Article Figures & Data
Figures
- Figure 1.
STROBE Diagram
COPD = chronic obstructive pulmonary disease; ICS = inhaled corticosteroid; LABA = long-acting beta-adrenoceptor agonist; LAMA = long-acting muscarinic antagonist; SABA = short-acting beta-adrenoceptor agonist; SAMA = short-acting muscarinic antagonist; STROBE = Strengthening the Reporting of Observational Studies in Epidemiology.
a Categories not mutually exclusive; 1 patient can be listed in >1 medication category.
Tables
Characteristic Prevalent (n = 318,385) Matched prevalent (n = 291,687) Inception (n = 209,062) Matched inception (n = 185,346) Age, y (mean, SD) 68.5 (10.9) 68.5 (10.9) 68.7 (10.8) 68.7 (10.8) Age, y (median) 68.0 68.0 69.0 69.0 Sex, No. (%) Female 52.5 (167,115) 52.9 (154,182) 52.7 (110,137) 53.0 (98,293) Male 47.5 (151,270) 47.1 (137,505) 47.3 (98,925) 47.0 (87,053) Race, No. (%) African American or Black 5.6 (17,983) 5.0 (14,694) 6.0 (12,482) 5.2 (9,632) White 62.6 (199,360) 61.8 (180,245) 63.7 (133,164) 62.4 (115,681) Other 2.0 (6,504) 1.8 (5,288) 1.8 (3,775) 1.5 (2,789) Unknown 29.7 (94,538) 31.4 (91,460) 28.5 (59,641) 30.9 (57,244) Ethnicity, No. (%) Hispanic 3.3 (10,449) 2.9 (8,557) 3.4 (7,057) 2.9 (5,401) Non-Hispanic 64.4 (204,888) 63.6 (185,584) 65.0 (135,972) 63.9 (118,443) Unknown 32.4 (103,048) 33.4 (97,546) 31.6 (66,033) 33.2 (61,502) Smoking status, No. (%) Never smoker 10.7 (34,044) 10.6 (30,776) 11.8 (24,661) 11.8 (21,906) Past smoker 32.5 (103,430) 32.1 (93,545) 33.7 (70,469) 33.8 (62,615) Current smoker 24.0 (76,432) 22.9 (66,674) 25.0 (52,331) 24.5 (45,325) Unknown 32.8 (104,479) 34.5 (100,692) 29.5 (61,601) 29.9 (55,500) Body mass index, No. (%) <20 5.2 (16,441) 4.6 (13,417) 5.4 (11,366) 4.9 (9,095) 20-29 27.3 (86,955) 25.5 (74,383) 30.1 (63,026) 28.3 (52,478) 30-34 10.6 (33,801) 10.3 (29,925) 11.9 (24,799) 11.6 (21,463) >35 10.1 (32,149) 9.8 (28,490) 11.2 (23,509) 11.0 (20,445) Missing 46.8 (149,039) 49.9 (145,472) 41.3 (86,362) 44.2 (81,865) Charlson-Deyo Index score (at index), No. (%) ≤2 76.3 (242,873) 77.7 (226,760) 72.5 (151,637) 74.3 (137,748) >2 23.7 (75,512) 22.3 (64,927) 27.5 (57,425) 25.7 (47,598) Select Charlson-Deyo comorbidities (at index), No. (%) Congestive heart failure 14.6 (46,420) 14.1 (41,166) 15.7 (32,731) 15.0 (27,877) Peripheral vascular disease 11.2 (35,758) 10.5 (30,745) 13.0 (27,082) 12.1 (22,439) Diabetes without complications 18.8 (59,827) 18.4 (53,529) 21.3 (44,448) 20.7 (38,429) Cerebrovascular disease 9.5 (30,339) 8.8 (25,562) 11.4 (23,766) 10.4 (19,347) Renal disease 11.3 (35,838) 10.7 (31,097) 13.5 (28,167) 12.7 (23,572) Liver disease 2.8 (8,888) 2.6 (7,500) 3.2 (6,782) 2.9 (5,449) Rheumatic disease 4.9 (15,663) 4.7 (13,775) 5.5 (11,430) 5.2 (9,642) Dementia 2.4 (7,546) 2.3 (6,647) 2.6 (5,333) 2.5 (4,554) Diabetes with complications 2.3 (7,208) 2.2 (6,279) 3.1 (6,400) 2.9 (5,413) Peptic ulcer 1.3 (4,038) 1.2 (3,524) 1.6 (3,288) 1.5 (2,769) Prevalent cohort (n = 291,687) Inception cohort (n = 185,346) Long-term ICS exposure Intermediate-term ICS exposure Short-term ICS exposure Long-term ICS exposure Intermediate-term ICS exposure Short-term ICS exposure Cohort size 67,448 54,978 169,261 48,806 35,039 101,501 Age, y (mean, SD) 67.4 (10.2) 69.0 (10.8) 68.7 (11.2) 67.4 (10.2) 69.0 (10.8) 69.3 (11.1) Age, y (median) 67.0 69.0 69.0 67.0 69.0 69.0 Sex, No. (%) Female 53.8 (36,268) 53.2 (29,222) 52.4 (88,692) 53.8 (26,261) 53.2 (18,646) 52.6 (53,386) Male 46.2 (31,180) 46.8 (25,756) 47.6 (80,569) 46.2 (22,545) 46.8 (16,393) 47.4 (48,115) Charlson-Deyo Index score (at index), No. (%) ≤2 78.0 (52,596) 80.4 (44,217) 76.8 (129,947) 76.1 (37,128) 78.5 (27,507) 72.0 (73,113) >2 22.0 (14,852) 19.6 (10,761) 23.2 (39,314) 23.9 (11,678) 21.5 (7,532) 28.0 (28,388) ICS = inhaled corticosteroid.
Prevalent cohort (COPD alone) Prevalent cohort (COPD + asthma) Inception cohort (COPD alone) Inception cohort (COPD + asthma) Outcome Long-term ICS exposure Short-term ICS exposure Long-term ICS exposure Short-term ICS exposure Long-term ICS exposure Short-term ICS exposure Long-term ICS exposure Short-term ICS exposure Diabetes No. (%) 8,324 (13.53) 4,086 (3.86) 11,321 (14.65) 5,327 (3.79) 6,037 (14.98) 2,933 (4.03) 11,135 (15.78) 2,672 (4.22) Total 61,540 105,750 77,290 140,489 40,288 72,752 70,570 63,341 Test X2(1) = 5,286.8; P < .001 X2(1) = 8,321; P < .001 X2(1) = 4,256.5; P < .001 X2(1) = 4,822.6; P < .001 Osteoporosis No. (%) 5,017 (7.93) 3,261 (2.69) 6,796 (8.56) 4,323 (2.85) 3,578 (8.63) 2,361 (3.03) 6,689 (9.25) 2,186 (3.22) Total 63,250 121,191 79,350 151,462 41,450 78,039 72,296 67,972 Test X2(1) = 2,662; P < .001 X2(1) = 3,701.8; P < .001 X2(1 ) = 1,800.5; P < .001 X2(1) = 2,152.7; P < .001 Fracture No. (%) 1,802 (2.79) 1,634 (1.28) 2,399 (2.96) 2,237 (1.40) 1,295 (3.06) 1,246 (1.51) 2,356 (3.20) 1,143 (1.58) Total 64,483 127,347 80,923 159,576 42,253 82,904 73,714 72,356 Test X2(1) = 555.05; P < .001 X2(1) = 692.77; P < .001 X2(1) = 342.5; P < .001 X2(1) = 407.38; P < .001 Cataracts No. (%) 406 (0.63) 631 (0.49) 555 (0.68) 897 (0.56) 293 (0.69) 417 (0.50) 613 (0.83) 412 (0.56) Total 64,680 128,371 81,184 160,835 42,437 84,014 73,976 73,135 Test X2(1) = 14.671; P < .001 X2(1) = 14.134; P < .001 X2(1 ) = 18.677; P < .001 X2(1) = 37.03; P < .001 Pneumonia No. (%) 8,263 (13.01) 6,017 (4.91) 11,263 (14.15) 8,077 (5.28) 5,828 (14.01) 4,144 (5.24) 10,804 (14.89) 3,855 (5.60) Total 63,530 122,501 79,593 153,024 41,591 79,159 72,561 68,878 Test X2(1) = 3,866.7; P .001 X2(1) = 5,406; P < .001 X2(1 ) = 2,771.5; P < .001 X2(1) = 3,283.6; P < .001 COPD = chronic obstructive pulmonary disease; ICS = inhaled corticosteroid.
- Table 4.
Primary and Sensitivity Analysis Results of Cox Proportional Hazards Regression Models of Prevalent and Inception Cohorts
Primary analysis (COPD alone) Sensitivity analysis (COPD + asthma) Outcome Prevalenta HR (95% Cl) Inceptiona HR (95% Cl) Prevalenta HR (95% Cl) lnceptiona HR (95% Cl) Composite dichotomous Overall 2.65 (2.62-2.68) 2.60 (2.56-2.64) 2.72 (2.69-2.75) 2.75 (2.71-2.79) Individual outcomes Diabetes 2.56 (2.53-2.59) 2.48 (2.45-2.52) 2.63 (2.60-2.66) 2.64 (2.61-2.68) Fracture 2.45 (2.42-2.47) 2.39 (2.36-2.42) 2.52 (2.50-2.55) 2.54 (2.50-2.57) Osteoporosis 2.50 (2.50-2.50) 2.44 (2.41-2.48) 2.57 (2.55-2.60) 2.59 (2.56-2.62) Pneumonia 2.48 (2.45-2.50) 2.40 (2.37-2.44) 2.54 (2.52-2.57) 2.54 (2.50-2.57) Cataracts 2.45 (2.43-2.48) 2.39 (2.36-2.43) 2.53 (2.50-2.55) 2.54 (2.50-2.57) Individual recurrent outcomes Pneumonia 2.88 (2.62-3.16) 2.85 (2.53-3.22) 2.87 (2.63-3.12) 3.26 (2.92-3.63) Nontraumatic fracture 1.77 (1.42-2.21) 1.57 (1.20-2.06) 1.92 (1.58-2.33) 1.83 (1.43-2.33) COPD = chronic obstructive pulmonary disease; HR = hazard ratio.
↵a All outcomes statistically significant (P < .001).
SUPPLEMENTAL DATA IN PDF FILE BELOW
- Pace_Supp_Tables_1-4_Fig_1.pdf -
Supplemental Table 1. COPD ICD9 and ICD10 codes
Supplemental Table 2. Demographics after genetic matching
Supplemental Table 3. Univariate analysis for directionality.
Supplemental Table 4. Cox regression analysis
Supplemental Figure 1. Time to first event - inception cohort short-term ICS vs long-term ICS
- Pace_Supp_Tables_1-4_Fig_1.pdf -
PLAIN-LANGUAGE SUMMARY
Original Research
Long-Term Inhaled Corticosteroid Use for COPD Linked to Serious Long-Term Health Risks
Background and Goal:
Inhaled corticosteroids (ICS) are commonly prescribed for chronic obstructive pulmonary disease (COPD), but not recommended as first-line treatment unless patients have asthma/COPD overlap or frequent exacerbations. This study examined whether long-term ICS use (more than two years) increases the risk of serious health conditions compared with short-term use (less than 4 months).
Study Approach:
Researchers analyzed electronic health records from over 20 million patients, focusing on individuals aged 45 and older with a COPD diagnosis. They created two groups: a prevalent cohort (patients diagnosed with COPD at any time) and an inception cohort (patients newly diagnosed with COPD). Using propensity score matching, they compared long-term ICS users with those with short-term use. Researchers examined the risk of developing both a composite outcome (any of five major conditions) and individual health outcomes for over 500,000 people.
Main Results:
Overall Risk (Composite Outcome)
-
Compared with those using ICS for a short time, long-term ICS users were more than twice as likely to develop at least one of the following conditions:
-
Type 2 diabetes
-
Cataracts
-
Pneumonia
-
Osteoporosis
-
Nontraumatic fractures
-
Some conditions were also analyzed for recurrent cases (meaning patients who developed them more than once):
-
People on long-term ICS were nearly three times as likely to have repeated cases of pneumonia Fractures were also more likely to happen again
Prevalent cohort (all COPD patients): risk was 165% higher Inception cohort (newly diagnosed COPD patients): risk was 160% higher Among newly diagnosed COPD patients, 1 in 5 long-term ICS users developed at least one of these conditions.
Recurrent Cases
-
Some conditions were also analyzed for recurrent cases (meaning patients who developed them more than once):
-
People on long-term ICS were nearly three times as likely to have repeated cases of pneumonia Fractures were also more likely to happen again
Individual Outcomes
Diabetes, cataracts, pneumonia, osteoporosis, and fractures were also more common in long-term ICS users before being combined into the composite outcome.
Why It Matters: Many patients with COPD are prescribed ICS even when not medically necessary, exposing them to serious long-term risks.
Adverse Outcomes Associated With Inhaled Corticosteroid Use in Individuals With Chronic Obstructive Pulmonary Disease
Wilson D. Pace, MD, et al
DARTNet Institute, Aurora, Colorado
-
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.