Statins have pleiotropic properties that may affect the development of connective tissue diseases (CTD). The objective of this study was to compare the risk of CTD diagnoses in statin users and nonusers. This study was a propensity score-matched analysis of adult patients (30 to 85 years old) in the San Antonio military medical community. The study was divided into baseline (October 1, 2003 to September 30, 2005), and follow-up (October 1, 2005 to March 5, 2010) periods. Statin users received a statin prescription during fiscal year 2005. Nonusers did not receive a statin at any time during the study. The outcome measure was the occurrence of 3 diagnosis codes of the International Classification of Diseases, 9th Revision, Clinical Modification consistent with CTD. We described co-morbidities during the baseline period using the Charlson Comorbidity Index. We created a propensity score based on 41 variables. We then matched statin users and nonusers 1:1, using a caliper of 0.001. Of 46,488 patients who met study criteria (13,640 statin users and 32,848 nonusers), we matched 6,956 pairs of statin users and nonusers. Matched groups were similar in terms of patient age, gender, incidence of co-morbidities, total Charlson Comorbidity Index, health care use, and medication use. The odds ratio for CTD was lower in statin users than nonusers (odds ratio: 0.80; 95% confidence interval: 0.64 to 0.99; p = 0.05). Secondary analysis and sensitivity analysis confirmed these results. In conclusion, statin use was associated with a lower risk of CTD.
Statins (hydroxyl-methyl-glutaryl-coenzyme A reductase inhibitors) have been shown to interfere with downstream signaling molecules that have been implicated in both pro-inflammatory and anti-inflammatory processes. Specifically, rheumatologic diseases are characterized by both systemic inflammation and an increased risk of cardiovascular disease, making these diseases an attractive area of statin research. The effects of statins on the development of connective tissue disease (CTD) have been debated. Some studies have noted that statins may be protective against the development of rheumatoid arthritis (RA), whereas others did not observe a link between statin use and RA. Furthermore, a recent case-control study concluded that statin use was associated with an increased risk of developing RA. The objective of this study was to examine the association of statin therapy with CTD in a propensity score-matched cohort of statin users and nonusers from a military health care system, where patients have similar access and standards of care.
Methods
This study was approved by the Institutional Review Board at the Brooke Army Medical Center. This is a retrospective cohort analysis of patients who were enrolled as Tricare Prime or Tricare Plus in the San Antonio area military health care system. The database and study population have been described elsewhere. Briefly, the extracted data included outpatient medical records, inpatient medical records, administrative data of services offered outside military facilities, and pharmacy data. Outpatient medical records and inpatient medical records contain all medical services activities, diagnosis codes, and procedure codes. Pharmacy data include dispensed medications, regardless of the pharmacy location or affiliation. The Management Analysis and Reporting Tool was used to access and retrieve all patient encounter data and prescription history regardless of encounters location or affiliation. The utility and reliability of this tool in medical research is well described in the literature.
The study was divided into baseline period (October 1, 2003 to September 30, 2005), which was used to describe baseline characteristics and follow-up period (October 1, 2005 to March 5, 2010), which was used to identify outcome events. During the baseline period, we identified 2 patient groups, statin users and nonusers. Statin users received a statin prescription of at least 90-day supply during the fiscal year 2005 (October 1, 2004 to September 30, 2005); nonusers did not receive a statin at any time during the study.
Patients had to be 30 to 85 years of age, enrolled in Tricare Prime or Tricare Plus in the San Antonio area military health care system until the date of data extraction, had to have ≥1 outpatient visit during the baseline period and ≥1 outpatient visit during the follow-up period, and had to receive ≥1 prescription medication during the baseline period. Hence, our cohort had complete data throughout the study period.
We excluded burn and trauma patients; these patients were identified based on the International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) codes. Codes for burn patients were those identified by the Agency for Health Research and Quality—Clinical Classifications Software (AHRQ-CCS), category 240 ; trauma codes were compiled from ICD-9 manual and previous publications. We also excluded patients who received a statin for <90 days or those who started a statin after the baseline period to allow equal follow-up periods in both patient groups.
The outcome measure was the occurrence of 3 separate ICD-9-CM codes, during the follow-up period in either the inpatient or outpatient setting, consistent with CTD as identified by AHRQ-CCS categories 202, 210, and 211, except for V-codes because they signify previous conditions ( Appendix A ).
We described patients’ co-morbidities using the Charlson Comorbidity Index, Deyo et al method. A propensity score-matched cohort of statin users and nonusers was created using 41 variables (age, gender, 17 co-morbid conditions as listed in Table 1 and identified from ICD-9-CM diagnoses of inpatient or outpatients medical encounters, total Charlson Comorbidity Index using Deyo method, health care utilization, and the use of 14 medication groups as listed in Table 1 ).
| Variable | Users (n = 13,640) | Nonusers (n = 32,848) | p Value |
|---|---|---|---|
| Age (yrs), mean (SD) | 60 (12) | 45 (11) | <0.0001 |
| Male gender | 7,957 (58.3%) | 14,387 (43.8%) | <0.0001 |
| Co-morbid conditions | |||
| Acute myocardial infarction ∗ | 798 (5.9%) | 121 (0.4%) | <0.0001 |
| Congestive heart failure ∗ | 747 (5.5) | 164 (0.5%) | <0.0001 |
| Peripheral vascular disease ∗ | 859 (6.3%) | 190 (0.6%) | <0.0001 |
| Cerebrovascular disease ∗ | 553 (4%) | 226 (0.7%) | <0.0001 |
| Dementia ∗ | 58 (0.4%) | 45 (0.1%) | <0.0001 |
| Chronic obstructive pulmonary diseases ∗ | 2,062 (15.1%) | 2,462 (7.5%) | <0.0001 |
| Rheumatologic diseases ∗ | 290 (2.1%) | 472 (1.4%) | <0.0001 |
| Peptic ulcer disease ∗ | 220 (1.6%) | 264 (0.8%) | <0.0001 |
| Mild liver disease ∗ | 48 (0.4%) | 116 (0.4) | >0.99 |
| Diabetes mellitus ∗ | 4,389 (32.2%) | 859 (2.6%) | <0.0001 |
| Diabetes mellitus with complications ∗ | 1,664 (12.2%) | 179 (0.5%) | <0.0001 |
| Hemiplegia/paraplegia | 50 (0.4%) | 27 (0.1%) | <0.0001 |
| Renal disease ∗ | 471 (3.5%) | 117 (0.4%) | <0.0001 |
| Malignancy ∗ | 1,010 (7.4%) | 1,102 (3.4%) | <0.0001 |
| Liver disease (moderate/severe) ∗ | 8 (0.1) | 41 (0.1%) | 0.06 |
| Metastatic neoplasm ∗ | 48 (0.4%) | 95 (0.3%) | 0.3 |
| HIV ∗ | 13 (0.1%) | 39 (0.1%) | 0.5 |
| Illicit drug use | 20 (0.1%) | 65 (0.2%) | 0.3 |
| Alcohol abuse/dependence | 133 (1%) | 240 (0.7%) | .008 |
| Smoker | 1,229 (9.0%) | 1,911 (5.8%) | <0.0001 |
| Charlson Comorbidity Index score, ∗ mean (SD) | 1.2 (1.6) | 0.3 (0.8) | <0.0001 |
| Health care utilization | |||
| Number of outpatient visits during baseline period, mean (SD) | 41 (5) | 23 (32) | <0.0001 |
| Number of admission during follow-up period, mean (SD) | 0.4 (1.0) | 0.2 (0.6) | <0.0001 |
| Number of outpatient visits during follow-up period, mean (SD) | 119 (149) | 64 (79) | <0.0001 |
| Number of admission during baseline period, mean (SD) | 3 (3.1) | 2 (2) | <0.0001 |
| Medications | |||
| Beta blocker | 3,911 (28.7%) | 2,167 (6.6%) | <0.0001 |
| Diuretic | 5,121 (37.5%) | 3,421 (10.4%) | <0.0001 |
| Calcium antagonist | 3,516 (25.8%) | 1,648 (5.0%) | <0.0001 |
| Nonstatin lipid-lowering drugs | 2,324 (17.0%) | 575 (1.8%) | <0.0001 |
| Angiotensin-receptor blockers/angiotensin converting enzyme inhibitors | 7,988 (58.6%) | 3,483 (10.6%) | <0.0001 |
| Oral hypoglycemic | 2,821 (20.7%) | 385 (1.2%) | <0.0001 |
| Cytochrome p450 | 1,466 (10.7%) | 1,410 (4.3%) | <0.0001 |
| Aspirin | 7,279 (53.4%) | 2,667 (8.1%) | <0.0001 |
| Nonsteroidal anti-inflammatory drugs | 7,572 (55.5%) | 20,244 (61.6%) | <0.0001 |
| Selective serotonin reuptake inhibitors | 2,514 (18.4%) | 4,321 (13.2%) | <0.0001 |
| Systemic corticosteroid | 532 (3.9%) | 1,372 (4.2%) | 0.08 |
| Antipsychotic | 180 (1.3%) | 326 (1.0%) | 0.001 |
| Sedatives | 2,864 (21.0%) | 5,450 (16.6%) | <0.0001 |
| Tricyclic antidepressants | 35 (0.3%) | 58 (0.1%) | 0.09 |
| Mean HDL in baseline period (mg/dl) † | 53 (15) | 59 (18) | <0.0001 |
| Mean HDL in follow-up period (mg/dl) † | 51 (14) | 57 (17) | <0.0001 |
| Mean LDL in baseline period (mg/dl) † | 105 (34) | 111 (28) | <0.0001 |
| Mean LDL in follow-up period (mg/dl) † | 98 (31) | 112 (27) | <0.0001 |
∗ Diagnosis is based on ICD-9-CM codes as identified in the Deyo method for applying the Charlson Comorbidity Index score.
† Values for these laboratory measurements were missing in 8,647-7,520 patients in statin users and 26,546-18,619 patients in the nonusers.
We performed the following analyses: primary analysis in which we determined the risk of CTD in the propensity score-matched cohort; secondary analysis in which we determined the risk of CTD in relation to statin use in all patients who met study criteria (unmatched cohort); and sensitivity analysis in which we excluded patients with previous diagnosis of CTD from the propensity score-matched cohort and determined the risk of incidence of CTD diagnosis (propensity score incidence cohort).
Baseline characteristics of statin users and nonusers were compared using chi-square for categorical variables and Student’s t test for continuous variables. Comparisons were considered to be statistically significant if the calculated p value was ≤0.05. We used logistic regression to create the propensity score and test the balance of covariates in our models using the routines developed by Becker and Ichino. We then used the routine by Leuven and Sianesi to perform nearest number matching with a caliper of 0.001.
For our secondary analysis, we used logistic regression analysis to examine the odds ratios (OR) of outcome. Potential confounders (as listed in Table 1 ) were introduced as covariates in the models. Statistical analyses were performed using STATA 12 (StataCorp, College Station, Texas) and SPSS statistical software version 19 (IBM, Armonk, New York).
Results
A total of 59,604 patients met inclusion criteria: 13,116 were excluded (2,124 burn or trauma patients, 516 who received <90 days of statins, and 10,476 who received statins after September 30, 2005). Of the remaining 46,488 patients, 13,640 were statin users and 32,848 were nonusers. The mean ± SD of cumulative duration of statin use among statin users was 1,694 ± 663 days. Table 1 depicts baseline characteristics of this cohort.
We matched 6,956 pairs of statin users and nonusers using propensity scores. The matched groups had similar baseline characteristics ( Table 2 ). Among statin users, mean total duration of statin use was 1,597 days; SD was 696 days (median = 1,740 days, interquartiles = 1,097 and 2,160 days). Approximately 26% of statin users received the maximum dose of their statin, defined as 80 mg of simvastatin, 80 mg of pravastatin, 80 mg of atorvastatin, and 40 mg of rosuvastatin. Table 3 depicts the prevalence of selected ICD-9-CM codes.
| Variable | Users (n = 6,956) | Nonusers (n = 6,956) | p Value |
|---|---|---|---|
| Age (yrs), mean (SD) | 57 (13) | 57 (12) | 0.2 |
| Male gender | 3,759 (54%) | 3,816 (55%) | 0.3 |
| Co-morbid conditions | |||
| Acute myocardial infarction ∗ | 73 (1.0%) | 68 (1.0%) | 0.7 |
| Congestive heart failure ∗ | 144 (2.1%) | 124 (1.8%) | 0.2 |
| Peripheral vascular disease ∗ | 144 (2.1%) | 134 (1.9%) | 0.6 |
| Cerebrovascular disease ∗ | 153 (2.2%) | 148 (2.1%) | 0.8 |
| Dementia ∗ | 33 (0.5%) | 28 (0.4%) | 0.6 |
| Chronic obstructive pulmonary disease ∗ | 836 (12.0%) | 876 (12.6%) | 0.3 |
| Rheumatologic diseases ∗ | 158 (2.3%) | 151 (2.2%) | 0.7 |
| Peptic ulcer disease ∗ | 97 (1.4%) | 94 (1.4%) | 0.8 |
| Mild liver disease ∗ | 30 (0.4%) | 31 (0.4%) | 1 |
| Diabetes mellitus ∗ | 697 (10.0%) | 671 (9.6%) | 0.5 |
| Diabetes mellitus with complications ∗ | 198 (2.8%) | 164 (2.4%) | 0.08 |
| Hemiplegia/paraplegia ∗ | 12 (0.2%) | 8 (0.1%) | 0.5 |
| Renal disease ∗ | 96 (1.4%) | 87 (1.3%) | 0.6 |
| Malignancy ∗ | 440 (6.3%) | 439 (6.3%) | 1 |
| Liver disease (moderate/severe) ∗ | 4 (0.1%) | 5 (0.1%) | 1 |
| Metastatic neoplasm ∗ | 24 (0.3%) | 22 (0.3%) | 0.9 |
| HIV ∗ | 10 (0.1%) | 8 (0.1%) | 0.6 |
| Alcohol abuse/dependence | 66 (0.9) | 71 (1) | 0.7 |
| Smoker | 588 (8.5) | 609 (8.8) | 0.5 |
| Illicit drug use | 12 (0.2) | 11 (0.2) | 0.8 |
| Charlson Comorbidity Score, mean (SD) | 0.59 (1.1) | 0.56 (1.3) | 0.6 |
| Health care utilization | |||
| Number of outpatient visits during baseline period, mean (SD) | 32 (33) | 32 (52) | 0.6 |
| Number of admission during baseline period, mean (SD) | 0.3 (0.8) | 0.3 (0.8) | 0.2 |
| Number of outpatient visits during follow-up period, mean (SD) | 89 (83) | 89 (124) | 0.8 |
| Number of admission during follow-up period, mean (SD) | 0.8 (2) | 0.8 (2) | 0.4 |
| Medications | |||
| Beta blocker | 1,282 (18.4%) | 1,279 (18.4%) | 0.9 |
| Diuretic | 1,967 (28.3%) | 1,942 (27.9%) | 0.7 |
| Calcium antagonist | 1,141 (16.4%) | 1,091 (15.7%) | 0.2 |
| Nonstatin lipid-lowering drugs | 530 (7.6%) | 495 (7.1%) | 0.3 |
| Angiotensin-receptor blockers/angiotensin converting enzyme inhibitors | 2,420 (34.8%) | 2,416 (34.7%) | 1.0 |
| Oral hypoglycemic | 326 (4.7%) | 292 (4.2%) | 0.2 |
| Cytochrome p450 | 447 (6.4%) | 450 (6.5%) | 0.9 |
| Aspirin | 2,207 (31.7%) | 2,219 (30.5%) | 0.1 |
| Nonsteroidal anti-inflammatory drugs | 3,998 (57.5%) | 3,965 (57.0%) | 0.6 |
| Selective serotonin reuptake inhibitors | 1,166 (16.8%) | 1,135 (16.3%) | 0.5 |
| Systemic corticosteroid | 275 (4.0%) | 272 (3.9%) | 0.9 |
| Antipsychotic | 95 (1.4%) | 100 (1.4%) | 0.7 |
| Sedatives | 1,375 (19.8%) | 1,342 (19.3%) | 0.5 |
| Tricyclic antidepressants | 16 (0.2%) | 12 (0.2%) | 0.5 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
