOI] [PubMed] [Google Scholar]
  • 48.Zigmond AS, Snaith RP. The hospital anxiety and depression scale. Acta Psychiatr Scand. 1983;67(6):361 – 70. 10.1111/j.1600-0447.1983.tb09716.x. PubMed PMID: 6880820. [DOI] [PubMed]
  • 49.Ewing JA. Detecting alcoholism. The CAGE questionnaire. JAMA. 1984;252(14):1905–7. 10.1001/jama.252.14.1905 . PubMed PMID: 6471323. [DOI] [PubMed] [Google Scholar]
  • 50.Vestbo J, Hurd SS, Agusti AG, Jones PW, Vogelmeier C, Anzueto A, Barnes PJ, Fabbri LM, Martinez FJ, Nishimura M, Stockley RA, Sin DD, Rodriguez-Roisin R. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: GOLD executive summary. Am J Respir Crit Care Med. 2013;187(4):347–65. 10.1164/rccm.201204-0596. Epub 20120809. [DOI] [PubMed] [Google Scholar]
  • 51.Baraghoshi D, Strand M, Humphries SM, San Jose Estepar R, Vegas Sanchez-Ferrero G, Charbonnier JP, Latisenko R, Silverman EK, Crapo JD, Lynch DA. Quantitative CT Evaluation of Emphysema Progression over 10 Years in the COPDGene Study. Radiology. 2023;307(4):e222786. 10.1148/radiol.222786. Epub 20230411. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Dirksen A, Dijkman JH, Madsen F, Stoel B, Hutchison DC, Ulrik CS, Skovgaard LT, Kok-Jensen A, Rudolphus A, Seersholm N, Vrooman HA, Reiber JH, Hansen NC, Heckscher T, Viskum K, Stolk J. A randomized clinical trial of alpha(1)-antitrypsin augmentation therapy. Am J Respir Crit Care Med. 1999;160(5 Pt 1):1468–72. 10.1164/ajrccm.160.5.9901055 . PubMed PMID: 10556107. [DOI] [PubMed] [Google Scholar]
  • 53.Nakano Y, Wong JC, de Jong PA, Buzatu L, Nagao T, Coxson HO, Elliott WM, Hogg JC, Pare PD. The prediction of small airway dimensions using computed tomography. Am J Respir Crit Care Med. 2005;171(2):142-6. Epub 20041029. 10.1164/rccm.200407-874OC. PubMed PMID: 15516531. [DOI] [PubMed]
  • 54.Motahari A, Barr RG, Han MK, Anderson WH, Barjaktarevic I, Bleecker ER, Comellas AP, Cooper CB, Couper DJ, Hansel NN, Kanner RE, Kazerooni EA, Lynch DA, Martinez FJ, Newell JD Jr., Schroeder JD, Smith BM, Woodruff PG, Hoffman EA, Group S. Repeatability of Pulmonary Quantitative Computed Tomography Measurements in Chronic Obstructive Pulmonary Disease. Am J Respir Crit Care Med. 2023;208(6):657–65. 10.1164/rccm.202209-1698. PubMed PMID: 37490608; PMCID: PMC10515564. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Haslbeck JMB, Waldorp LJ. mgm: Estimating time-varying mixed graphical models in high-dimensional data. J Stat Softw. 2020;93(8):1–46. [Google Scholar]
  • 56.Epskamp S, Cramer AOJ, Waldorp LJ, Schmittmann VD, Borsboom D. qgraph: Network visualizations of relationships in psychometric data. J Stat Softw. 2012;48(4):1–18. 10.18637/jss.v048.i04. [DOI] [Google Scholar]
  • 57.Bates D, Maechler M, Bolker B, Walker S. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67(1):1–48. 10.18637/jss.v067.i01. [DOI] [Google Scholar]
  • 58.Wickham H. ggplot2: Elegant Graphics for Data Analysis. New York: Springer-; 2016. [Google Scholar]
  • 59.Dodd JW. Lung disease as a determinant of cognitive decline and dementia. Alzheimers Res Ther. 2015;7(1):32. 10.1186/s13195-015-0116-3. Epub 20150321. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Wan ES, Fortis S, Regan EA, Hokanson J, Han MK, Casaburi R, Make BJ, Crapo JD, DeMeo DL, Silverman EK, Investigators CO. Longitudinal Phenotypes and Mortality in Preserved Ratio Impaired Spirometry in the COPDGene Study. Am J Respir Crit Care Med. 2018;198(11):1397 – 405. 10.1164/rccm.201804-0663OC. PubMed PMID: 29874098; PMCID: PMC6290948. [DOI] [PMC free article] [PubMed]
  • 61.Xiao T, Wijnant SRA, van der Velpen I, Terzikhan N, Lahousse L, Ikram MK, Vernooij MW, Brusselle GG, Ikram MA. Lung function impairment in relation to cognition and vascular brain lesions: the Rotterdam Study. J Neurol. 2022;269(8):4141–53. 10.1007/s00415-022-11027-9. Epub 20220310. [DOI] [PubMed] [Google Scholar]
  • 62.Kawatoko K, Washio Y, Ohara T, Fukuyama S, Honda T, Hata J, Nakazawa T, Kan OK, Inoue H, Matsumoto K, Nakao T, Kitazono T, Okamoto I, Ninomiya T. Risks of dementia in a general Japanese older population with preserved ratio impaired spirometry: The Hisayama Study. J Epidemiol. 2023. Epub 20231202. 10.2188/jea.JE20230207. PubMed PMID: 38044087. [DOI] [PMC free article] [PubMed]
  • 63.Marott JL, Ingebrigtsen TS, Colak Y, Vestbo J, Lange P. Trajectory of Preserved Ratio Impaired Spirometry: Natural History and Long-Term Prognosis. Am J Respir Crit Care Med. 2021;204(8):910 – 20. 10.1164/rccm.202102-0517OC. PubMed PMID: 34233141. [DOI] [PubMed]
  • 64.Mannino DM, Ford ES, Redd SC. Obstructive and restrictive lung disease and markers of inflammation: data from the Third National Health and Nutrition Examination. Am J Med. 2003;114(9):758–62. 10.1016/s0002-9343(03)00185-2 . PubMed PMID: 12829203. [DOI] [PubMed] [Google Scholar]
  • 65.Hainsworth AH, Markus HS, Schneider JA. Cerebral Small Vessel Disease, Hypertension, and Vascular Contributions to Cognitive Impairment and Dementia. Hypertension. 2024;81(1):75–86. 10.1161/HYPERTENSIONAHA.123.19943. Epub 20231129. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Stuart MJ, Baune BT. Chemokines and chemokine receptors in mood disorders, schizophrenia, and cognitive impairment: a systematic review of biomarker studies. Neurosci Biobehav Rev. 2014;42:93–115. 10.1016/j.neubiorev.2014.02.001. Epub 20140208. [DOI] [PubMed] [Google Scholar]
  • 67.Henkle BE, Colangelo LA, Dransfield MT, Hou L, Jacobs DR Jr., Joyce BT, Pistenmaa CL, Putman RK, Sidney S, Thyagarajan B, Washko GR, Yaffe K, Kalhan R, Kunisaki KM. The presence of emphysema on chest imaging and mid-life cognition. ERJ Open Res. 2021;7(1). Epub 20210315. 10.1183/23120541.00048-2021. PubMed PMID: 33748259; PMCID: PMC7957295. [DOI] [PMC free article] [PubMed]
  • 68.Simargi Y, Turana Y, Icksan AG, Harahap AR, Siste K, Mansyur M, Damayanti T, Maryastuti M, Fazharyasti V, Dewi IP, Ramli Y, Prasetyo M, Rumende CM. A Multicenter Study of COPD and Cognitive Impairment: Unraveling the Interplay of Quantitative CT, Lung Function, HIF-1alpha, and Clinical Variables. Int J Chron Obstruct Pulmon Dis. 2024;19:1741–53. Epub 20240729. doi: 10.2147/COPD.S466173. PubMed PMID: 39099608; PMCID: PMC11296506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69.Yang HM, Ryu MH, Carey VJ, Young K, Kinney GL, Dransfield MT, Wade RC, Wells JM, Budoff MJ, Castaldi PJ, Hersh CP, Silverman EK. Differential Association of COPD Subtypes With Cardiovascular Events and COPD Exacerbations. Chest. 2024;166(6):1360–70. PubMed PMID: 39094733; PMCID: PMC11638542. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 70.Chen X, Yu Z, Liu Y, Zhao Y, Li S, Wang L. Chronic obstructive pulmonary disease as a risk factor for cognitive impairment: a systematic review and meta-analysis. BMJ Open Respir Res. 2024;11(1). Epub 20241128. 10.1136/bmjresp-2023-001709. PubMed PMID: 39608869; PMCID: PMC11603807. [DOI] [PMC free article] [PubMed]
  • 71.Yohannes AM, Holbrook MNE, Sugar JT, Henderson EA, Baker R, Casper AM, Kaminsky AS, Rea DA, Mathews AL, Que AM, Ramsdell LG, Gerald JW, Wise LB, Hanania RA. Association of mild cognitive impairment and characteristic of COPD and overall health status in a cohort study. Expert Rev Respir Med. 2021;15(1):153–9. Epub 20201118. doi: 10.1080/17476348.2021.1838278. PubMed PMID: 33054443. [DOI] [PubMed] [Google Scholar]
  • 72.Adhikari SP, Dev R, Borson S. Modifying the Mini-Cog to Screen for Cognitive Impairment in Nonliterate Individuals. Int J Alzheimers Dis. 2021;2021:5510093. Epub 20210816. 10.1155/2021/5510093. PubMed PMID: 34447592; PMCID: PMC8384546. [DOI] [PMC free article] [PubMed]
  • 73.McDonald MN, Wouters EFM, Rutten E, Casaburi R, Rennard SI, Lomas DA, Bamman M, Celli B, Agusti A, Tal-Singer R, Hersh CP, Dransfield M, Silverman EK. It’s more than low BMI: prevalence of cachexia and associated mortality in COPD. Respir Res. 2019;20(1):100. 10.1186/s12931-019-1073-3. Epub 20190522. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 74.Chodosh J, Petitti DB, Elliott M, Hays RD, Crooks VC, Reuben DB, Galen Buckwalter J, Wenger N. Physician recognition of cognitive impairment: evaluating the need for improvement. J Am Geriatr Soc. 2004;52(7):1051-9. 10.1111/j.1532-5415.2004.52301.x. PubMed PMID: 15209641. [DOI] [PubMed]
  • 75.Visser FCW, van Eersel MEA, Hempenius L, Verwey NA, Band C, van der Bol JM, Boudestein K, van Dijk SC, Gobbens R, van der Hooft CS, Kamper AM, Ruiter R, Sipers W, Spoelstra BNA, Stoffels J, Stolwijk-Woudstra DJ, van Stralen KJ, van Strien AM, Wijngaarden MA, Winters M, Strijkert F, van Munster BC. Recognition of cognitive dysfunction in hospitalised older patients: a flash mob study. BMC Geriatr. 2024;24(1):66. 10.1186/s12877-023-04588-5. Epub 20240116. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

12931_2026_3803_MOESM1_ESM.docx (132.8KB, docx)

Supplementary Material 1: Supplementary Table 1. Frequencies of history of cognitive concerns and probable cognitive impairment based on the Mini-Cog. Supplementary Table 2. Frequencies of Mini-Cog Clock Drawing Error Types by Smoking/COPD Group. Supplementary Table 3. Interaction of Race x HADS-A. Supplementary Figure 1. Mixed graphical model for demographic and clinical variables and probable cognitive impairment (pCI) defined by Mini-Cog ≤3 (diamond node labeled ‘1’). Lines or edges connecting nodes show a dependency between two variables, conditional because it is the relationship after adjusting for all other variables and their pairwise interactions. The strongest 50% of the pair-wise relationships in the network are plotted with edges connecting the nodes. Thicker lines correspond to stronger dependencies. Dark grey edges connect pCI to the candidate variables and the lighter orange color depict edges between two non-Mini-Cog variables. Supplementary Figure 2. Side-by-side boxplots of smoking/COPD group by three standardized clinical variables: FEV1pp (left panel), Perc15 (center panel), and Pi10 (right panel). Values were standardized to facilitate comparison across measures.

12931_2026_3803_MOESM2_ESM.docx (19.8KB, docx)

Supplementary Material 2: Appendix A. COPDGene Sites and Investigators. Appendix B. COPDGene Phase 3 Mini-Cog Clock Drawing Error Classification.

Data Availability Statement

Data for the COPDGene study are deposited in dbGAP. The specific combination of data analyzed during the current study are available from the corresponding author on reasonable request.

Articles from Respiratory Research are provided here courtesy of BMC

RESOURCES