Seasonal Variation in Lipids in Patients Following Acute Coronary Syndrome on Fixed Doses of Pravastatin (40 mg) or Atorvastatin (80 mg) (from the Pravastatin or Atorvastatin Evaluation and Infection Therapy–Thrombolysis In Myocardial Infarction 22 [PROVE IT-TIMI 22] Study)
Article Outline
Previous studies have shown seasonal variation in lipids. To understand whether this variation exists in patients with acute coronary syndromes receiving statins, we examined data from the PROVE IT-TIMI 22 Study. At baseline, no significant difference in low-density lipoprotein (LDL) cholesterol was observed when stratified by season. However, a statistically significant difference in high-density lipoprotein cholesterol between winter (37 mg/dl) and summer (39 mg/dl) was observed (p <0.001) at baseline. On treatment, median LDL cholesterol was 102 mg/dl in winter versus 96 mg/dl in summer (p <0.001) for the pravastatin group and 68 mg/dl in winter versus 62 mg/dl in summer (p <0.001) for the atorvastatin group. Median high-density lipoprotein cholesterol was 43 mg/dl in summer versus 41 mg/dl in winter in the pravastatin group and 42 mg/dl in summer versus 39 mg/dl in winter in the atorvastatin group (p <0.001). More patients achieved LDL cholesterol <100 mg/dl in summer at 56% versus 47% in winter in the pravastatin group (p <0.001) and 89% versus 87% in winter for the atorvastatin group (p = 0.11). Achievement of LDL cholesterol <70 mg/dl was also higher in summer than winter. In conclusion, this was the first evidence of seasonal variability in cholesterol in patients with acute coronary syndromes treated with statins. This variability affected achievement of National Cholesterol Education Program goals and may affect management decisions based on season of collection.
Seasonal variations in myocardial ischemic events have been reported, with a peak in winter and nadir in summer in both northern and southern hemispheres.1, 2, 3, 4 During the past 5 decades, small longitudinal and larger cross-sectional studies have shown seasonal variation in serum lipids that appeared to be independent of ambient temperature, diet, and physical activity.1, 2, 3, 5, 6, 7 However, many reported differences in the timing of peak and trough cholesterol and a review of 19 early studies raised method concerns and questioned the evidence for seasonal cholesterol fluctuation.8 Moreover, although Manttari et al5 reported seasonal variation in high-density lipoprotein (HDL) cholesterol in patients treated with gemfibrozil, many studies included healthy volunteers or subjects with untreated dyslipidemia. To understand whether this variation exists in patients with active coronary disease treated with statins, we examined serum lipids in the Pravastatin or Atorvastatin Evaluation and Infection Therapy–Thrombolysis In Myocardial Infarction 22 (PROVE IT-TIMI 22), a large multicenter trial comparing atorvastatin 80 mg with pravastatin 40 mg in patients with acute coronary syndromes (ACSs).
Methods
A total of 4,162 subjects stabilized from ACSs were enrolled in PROVE IT-TIMI 22, with 2,099 subjects randomly assigned to intensive therapy (atorvastatin 80 mg) and 2,063 randomly assigned to moderate therapy (pravastatin 40 mg).9, 10 Patients were enrolled within 10 days of presentation and were eligible if total cholesterol (TC) was <240 or <200 mg/dl if already treated with cholesterol-lowering therapy, including statins. Patients were followed up for a mean of 2 years. End points were adjudicated by an independent clinical events committee. Definitions of trial end points have been described previously.9, 10 Primary analyses included both the moderate- and intensive-treatment arms.
Seasons were defined using a light seasons model centered on the equinoxes. Accordingly, winter was defined as November 6 to February 4; spring, February 5 to May 6; summer, May 7 to August 5; and fall, August 6 to November 5. Seasons were reversed for subjects enrolled in the southern hemisphere. Lipids measured before therapy were used for baseline measures. For the present report, on-treatment analyses were confined to lipids measured ≥4 months after enrollment because subjects would have reached steady state after this time. All lipids obtained at each visit from ≥4 months for a given patient were included in analyses. Therefore, a given patient would contribute data to >1 season.
Continuous lipid values were reported as median and compared using Kruskal-Wallis test. Comparisons among percentages of patients who achieved the low-density lipoprotein (LDL) cholesterol goal by season were made using chi-square test for heterogeneity. All on-treatment analyses were stratified by randomized treatment group. Two-tailed p <0.05 was considered significant. Seasonal differences in adverse events were not examined because of timing constraints of a clinical trial, in which patients were most likely to have recurrent clinical events in the period immediately after ACS, thereby limiting the utility of stratifying events by season.
Results
A total of 4,162 subjects were enrolled in the study and 18,075 lipid measurements were obtained. Baseline characteristics of this study population have been described elsewhere.10 Median baseline TC for all subjects was 180 mg/dl and showed minimal baseline differences stratified by season. Median TC was 179 mg/dl in winter, 182 mg/dl in spring, 182 mg/dl in summer, and 179 mg/dl in fall. Median baseline LDL cholesterol for all subjects was 106 mg/dl and showed no statistically significant difference stratified by season. Median LDL cholesterol was 109 mg/dl in winter, 106 mg/dl in spring, 105 mg/dl in summer, and 106 mg/dl in fall (p = NS). Median baseline HDL cholesterol for all subjects was 38 mg/dl and showed a statistically significant difference between winter and summer at baseline. Median HDL cholesterol was 37 mg/dl in winter, 40 mg/dl in spring, 39 mg/dl in summer, and 36 mg/dl in fall (p <0.001). Figure 1 shows baseline LDL and HDL cholesterol by season.
Figure 1.
Baseline LDL (LDL-C) and HDL cholesterol (HDL-C) by season.
A total of 14,099 measurements from patients on lipid-lowering therapy were obtained for this analysis. This included 2,838 in winter, 2,959 in spring, 3,049 in summer, and 5,253 in fall. Median TC was 174 mg/dl for the pravastatin group and 135 mg/dl for the atorvastatin group. Stratified by season, TC was highest in winter and lowest in summer. In the pravastatin 40-mg group, median LDL cholesterol after ≥4 months of treatment was 98 mg/dl versus 66 in the atorvastatin 80-mg group. Stratified by season, median LDL cholesterol was 6% higher (102 vs 96 mg/dl; p <0.001) for the pravastatin group and 10% higher (68 vs 62 mg/dl; p <0.001) in the atorvastatin group in winter than summer. Median LDL cholesterol by treatment group and season is shown in Figure 2. Median HDL cholesterol after ≥4 months of treatment was 42 mg/dl for the pravastatin group and 41 mg/dl for the atorvastatin group. In contrast to other lipid parameters, HDL cholesterol was higher in summer in both groups. In the pravastatin 40-mg group, HDL cholesterol was 41 mg/dl in winter and 43 mg/dl in summer (p <0.001). For the atorvastatin 80-mg group, HDL cholesterol was 39 mg/dl in winter versus 42 mg/dl in summer (p <0.001). Median HDL cholesterol by treatment group and season is shown in Figure 2. The ratio of TC to HDL cholesterol followed a seasonal pattern, with peak values in winter and lowest values in summer. For the pravastatin 40-mg therapy group, TC/HDL cholesterol ratio was 4.2 in winter and 3.9 in summer (p <0.001). For the atorvastatin 80-mg group, TC/HDL cholesterol ratio was 3.4 in winter and 3.1 in summer (p <0.001).
Figure 2.
On-treatment (≥4 months) median (A) LDL cholesterol (milligrams per deciliter) and (B) HDL cholesterol (milligrams per deciliter) by season.
This variability in lipids resulted in a statistically significant difference in the percentage of decrease in LDL cholesterol from baseline by season. The decrease in LDL cholesterol from baseline after ≥4 months of treatment with pravastatin 40 mg was 3% in winter compared with 11% in summer (p <0.001). In the atorvastatin 80-mg group, the decrease in LDL cholesterol was 37% in winter versus 41% in summer (p <0.001). Change in LDL cholesterol from post-ACS baseline is shown in Figure 3.
Figure 3.
Median percentage of decrease in LDL cholesterol (≥4 months vs baseline).
When treatment targets were examined by season, rates of achievement were greater in summer than winter. In both treatment groups, more patients met the treatment goal of LDL cholesterol <100 mg/dl in summer than in all other seasons. In the pravastatin 40-mg group, 47% achieved LDL cholesterol <100 mg/dl in winter compared with 56% in summer (p <0.001). In the atorvastatin 80-mg group, 87% met treatment goals of LDL cholesterol <100 mg/dl in winter compared with 89% in summer (p = 0.11). Based in part on findings from PROVE IT-TIMI 22, the most recent American Heart Association Scientific Statement included an optional target of LDL cholesterol <70 mg/dl for patients with coronary artery disease at high risk, including after ACS.11 In the pravastatin 40-mg group, 11% achieved LDL cholesterol <70 mg/dl in winter versus 14% in summer (p = 0.02). Similarly, in the atorvastatin 80-mg group, 53% of subjects achieved LDL cholesterol <70 mg/dl in winter compared with 63% in summer (p <0.001). Figure 4 shows the percentage of patients achieving both treatment goals by season.
Figure 4.
On-treatment (≥4 months) percentages achieving LDL cholesterol <100 and <70 mg/dl.
Discussion
This analysis provided the first evidence of seasonal variation in lipids in patients treated with statin therapy. This study evaluated 2 statins of differing potency, and despite significant improvements in LDL cholesterol and other lipids, seasonal variability was noted in both the moderate- and intensive-therapy groups. Both groups showed decreases from baseline LDL cholesterol that were more favorable in summer than winter, and as such, differential achievement of National Cholesterol Education Program treatment goals by season.
Previous data have indicated that a difference in cardiac risk profiles existed in summer versus winter. Studies examining the role of body mass index, diet, and exercise reported that these variables did not fully explain the variation between seasons.1, 12 Although seasonal variation in cholesterol has been considered a factor, concerns regarding the methods of early studies questioned the evidence for a seasonal pattern of lipids.8 These concerns included the small numbers of subjects and inconsistent patterns of seasonal variation.8 Subsequent studies supported a consistent seasonal pattern in lipids. However, until this analysis, it was not known whether these differences existed in patients with ACS or those treated with lipid-lowering therapy.
The Cholesterol Treatment Trialists' Collaborators conducted a meta-analysis of >90,000 subjects in 14 randomized controlled clinical trials to assess the effect on different clinical outcomes of a 1-mmol/L (39 mg/dl) decrease in LDL cholesterol.13 This analysis showed a decrease in important vascular events that was directly proportional to the absolute decrease in LDL cholesterol. In particular, they found a 12% decrease in all-cause mortality and 19% decrease in coronary mortality for each 1-mmol/L decrease in LDL cholesterol.13 Based on this finding, even small seasonal differences in LDL cholesterol may therefore be clinically significant.
The relation of LDL cholesterol and cardiac risk combined with the understanding of seasonal variation in cholesterol has potential implications for patient management. Although this analysis did not examine seasonal differences in adverse events, increases in LDL cholesterol may be important in clinical practice. For example, if a patient is only slightly lower than a treatment goal in summer, they are likely to be higher than goal in winter, and consideration may be given to intensification of therapy. The seasonal variation in lipids may also affect the diagnosis and treatment of dyslipidemia under existing lipid guidelines because the National Cholesterol Education Program guidelines do not currently consider seasonal variation in cholesterol.
The present analysis extended observations regarding the seasonal variability in lipids in several important ways. First, the large study size, frequency of measurement, and use of a central laboratory with standard measurement technique confirmed previous epidemiologic and observational studies. Second, it provided clear evidence of the seasonal variability in lipids in patients with ACS treated with statins, which was not previously known. Finally, it showed seasonal variability in the achievement of clinical targets.
This was a post hoc analysis, and all hypotheses based on the results should be considered exploratory. Because clinical trial enrollment was not uniformly distributed over a single year, there were an unequal number of patients enrolled in any given season. In addition, because this trial was not designed to assess seasonal variability in lipids, limited information regarding diet, exercise, and other variables that may impact on lipids was obtained and did not allow for further mechanistic explanation of these findings. Additional studies are needed to determine whether diagnosis and treatment of hyperlipidemia based on multiple seasonal data points would result in improved outcomes and to obtain insight into seasonal differences in ischemic events.
References
- . Seasonal variation in plasma cholesterol distributions: implications for screening and referral. Am J Prev Med. 1992;8:360–366
- . Seasonal variation of serum lipids in an elderly population. Age Ageing. 1993;22:273–278
- . Seasonal cholesterol cycles: the Lipid Research Clinics Coronary Primary Prevention Trial placebo group. Circulation. 1987;76:1224–1231
- . Mechanism of seasonal variation in acute myocardial infarction. Int J Cardiol. 2005;100:163–164
- . Seasonal variation in high density lipoprotein cholesterol. Atherosclerosis. 1993;100:257–265
- . Seasonal variation in parameters related to coronary heart disease risk in young men. Atherosclerosis. 1996;126:117–129
- . Cyclic seasonal variation in plasma lipid and lipoprotein levels: the Lipid Research Clinics Coronary Primary Prevention Trial Placebo Group. J Clin Epidemiol. 1988;41:679–689
- . Overview: seasonal variations in cholesterol. Prev Med. 1981;10:655–659
- . Design of the Pravastatin or Atorvastatin Evaluation and Infection Therapy (PROVE IT)-TIMI 22 trial. Am J Cardiol. 2002;89:860–861
- . Intensive versus moderate lipid lowering with statins after acute coronary syndromes. N Engl J Med. 2004;350:1495–1504
- . Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation. 2005;112:2735–2752
- . Seasonal variations of serum lipids and apoproteins. Ann Nutr Metab. 1988;32:68–74
- . Efficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from 90,056 participants in 14 randomised trials of statins. Lancet. 2005;366:1267–1278
The PROVE IT-TIMI 22 Study was supported by a research grant from Bristol-Myers Squibb, New York, New York. Dr. Wiviott has received honoraria for speaking from Pfizer, New York, New York, and Merck and Co., Whitehouse Station, New Jersey, and consulting fees from Sanofi-Aventis, Bridgewater, New Jersey. Dr. Cannon currently receives research grants/support from Accumetrics, San Diego, California, AstraZeneca, Wilmington, Delaware, Bristol-Myers Squibb/Sanofi Partnership, New York, New York, Glaxo Smith Kline Brentford, Middlesex, United Kingdom, and Merck/Schering Plough Partnership, Kenilworth, New Jersey.
PII: S0002-9149(09)00042-3
doi:10.1016/j.amjcard.2008.12.034
© 2009 Elsevier Inc. All rights reserved.
