Impact of Statin Use on Lipid Levels in Statin-Naive Patients With Rheumatoid Arthritis Versus Non–Rheumatoid Arthritis Subjects: Results From a Population-Based Study

Authors


  • The views expressed herein are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.

Mayo Clinic, 200 First Street SW, Rochester, MN 55905. E-mail: crowson@mayo.edu

Abstract

Objective

To examine lipid profiles among statin-naive patients with rheumatoid arthritis (RA) and those without RA before and after the initiation of statins.

Methods

Information regarding lipid measures and statin use was gathered in a population-based incident cohort of patients with RA (1987 American College of Rheumatology criteria first met between January 1, 1988 and January 1, 2008) and in a cohort of non-RA subjects from the same underlying population. Only patients with no prior history of statin use were included.

Results

The study included 161 patients with RA (mean age 56.3 years, 57% female) and 221 non-RA subjects (mean age 56.0 years, 66% female). Prior to the start of statins, the levels of total cholesterol and low-density lipoprotein (LDL) cholesterol were lower in the RA versus the non-RA cohort (P < 0.001 and P = 0.003, respectively). The absolute and percentage change in LDL cholesterol after at least 90 days of statin use tended to be smaller in the RA versus the non-RA cohort (P = 0.03 and P = 0.09, respectively). After at least 90 days of statin use, patients with RA were less likely to achieve therapeutic goals for LDL cholesterol than the non-RA subjects (P = 0.046). Increased erythrocyte sedimentation rate (ESR) at baseline (odds ratio 0.47, 95% confidence interval 0.26–0.85) was associated with lower likelihood of achieving therapeutic LDL goals.

Conclusion

Patients with RA had lower total cholesterol and LDL cholesterol levels before statin initiation and lower likelihood of achieving therapeutic LDL goals following statin use than the non-RA subjects. Some RA disease characteristics, in particular ESR at baseline, may have an adverse impact on achieving therapeutic LDL goals.

INTRODUCTION

Increased risk of cardiovascular (CV) disease in patients with rheumatoid arthritis (RA) is well recognized ([1-3]). Unlike the general population, where increased serum cholesterol levels are associated with increased CV risk, the relationship between RA disease and lipid profile appears to be more complex and even paradoxical. Our previous studies suggest that lower total cholesterol and low-density lipoprotein (LDL) cholesterol are associated with increased CV risk in RA and this association may be confounded by inflammation ([4]).

3-hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors are a class of “statin” cholesterol-lowering medications with a spectrum of pleiotropic CV protective effects, including antiinflammatory and immunomodulatory properties ([5]). In the general population, statin use has been linked to substantial CV risk reduction, which is proportional to the degree of LDL cholesterol lowering ([6]). Additional CV protective mechanisms of statin use may be associated with reduction of inflammation as evidenced by the findings from the Justification for the Use of Statins in Prevention: An Intervention Trial Evaluating Rosuvastatin ([7]).

Lipid lowering with statins appears to be associated with favorable effect on CV disease in patients with RA as well ([8-10]). However, mechanisms underlying the effects of statins in RA are poorly understood and the impact of RA characteristics on the lipid-lowering effect of statins has not been defined. The aim of this study was to examine lipid profiles before and after the initiation of statins in a population-based inception cohort of patients with RA and subjects without RA from the same underlying population, as well as to determine RA disease–related predictors for meeting therapeutic LDL goals in patients with RA.

Box 1. Significance & Innovations

  • This retrospective population-based cohort study reports lower total cholesterol and low-density lipoprotein (LDL) cholesterol levels before statin initiation and lower likelihood of achieving therapeutic LDL cholesterol goals following statin use in statin-naive patients with rheumatoid arthritis (RA) versus non-RA subjects.
  • Some RA disease characteristics, in particular higher disease activity as reflected in an elevated erythrocyte sedimentation rate at baseline, may adversely affect the success of lipid lowering with statins in RA.

MATERIALS AND METHODS

Study setting and design

This population-based longitudinal study was performed using the resources of the Rochester Epidemiology Project (REP), a centralized community-wide medical record linkage system. The unique features of the REP and its capabilities for the population-based research in rheumatic diseases have been described in detail elsewhere ([11, 12]).

The study included a population-based incidence cohort of patients with RA who were Olmsted County, Minnesota residents ages ≥18 years and who first fulfilled the 1987 American College of Rheumatology (ACR) criteria for RA ([13]) between January 1, 1988 and January 1, 2008. The date when the patient fulfilled ≥4 ACR criteria for RA was considered the RA incidence date. For each subject with RA, a comparison subject without RA was randomly selected from Olmsted County residents of the same age and sex in the same calendar year that each patient developed RA. Each non-RA subject was assigned an index date corresponding to the RA incidence date of the designated RA patient. Only patients with no prior history of statin use who started a statin for dyslipidemia between 1 year prior to RA incidence/index date and last followup were included.

Information on the following CV risk factors was collected at baseline as previously described ([14]): family history of premature coronary heart disease (CHD), smoking (current/former), body mass index (BMI; kg/m2), hypertension/antihypertensive treatment, and diabetes mellitus. Data on personal history of CHD (i.e., angina pectoris, coronary artery disease, and myocardial infarction [including silent events and coronary revascularization procedures, i.e., coronary artery bypass graft, percutaneous angioplasty, insertion of stents, and atherectomy]) were also gathered from the medical records.

The results of all clinically performed fasting serum lipid measures including total cholesterol, LDL, high-density lipoproteins (HDL), and triglycerides (TG) were abstracted. Abnormal lipid levels for both cohorts were defined according to the National Cholesterol Education Program (NCEP) Adult Treatment Panel III (ATP III) guidelines ([15]) as total cholesterol ≥240 mg/dl, LDL ≥160 mg/dl, TG ≥200 mg/dl, or HDL <40 mg/dl. Desirable lipid levels were defined as follows: total cholesterol <200 mg/dl, TG <150 mg/dl, HDL ≥50 mg/dl for women and ≥40 mg/dl for men, and total cholesterol/HDL ratio <4. Therapeutic goals for LDL cholesterol were defined according to the patient's CV risk factor profile as recommended in the NCEP/ATP III guidelines ([15]). In patients with 0–1 risk factor, the LDL cholesterol goal was <160 mg/dl, in patients with ≥2 risk factors the goal was <130 mg/dl, and in patients with major comorbidities (i.e., CHD/diabetes mellitus) the goal was <100 mg/dl. Risk factors for defining LDL cholesterol goals included age (≥45 years in men and ≥55 years in women), family history of premature CHD, current cigarette smoking, hypertension/antihypertensive treatment, and low HDL cholesterol (<40 mg/dl) measured on ≥3 different occasions. High HDL cholesterol (≥60 mg/dl) at baseline was considered a protective CV risk factor and allowed subtraction of one other risk factor from the total count. Information on statin use, including start and stop dates for each statin, was gathered in both cohorts.

For patients with RA, information on RA characteristics (i.e., RA duration, rheumatoid factor positivity, erythrocyte sedimentation rate [ESR] and C-reactive protein level at the time of statin initiation [baseline], joint erosions/destructive changes on radiographs, large joint swelling, the presence of rheumatoid nodules, and severe extraarticular manifestations of RA) was collected ([16]). The data regarding the use of antirheumatic medications, i.e., methotrexate, hydroxychloroquine, other disease-modifying antirheumatic drugs (DMARDs; including sulfasalazine, leflunomide, azathioprine), biologic response modifiers (BRMs), corticosteroids, and nonsteroidal antirheumatic drugs, including coxibs, were also gathered. Data on the use of acetylsalicylic acid (ASA) for arthritis (the use of >6 tablets/day of ASA [>1,950 mg/day] for ≥3 months) were recorded. The study protocol was approved by the Institutional Review Boards from Mayo Clinic and Olmsted Medical Center.

Statistical methods

The absolute and percent changes in lipid levels were calculated as the change from baseline to ≥90 days of followup. The most recent lipid measure before the day of statin initiation and the closest lipid measure to 90 days of followup were used for the analyses. T-tests and linear regression models were used to compare changes in lipid profiles between the RA and non-RA cohorts. Logistic regression models were used to examine factors associated with meeting LDL goals in RA, adjusting for age, sex, and duration of RA at statin initiation; each risk factor was examined individually.

RESULTS

A total of 161 patients with RA and 221 non-RA subjects without a history of statin use earlier than 1 year prior to RA diagnosis/index date were included. Table 1 shows baseline characteristics for both cohorts. Patients with RA had a less favorable CV profile than non-RA subjects (P = 0.02) (Table 1).

Table 1. Baseline characteristics of statin-naive patients with RA and those without RA who started a statin between 1 year prior to RA incidence/index date and last followup*
CharacteristicRA (n = 161)Non-RA (n = 221)P
  1. Values are the number (percentage) unless indicated otherwise. RA = rheumatoid arthritis; CHD = coronary heart disease; BMI = body mass index; HDL = high-density lipoprotein; CV = cardiovascular.
  2. aStatistically significant difference (P < 0.05).
  3. bDyslipidemia defined as total cholesterol ≥240 mg/dl, low-density lipoprotein cholesterol >160 mg/dl, triglycerides ≥200 mg/dl, or HDL cholesterol <40 mg/dl.
  4. cRisk factors included age (≥45 years in men, ≥55 years in women), family history of premature CHD, current cigarette smoking, hypertension or antihypertensive treatment, and low HDL cholesterol (<40 mg/dl) measured on ≥3 different occasions. High HDL cholesterol (≥60 mg/dl) at baseline was considered a protective CV risk factor and allowed subtraction of 1 other risk factor from the total count. Comorbidities included CHD and/or diabetes mellitus.
Age, mean ± SD years56.3 ± 13.256.0 ± 12.10.86
Female sex91 (57)146 (66)0.058
Smoking  0.002a
Never57 (35)119 (54)
Former61 (38)62 (28)
Current43 (27)40 (18)
Family history of CHD41 (25)67 (30)0.30
BMI ≥30 kg/m2, ever74 (46)108 (49)0.57
Diabetes mellitus43 (27)56 (25)0.76
Hypertension149 (93)173 (78)< 0.001a
Dyslipidemiab159 (99)221 (100)0.10
HDL cholesterol <40 mg/dl47 (31)38 (19)0.007a
HDL cholesterol ≥60 mg/dl41 (27)61 (30)0.56
CHD61 (38)51 (23)0.002a
Patient categories according to no. of CV risk factors and comorbiditiesc  0.02a
History of CHD or diabetes mellitus90 (56)94 (43)
≥2 risk factors32 (20)48 (22)
0–1 risk factor39 (24)79 (36)

The mean ± SD time from index date to the start of statins was similar in the RA versus non-RA cohort: 8.0 ± 6.6 years versus 8.8 ± 6.7 years (P = 0.17). The mean time from the baseline lipid test to the start of statins was also similar in both cohorts (P = 0.56). The median (range) followup time after the initiation of statin treatment was 7.0 years (0.1–20.2 years) in patients with RA versus 8 years (0.0–24.1 years) in the non-RA subjects.

Table 2 shows lipid levels at baseline and their change during the followup in both cohorts. Prior to the start of statins, the levels of total cholesterol and LDL in RA were significantly lower than in the non-RA cohort (P < 0.001 and P = 0.003, respectively); HDL cholesterol level was marginally lower in the RA versus non-RA cohort (P = 0.05). Following ≥90 days of statin use, the level of total cholesterol remained lower in RA versus non-RA subjects (P = 0.006). The decrease in absolute LDL cholesterol values after ≥90 days of statin use was less pronounced in the RA versus non-RA cohort (P = 0.03); a similar trend was observed for percentage change in LDL cholesterol, but this did not reach statistical significance (P = 0.09).

Table 2. Lipid levels at baseline and their change during the followup in patients with RA and non-RA subjects*
VariableNo.RANo.Non-RAP
  1. Values are the mean ± SD unless otherwise indicated. RA = rheumatoid arthritis; LDL = low-density lipoprotein; HDL = high-density lipoprotein; TG = triglycerides.
  2. aStatistically significant difference (P < 0.05).
  3. bReported as median (range) due to the extremely skewed distribution of values.
Baseline, mg/dl     
Total cholesterol153224.2 ± 54.0208241.1 ± 45.5< 0.001a
LDL cholesterol151139.9 ± 52.5204154.1 ± 40.30.003a
HDL cholesterol15350.3 ± 16.120653.0 ± 15.20.05
TG154179.1 ± 131.1207173.2 ± 99.60.7
Total/HDL cholesterol1534.8 ± 1.62064.9 ± 1.60.53
Followup, mg/dl     
Total cholesterol153186.8 ± 48.5207197.5 ± 44.40.006a
LDL cholesterol149104.7 ± 43.7203109.1 ± 36.70.14
HDL cholesterol15251.4 ± 16.220553.1 ± 15.10.18
TG153155.4 ± 91.7204171.5 ± 119.10.27
Total/HDL cholesterol1523.9 ± 1.42054.0 ± 1.70.92
Absolute change, mg/dl     
Total cholesterol152−37.1 ± 46.4207−43.2 ± 41.60.15
LDL cholesterol148−34.8 ± 42.7201−44.5 ± 38.30.03a
HDL cholesterol1511.2 ± 10.12030.6 ± 7.60.41
TG153−24.1 ± 105.3203−2.2 ± 83.90.2
Percentage change, mg/dl     
Total cholesterol152−14.3 ± 21.1207−16.8 ± 16.90.37
LDL cholesterol148−16.7 ± 48.6201−26.8 ± 23.60.09
HDL cholesterol1514.3 ± 20.72032.0 ± 14.20.29
TG, median (range)b153−9.8 (−72.2, −192.2)203−5.4 (−60.1, −347.1)0.31

After adjusting for age, sex, smoking, diabetes mellitus, hypertension, and CHD, the absolute change in total cholesterol and LDL cholesterol from baseline to followup was significantly smaller in the RA versus the non-RA cohort (P = 0.003 versus P = 0.002). After adjusting for age, sex, smoking, diabetes mellitus, hypertension, and CHD, the percentage change in total cholesterol, but not in other lipids, was significantly smaller (by 18%) in RA versus non-RA cohort (P = 0.007). There were no statistically significant differences in baseline and/or followup levels of HDL cholesterol and TG, as well as total cholesterol/HDL cholesterol ratio in the RA versus the non-RA cohort.

Table 3 shows proportions of subjects who had desirable lipid levels at baseline and during the followup. Total cholesterol <200 mg/dl at baseline was more common in the RA versus the non-RA cohort (P < 0.001). The likelihood of achieving desirable levels of total cholesterol, TG, HDL cholesterol, and total cholesterol/HDL cholesterol during the followup was similar in RA and non-RA cohorts. However, patients with RA, in particular those with altered LDL cholesterol levels at baseline, were less likely to reach therapeutic LDL cholesterol goals during the followup than the non-RA subjects (Table 3).

Table 3. Number (percentage) of subjects with desirable lipid levels in RA vs. non-RA cohorts*
VariableNo.RANo.Non-RAP
  1. RA = rheumatoid arthritis; LDL = low-density lipoprotein; HDL = high-density lipoprotein; TG = triglycerides.
  2. aStatistically significant difference (P < 0.05).
  3. bTherapeutic goals for LDL cholesterol were defined according to the National Cholesterol Education Program Adult Treatment Panel III guidelines based on the number of cardiovascular risk factors other than LDL cholesterol. In patients with 0–1 risk factor, the LDL cholesterol goal was <160 mg/dl, in patients with multiple (≥2) risk factors <130 mg/dl, and in patients with major comorbidities (i.e., coronary heart disease or diabetes mellitus) <100 mg/dl.
  4. cHDL cholesterol ≥50 mg/dl for women and ≥40 mg/dl for men.
  5. dDefined for each category of lipids separately, e.g., no. (%) of patients who met LDL cholesterol goals at followup among those who did not meet LDL cholesterol goals at baseline.
Baseline     
Total cholesterol <200 mg/dl15354 (350)20828 (130)< 0.001a
Therapeutic goals for LDL cholesterolb15416 (110)20825 (120)0.63
HDL cholesterolc15388 (580)206134 (650)0.15
TG <150 mg/dl15483 (540)207103 (500)0.44
Total/HDL cholesterol <4.015359 (290)20654 (350)0.18
Followup     
Total cholesterol <200 mg/dl153103 (670)207119 (570)0.06
Therapeutic goals for LDL cholesterolb15232 (210)20363 (310)0.046a
HDL cholesterolc15296 (630)205136 (660)0.53
TG <150 mg/dl15387 (570)204109 (530)0.52
Total/HDL cholesterol <4.0152121 (590)20587 (570)0.73
At followup, among those with poor lipid levels at baselined     
Total cholesterol <200 mg/dl9854 (550)17995 (530)0.78
Therapeutic goals for LDL cholesterolb13117 (130)18240 (220)0.026a
HDL cholesterolc6418 (280)7120 (280)0.99
TG <150 mg/dl7223 (320)10429 (280)0.52
Total/HDL cholesterol <4.014564 (440)10037 (370)0.34

A subgroup analysis of patients with RA (n = 100) and non-RA subjects (n = 170) without CHD revealed that baseline mean ± SD cholesterol levels were lower in RA subjects (231.2 ± 53.1 mg/dl for total cholesterol and 146.9 ± 51.4 mg/dl for LDL cholesterol) versus non-RA subjects (249.6 ± 39.4 mg/dl for total cholesterol [P = 0.004] and 160.8 ± 38.2 mg/dl [P = 0.02] for LDL cholesterol). A similar trend was noted for total cholesterol and LDL cholesterol at 90 days of followup: mean ± SD total cholesterol levels were 189.4 ± 47.7 mg/dl in RA subjects versus 202.6 ± 41.1 mg/dl in the non-RA subjects (P = 0.004) and mean ± SD LDL cholesterol levels were 107.1 ± 44.7 mg/dl in the RA cohort and 113.6 ± 36.9 mg/dl in the non-RA cohort (P = 0.08). There were no statistically significant differences in absolute and percentage change in total cholesterol and LDL in RA versus non-RA subjects without CHD (all P > 0.05).

Among the subjects without CHD, patients with RA were more likely to have desirable total cholesterol levels at baseline; total cholesterol <200 mg/dl was found in 29 (30%) patients with RA versus 13 (8%) non-RA subjects (P < 0.001). During the followup, total cholesterol <200 mg/dl was found in 64 (66%) patients with RA versus 82 (52%) non-RA subjects (P = 0.03). There were no statistically significant differences in the likelihood of achieving therapeutic LDL cholesterol goals at 90 days in RA versus non-RA patients without CHD: 32 (34%) versus 63 (40%) (P = 0.29).

We examined the impact of each of the RA disease characteristics individually on achieving LDL cholesterol goals in RA, adjusting for age, sex, and RA duration at statin initiation (Table 4). Increased ESR at baseline was associated with significantly lower likelihood of achieving LDL cholesterol goals (odds ratio [OR] per 10 mm/hour increase in ESR 0.47, 95% confidence interval [95% CI] 0.26–0.85). There was a trend toward lower likelihood of achieving LDL goals in patients with longer RA duration and in those using other DMARDs, but this did not reach statistical significance (P = 0.12 and P = 0.09, respectively). There was no statistically significant association between the calendar year of statin initiation and the likelihood of achieving LDL goals (P = 0.39) (Table 4).

Table 4. Association between individual RA disease characteristics and meeting therapeutic LDL cholesterol goals by 90 days of statin use among patients with RA (n = 161)*
RA characteristicsValueOR (95% CI)aP
  1. Values are the number (percentage) unless indicated otherwise. RA = rheumatoid arthritis; LDL = low-density lipoprotein; OR = odds ratio; 95% CI = 95% confidence interval; ESR = erythrocyte sedimentation rate; CRP = C-reactive protein; DMARDs = disease-modifying antirheumatic drugs; ASA = acetylsalicylic acid; NSAIDs = nonsteroidal antiinflammatory drugs.
  2. aAdjusted for age, sex, and duration of RA at statin initiation.
  3. bOR per 10 mm/hour increase.
  4. cStatistically significant difference (P < 0.05).
  5. dNone of patients with RA who had severe extraarticular manifestations achieved the LDL cholesterol goals, thus the OR was not estimated for this variable.
  6. eMedication use at any time prior to initiation of statin therapy.
  7. fUse of ASA for arthritis, i.e., the use of >6 tablets of ASA per day (>1,950 mg/day) for ≥3 months.
  8. gOR per 1,000 gram increase.
Duration of RA, mean ± SD years7.8 ± 6.80.95 (0.88–1.02)0.12
Calendar year of statin initiation, mean ± SD2002 ± 4.41.05 (0.94–1.17)0.39
ESR at baseline, mean ± SD mm/hour18.3 ± 18.60.47 (0.26–0.85)b0.012c
CRP level at baseline, mean ± SD mg/literd6.4 ± 24.50.02 (0.0–55.04)0.34
Rheumatoid factor positivity106 (66)0.64 (0.26–1.58)0.33
Joint erosions/destructive changes71 (44)0.58 (0.24–1.41)0.23
Large joint swelling112 (70)1.33 (0.46–3.87)0.60
Severe extraarticular manifestations of RAd15 (9)0.97
Rheumatoid nodules42 (26)1.21 (0.47–3.07)0.69
Antirheumatic treatmentse   
Methotrexate74 (46)0.85 (0.36–2.01)0.71
Hydroxychloroquine78 (48)0.71 (0.29–1.73)0.45
Other DMARDs52 (32)0.42 (0.15–1.16)0.09
Biologic agents18 (11)0.98 (0.29–3.32)0.97
Corticosteroids90 (56)0.76 (0.31–1.82)0.53
Coxibs71 (44)1.10 (0.45–2.67)0.84
ASAf70 (44)0.97 (0.35–2.68)0.95
NSAIDs137 (85)1.79 (0.40–7.91)0.44
Cumulative dose of corticosteroids, mean ± SD gm8,224.0 ± 9,813.80.99 (0.92–1.06)g0.75

Of traditional CV risk factors, older age (OR 0.43, 95% CI 0.29–0.65), male sex (OR 0.39, 95% CI 0.16–0.97), and BMI ≥30 kg/m2 (OR 0.27, 95% CI 0.10–0.71), but not other CV risk factors, were significantly associated with a lower likelihood of meeting LDL cholesterol goals. LDL cholesterol level at baseline was not associated with the likelihood of achieving LDL goals (OR 1.03, 95% CI 0.95–1.11; P = 0.51).

When statin use was compared in the RA versus non-RA cohorts, the rates of prescription of each statin (i.e., atorvastatin, cerivastatin, fluvastatin, lovastatin, pravastatin, rosuvastatin, and simvastatin with or without ezetimibe) and the frequency of changes in statin use were similar in the RA versus non-RA cohorts (data not shown). There were no apparent changes in the use of DMARDs, BRMs, and corticosteroids in patients with RA from time before the start of statin therapy to ≥90 days of followup. At baseline, 44 (27%) patients used methotrexate, 22 (14%) used hydroxychloroquine, 13 (8%) used other DMARDs, 10 (6%) received BRMs, and 55 (34%) used corticosteroids; after ≥90 days of statin use the numbers were 45 (28%), 25 (16%), 12 (7%), 10 (6%), and 58 (36%), respectively.

DISCUSSION

This retrospective population-based study found that patients with RA have lower total cholesterol and LDL cholesterol levels before statin initiation than the non-RA subjects. Absolute and relative decreases in total cholesterol and LDL cholesterol values following statin initiation tended to be smaller in RA versus non-RA subjects. Consequently, patients with RA had a lower likelihood of achieving therapeutic LDL cholesterol goals following statin use than the non-RA subjects.

The impact of statin treatment on lipid profile in RA and other inflammatory joint diseases has been reported in a series of studies using post hoc analysis data from 2 large clinical trials: the Incremental Decrease in Endpoints Through Aggressive Lipid Lowering (IDEAL) study and the Treating to New Targets (TNT) study ([9, 17]). Similar to our study, baseline levels of total cholesterol and LDL cholesterol in the IDEAL study were lower in RA subjects versus non-RA subjects. Except for the higher increase in HDL cholesterol and apolipoprotein A-1 levels in atorvastatin users, there were no statistically significant differences in changes of absolute lipid levels in RA subjects versus non-RA subjects followed for 4.8 years in that study. This was different from our study, in which decreases in total cholesterol and LDL cholesterol during ≥90 days of statin treatment tended to be smaller in RA subjects versus non-RA subjects. Longer followup and lower proportion of patients with CHD in our study may at least in part account for these differences. Alternatively, these seemingly discrepant findings could be explained by differences in the lipid-lowering effect of an intensive versus conventional regimen of statin treatment.

Concordant with the findings from the Reversal of Atherosclerosis with Aggressive Lipid Lowering trial ([18]), the post hoc analyses data from the TNT and IDEAL trials showed that intensive lipid lowering with atorvastatin 80 mg induced the highest reduction of total cholesterol and LDL cholesterol in patients with and without inflammatory joint diseases compared to conventional lipid-lowering regimens ([17]). While complete information regarding statin doses was not available in our study, it is unlikely that a substantial proportion of patients in our observational study received the intensive lipid-lowering treatment used in clinical trials and thus smaller lipid-lowering effect of statins could be expected. Nevertheless, the degree of total cholesterol lowering following statin use in our study (14% decrease in the entire RA cohort and 16% decrease in RA patients without CHD) was similar to that reported in a retrospective population-based study of RA subjects from Scotland, where a 15% decrease in total cholesterol level was reported for patients with CHD and a 16% decrease was found in patients without CHD ([10]).

Despite the smaller decrease in total cholesterol levels, patients with RA and non-RA subjects in our study had a similar likelihood of achieving desirable total cholesterol levels, which is a secondary target for lipid-lowering therapy. In contrast to what we observed regarding total cholesterol reduction, the rate of decrease of LDL cholesterol in patients with RA was not sufficient to achieve the primary target for statin intervention, i.e., therapeutic LDL cholesterol goals, with a similar likelihood as in non-RA subjects. Indeed, the majority (79%) of patients with RA did not meet LDL cholesterol goals after ≥90 days of followup.

This lack of success in achieving the LDL cholesterol targets after statin treatment has been reported in the general population, particularly in patients with high CV risk ([19, 20]). Extending these observations, our study shows that the risk of not achieving LDL cholesterol goals may be even higher in RA subjects versus non-RA subjects, highlighting the gap in the success of lipid-lowering treatment in RA versus the general population. It could be suggested that intensive lipid-lowering treatment in RA may be beneficial in some patients with RA, especially in those who failed to achieve LDL cholesterol goals with a conventional lipid-lowering regimen. Concordant with this suggestion, a treatment to target strategy for statin use, where lipid-lowering treatment was adjusted until lipid goals were achieved, was found to be successful in 92% of patients with RA referred to a preventive cardiorheumatology clinic ([21]).

The lack of success in meeting LDL cholesterol goals in our study was not apparent in patients without CHD, where both patients with RA and non-RA subjects had a similar chance of achieving the LDL goals. This finding suggests important benefits of statin treatment in primary prevention in RA and may at least in part explain significant reduction of CV disease and mortality following statin use in RA in primary, but not secondary, prevention ([10]).

Some studies from the general population suggest that the degree of lipid lowering may be less pronounced in patients with decreased baseline lipid levels. Indeed, clinical trial data from statin-naive patients with recent coronary syndrome suggest that the degree of LDL cholesterol lowering during the 4 months of statin use decreased significantly from highest to lowest baseline LDL cholesterol quartile ([22]). In our study, LDL cholesterol level at baseline was not a predictor of meeting LDL cholesterol goals, suggesting that factors associated with the success of lipid lowering in RA may be different from the general population.

To better understand the mechanisms underlying lipid lowering in statin users with RA, we investigated the impact of RA characteristics on achieving therapeutic LDL cholesterol goals. Increased ESR at baseline was associated with decreased likelihood of meeting the LDL cholesterol goals after ≥90 days of statin use in our study. Statin treatment has been previously associated with a decrease in RA disease activity in the Trial of Atorvastatin in Rheumatoid Arthritis ([8]). The results of our study to some extent explicate these findings and suggest that there may be also an association between inflammation and degree of lipid lowering where inflammation may adversely affect the success of statin treatment in patients with RA. These findings support the need for more stringent control of inflammation and scrupulous lipid monitoring in RA, and raise the possibility that tailoring of the statin treatment regimen to the degree of inflammatory activity may be advantageous in patients with RA.

While the underlying mechanisms for the association of ESR with achieving LDL cholesterol goals are unclear, impaired cholesterol transport in patients with increased ESR may be a contributory factor. This hypothesis is in line with the observations of increased cholesterol ester transfer protein activity in patients with active RA versus controls ([23]), as well as with the recent reports on associations between measures of RA disease activity, including ESR, and impaired cholesterol efflux ([24]). Certainly some antirheumatic treatments may also affect the degree of lipid lowering in statin users as suggested by a somewhat lower likelihood of achieving the LDL cholesterol goals in patients with RA patients who used DMARDs other than methotrexate and hydroxychloroquine. A better understanding of the association between inflammation and lipid lowering in RA and the clinical implications of these findings awaits further investigation.

Our study has several potential limitations. Full information on statin doses used in the study populations, as well as information on the type of health care provider (e.g., rheumatologist/internist) and information regarding patients' compliance with statin treatment was not available in this retrospective study. Therefore, we were not able to compare in detail statin treatment regimens in RA and non-RA subjects. However, there was no statistically significant difference in the rates of prescription of each statin and the frequency of changes in statin use in RA versus non-RA cohorts, which we believe minimizes this limitation. Since the study aimed to explore lipid lowering with statins, other approaches for lipid lowering including lifestyle modifications were not assessed. Optimal lipid values were defined according to the NCEP/ATP III guidelines, which do not account for CV risk associated with RA. Some authors have previously attempted to include RA as an additional risk factor while defining therapeutic LDL cholesterol goals with NCEP/ATP III guidelines, which resulted in the more stringent treatment goals ([25]). However, we chose to use standard NCEP/ATP III guidelines to allow fair comparison of the cohorts based on the traditional CV risk factor profile.

Unfavorable changes in composition and impaired functional properties of lipids, in particular HDL, in patients with RA are the subject of active research ([24, 26]). Decrease in the atheroprotective effect of HDL has been suggested in patients with RA. Therefore, when defining LDL cholesterol goals, it is possible that the impact of HDL >60 mg/dl as a protective CV factor is different in RA versus non-RA subjects. However, the predictive value of changes in HDL properties on CV risk in RA is not fully understood and specific guidelines on defining LDL cholesterol goals in RA are lacking. Thus, a similar definition of therapeutic LDL cholesterol goals based on the NCEP/ATP III guidelines was applied to both patients with RA and non-RA subjects.

Statistical power may be limited in some analyses of associations between RA characteristics and LDL cholesterol goals, suggesting that these analyses should be interpreted with caution. Finally, during the period of investigation, the population of Olmsted County, Minnesota was predominantly white. Thus, the results may not be generalizable to nonwhite individuals.

To our knowledge, this is the first population-based study to report the impact of RA characteristics on the lipid-lowering effect of statins. Other important strengths of this study include its longitudinal population-based design, inclusion of 2 large National Institutes of Health–funded cohorts of patients with RA and non-RA subjects from the same underlying population, and the use of extensive data on RA characteristics and lipid measures available through the REP.

In conclusion, before statin initiation, patients with RA had significantly lower total cholesterol and LDL cholesterol levels than non-RA subjects. Absolute and relative decreases in LDL cholesterol values following statin initiation were smaller in RA than in non-RA subjects. Patients with RA were less likely to achieve therapeutic goals for LDL cholesterol compared to the non-RA subjects. Some RA disease characteristics, in particular higher disease activity as reflected in elevated ESR at baseline, may adversely affect the success of lipid lowering with statins in RA. More studies are needed to assess the mechanisms underlying the association of inflammation and lipid lowering in RA, and to determine the impact of improvement in lipid profile with statin use on CV risk reduction in RA.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Ms Crowson had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Myasoedova, Gabriel, Crowson.

Acquisition of data. Myasoedova, Gabriel, Crowson.

Analysis and interpretation of data. Myasoedova, Gabriel, Green, Matteson, Crowson.

ROLE OF THE STUDY SPONSOR

Pfizer had no role in the study design or in the collection, analysis, or interpretation of the data, the writing of the manuscript, or the decision to submit the manuscript for publication. Publication of this article was not contingent upon approval by Pfizer.

Ancillary