Association of Dyslipidemia in Moderate to Severe Meibomian Gland Dysfunction ADAM H. DAO, JORDAN D. SPINDLE, BLAKE A. HARP, ARUN JACOB, ALICE Z. CHUANG, AND RICHARD W. YEE ● PURPOSE:
To determine whether meibomian gland disease, a major contributor to dry eye syndrome, is associated with dyslipidemia. ● DESIGN: Retrospective case-control study. ● METHODS: SETTING: Clinical practice. PATIENT OR STUDY POPULATION: Sixty-six patients from January 2008 to July 2009 with moderate to severe meibomian gland disease whose serum lipid levels were obtained. We excluded patients who were already taking lipid-altering substances and patients with rheumatologic disease. We analyzed several parameters in prevalence of dyslipidemia (total cholesterol > 200 mg/dL, low-density lipoprotein [LDL] > 130 mg/dL, high-density lipoprotein [HDL] < 40 mg/dL, and triglycerides >150 mg/dL) in MGD patients and compared these patients to the general population as reported by data from the National Health and Nutrition Examination Survey (NHANES). MAIN OUTCOME MEASURE: The prevalence of dyslipidemia (elevated total cholesterol, elevated LDL, decreased HDL, or elevated triglycerides) in patients with moderate to severe MGD. ● RESULTS: Patients with moderate to severe MGD had a higher incidence of dyslipidemia with respect to elevated total cholesterol (>200 mg/dL), 67.4% to 45.1% (P ⴝ .0012) when compared to population controls. There was a smaller number of MGD patients with low HDL (HDL < 40 mg/dL), 6.5%, when compared to controls, 15.7% (P ⴝ .045). The incidence of increased LDL was not statistically significant (P ⴝ .184). There was a statistically smaller number of MGD patients with high triglycerides (TG > 150 mg/dL), 15.2%, when compared to controls, 33.1% (P ⴝ .0049). ● CONCLUSIONS: Patients with moderate to severe MGD have a higher incidence of dyslipidemia with respect to elevated total cholesterol than the general population. Surprisingly, the component of total cholesterol that contributed most to this increase in total cholesterol came from elevated serum HDL levels. To our knowledge, elevated HDL has not been associated with any pathologic state. Patients with MGD had a statistically Accepted for publication Apr 18, 2010. From the University of Texas Medical School at Houston, Department of Ophthalmology and Visual Sciences, Houston, Texas. Inquiries to Richard W. Yee, The University of Texas Medical School at Houston, Department of Ophthalmology and Visual Sciences, 6400 Fannin, 18th floor, Houston, TX 77030; e-mail: [email protected]
tmc.edu 0002-9394/$36.00 doi:10.1016/j.ajo.2010.04.016
significant lower incidence of hypoalphalipoproteinemia (low HDL) than the general population. Patients with MGD also had a lower incidence of hypertriglyceridemia than the general population. (Am J Ophthalmol 2010; 150:371–375. © 2010 by Elsevier Inc. All rights reserved.)
PHTHALMOLOGISTS RECOGNIZE THAT BLEPHARI-
tis, inflammation of the eyelid, is a common cause of dry eye syndrome. There are 2 types of blepharitis, anterior blepharitis and posterior blepharitis. Physical symptoms of both types of blepharitis include redness, foreign body sensation, itching, dry eye, crusting upon awakening, and blurred vision. Irritation of the outer eyelid is termed anterior blepharitis and is usually bacterial or parasitic in origin, while irritation of the inner eyelid is expressed as posterior blepharitis, caused primarily by the dysfunction of the meibomian glands. The meibomian glands are holocrine, tubulo-acinar structures that secrete meibum, a substance composed of lipids that forms the most anterior layer of the tear film, inhibits the evaporation of the underlying aqueous layer, and prevents the development of dry eye.1 A single row of meibomian glands are located along the lid margin of both the superior and inferior eyelids. The mucin and aqueous layers make up the remainder of the tear film. The mucin layer is secreted by goblet cells on the conjunctiva, which rests on the epithelial surface of the cornea and helps spread the aqueous layer evenly over the ocular surface. The aqueous layer is secreted by the lacrimal gland, represents 90% of the thickness of the tear film, and consists of a dilute solution of salts, minerals, and dissolved organic materials. An unstable tear film offers little protection, hydration, or lubrication to the ocular surface and is 1 of the major contributors to ocular surface disease. Meibomian gland dysfunction (MGD) results from an altered composition of meibum2 or from outflow obstruction of the gland.3 Altered composition of the lipid layer may result in uneven coverage of the aqueous layer. Obstruction prevents meibum secretion, decreasing the amount of lipid content of the tear film and subsequent protection against early evaporation and early tear breakup. Thus, the unprotected epithelium can become irritated and damaged, leading to ocular surface disease. Dysfunctional tear syndrome is an extremely common condition and has recently been defined to include lid margin
ELSEVIER INC. ALL
disease. Studies indicate that up to 20% of adults aged 45 years or more experience dry eye symptoms.4 Current treatment modalities include warm compresses, antibiotic ophthalmic drops, anti-inflammatory ophthalmic drops, and artificial tears. Recently, omega-3 fatty acid supplementation has been shown to be beneficial in the treatment of MGD.5 Despite these treatments, many patients do not experience complete resolution of their symptoms. Lipid research has been focused primarily on cardiovascular and cerebrovascular disease risk. Total cholesterol was found to be composed of 3 main types of lipoproteins—low-density lipoprotein (LDL), high-density lipoprotein (HDL), and very-low-density lipoprotein (VLDL). Patients with cholesterol and lipoprotein serum values outside the normal range are defined as having dyslipidemia. It is well understood from epidemiologic studies that having elevated levels of LDL or decreased levels of HDL, with or without other risk factors, increases the risk of a cardiovascular event.6 The prevalence of dyslipidemia in the general population is well described by current literature,7,8 which extrapolated data from the National Health and Nutrition Examination Survey (NHANES). The prevalence of total cholesterol greater than 200 is 45.1% and total cholesterol greater than 240 is 15.7%. The prevalence of LDL greater than 130 is 32.8%, HDL less than 40 is 15.5%, and triglycerides greater than 150 is 33.1%. The purpose of this study is to determine the association of MGD with serum lipid profiles.
gland dysfunction. We ordered lipid profiles on patients diagnosed with moderate to severe meibomian gland dysfunction at the time of diagnosis. In patients diagnosed with moderate to severe MGD, we included those 20 years of age or older and excluded those previously taking statins or omega-3 fatty acid supplements, which are known to alter lipid levels. We also excluded patients with known rheumatologic disease, as this may confound the causation of MGD. Age, sex, cholesterol-lowering medications, total cholesterol, HDL, LDL, and triglycerides were recorded. The percentage of patients with total cholesterol greater than 200 mg/dL, HDL less than 40 mg/dL, LDL greater than 130 mg/dL, and triglyceride levels greater than 150 mg/dL were determined and compared to historical controls.6,7 ● STATISTICAL METHODS:
The percentage of MGD patients whose total cholesterol level greater than or equal to 200 mg/dL was determined and a 1-sample z test was calculated to determine whether this percentage was significantly different than in the general population. Additionally, we stratified our patients by age and sex to compare to controls. Similarly, the percentages and z tests were used for the secondary outcome variables (LDL, HDL, and triglycerides). We define statistically significant as a P value of less than .05.
RESULTS MATERIALS AND METHODS
BASELINE LIPID LEVELS WERE RECORDED FOR 63 PATIENTS
with MGD. Of these, 2 patients were excluded because of rheumatologic disease and 15 patients were excluded from the analysis because of their taking either an omega-3 fatty acid supplement (n ⫽ 6) or a statin (n ⫽ 9). Of the 46 remaining patients, 22 were male and 24 were female, the average age was 52.0 years, and the range in age was 27 to 82. We stratified our patients into age brackets (age 20 – 44, n ⫽ 17 [male n ⫽ 8, female n ⫽ 9]; age 45– 64, n ⫽ 18 [male n ⫽ 7, female n ⫽ 11]; age 65⫹, n ⫽ 11 [male n ⫽ 7, female n ⫽ 4]) to compare to historical controls gathered from the NHANES, which was population age-adjusted (average age ⫽ 46.6). We found a statistically significant increase in the proportion of patients with elevated total cholesterol (total cholesterol ⬎ 200 mg/dL) (Figure 1), 67.39% (31/46) when compared to controls, 45.1% (P ⫽ .0012). There was a statistically smaller number of MGD patients with low HDL (HDL ⬍ 40 mg/dL), 6.5% (3/46), when compared to controls, 15.7% (P ⫽ .045). There was an elevation prevalence of elevated serum LDL (LDL ⬎ 130 mg/dL) between patients with MGD and those in the general population (39.1% [18/46] and 32.8%), but this was not statistically significant (P ⫽ .184). There was a statistically smaller number of MGD patients with high triglycerides
WE REVIEWED THE CHARTS OF PATIENTS IN OUR CLINIC
who were diagnosed with MGD since January 1, 2008. Treatment of our patients with MGD often includes the supplementation of 2000 to 4000 mg of an omega-3 fatty acid daily. Omega-3 fatty acid supplementation is known to alter lipid levels; therefore, we routinely order a baseline lipid panel, which includes HDL concentration, LDL concentration, triglycerides, and total cholesterol. Patients were diagnosed with MGD by clinical examination based on glandular obstruction and meibum quality. This grading was obtained by gently applying pressure with a cotton swab tip to the upper and lower eyelid while observing the ease of excretion and quality of the meibum under slit-lamp microscopy. Currently, there are no well-accepted clinical guidelines on describing meibomian gland obstruction or quality. We use a modified approach of Foulkes and Bron9 to grade meibomian gland obstruction on a scale from 0 (no obstruction) to 4 (complete obstruction). Meibum quality was graded on a scale from 0 (clear) to 4 (toothpaste-like). Patients graded 3 or higher on either of these scales in either eye during their baseline meibomian gland evaluation were diagnosed with moderate to severe meibomian 372
FIGURE 1. Prevalence of dyslipidemia in patients with meibomian gland dysfunction compared to normal population historical controls by lipid type. The differences between total cholesterol (P ⴝ .0012), high-density lipoprotein (P ⴝ 045), and triglycerides (P ⴝ .0049) were significant while the difference between low-density lipoprotein (P ⴝ .184) was not.
FIGURE 2. Prevalence of dyslipidemia with respect to elevated total cholesterol in our patients with meibomian gland dysfunction stratified by age compared to controls of normal population historical controls. The differences among those aged 20 to 44 (P ⴝ .084) and 45 to 64 (P ⴝ .012) was statistically significant* while the difference in those above 65 years of age (P ⴝ .46) was not.
(TG ⬎ 150 mg/dL), 15.2% (7/46), when compared to controls, 33.1% (P ⫽ .0049). In patients with moderate to severe MGD stratified according to age groups defined in our historical control, we found a statistically significant difference in 1 of 3 age groups (Figure 2). In patients aged 45 to 64 years (n ⫽ 18), 83.3% (15/18) of patients had elevated total cholesterol, vs 57.0% of controls (P ⫽ .012). In patients aged 20 to 44 years (n ⫽ 17), 64.7% (11/17) of patients had elevated total cholesterol, vs 48.0% of controls (P ⫽ .084). In patients aged 65⫹ (n ⫽ 11), 45.5% (5/11) of patients had elevated total cholesterol, vs 46.9% of controls (P ⫽ .46). We also stratified our patients with moderate to severe MGD according to sex and age and compared these VOL. 150, NO. 3
numbers to the historical control (Figure 3). In male patients, the prevalence of high cholesterol in those between the ages of 20 and 44 (n ⫽ 8) was statistically significantly increased relative to population controls (87.5% [7/8] to 41.2%, P ⫽ .0039). The prevalence was also increased in those men aged 45 to 64 (n ⫽ 7) (85.7% [6/7] to 52.9%, P ⫽ .041), and in those aged 65 years and older, it was increased (n ⫽ 7) (42.8% [3/7] to 36.3%, P ⫽ .46), although not statistically significantly. In female patients (Figure 4), the prevalence of high cholesterol in those between the ages of 20 and 44 (n ⫽ 9) was not statistically significantly increased relative to population controls (44.4% [4/9] to 34.9%, P ⫽ .22). The prevalence was not statistically significantly increased in MEIBOMIAN GLAND DYSFUNCTION
FIGURE 3. Prevalence of dyslipidemia with respect to elevated total cholesterol in male patients with meibomian gland dysfunction stratified by age compared to normal population historical controls. The differences among men aged 20 to 44 (P ⴝ .0039) and 45 to 64 (P ⴝ .041) were statistically significant* while the difference in men above 65 years of age (P ⴝ .46) was not.
FIGURE 4. Prevalence of dyslipidemia with respect to elevated total cholesterol in female patients with meibomian gland dysfunction stratified by age compared to normal population historical controls. The differences among women aged 45 to 64 (P ⴝ .078) were statistically significant* while the difference between women aged 20 to 44 (P ⴝ .22) and women above 65 years of age (P ⴝ .42) was not.
women aged 45 to 64 (n ⫽ 11) (81.8% [9/11] to 60.9%, P ⫽ .078). In those aged 65 years and older (n ⫽ 4), there was no statistical difference (50.0% [2/4] vs 55.1% in controls, P ⫽ .42).
level has been studied and hypothesized to be a factor in meibomian gland dysfunction.3,12 Because normal whole meibomian lipids have a melting point of 30°C to 34°C10 and cholesterol has a melting point of 148°C,10 increased concentration of cholesterol in meibomian lipid would increase the melting point of the meibomian lipid milieu, theoretically increasing viscosity and leading to plugging of the meibomian glands.3 The composition of normal meibum secretions has been well studied and cholesterol content has been reported to be 1% to 2%;11 however, to our knowledge, meibum compo-
DISCUSSION PATIENTS WITH MODERATE TO SEVERE MGD HAVE IN-
creased levels of total cholesterol compared to population controls. Increased cholesterol at the glandular secretion 374
sition in patients with meibomian gland dysfunction has not been studied. One study attempted to associate serum cholesterol and human tear fluid and found that cholesterol concentration in the tear film bore no correlation to serum cholesterol.13 However, this research measured the concentration of cholesterol in the aggregate tear film. It did not measure this concentration in the lipid layer, in the meibum, or in the meibomian gland. Altered meibomian lipid concentration, including an increase in meibum cholesterol, has been shown to cause dry eye symptoms. In another study, patients on antiandrogen therapy were found to have an altered meibomian lipid composition with a notable increase in the quantity of cholesterol.12 These patients experienced a subjective increase in dry eye symptoms, including light sensitivity, painful eyes, and blurred vision. Under biomicroscopy, the patients were found to have an increase in tear film debris, the presence of an abnormal tear film meniscus, irregular posterior lid margins, increased vital dye stain, conjunctival injection, and a significantly decreased tear break-up time, all signs of dry eye disease.12 Furthermore, it has been shown that patients with Sjögren syndrome have increased cholesterol content in their tear film.14 We found a statistically significant decreased prevalence of patients with low HDL in patients with moderate to severe MGD relative to the population. The cause of this
is unknown, but our findings suggest that elevated HDL may be a risk factor for the development of meibomian gland disease, despite its cardioprotective effects. More studies, both prospective and at the basic science level, will be needed to examine the significance of this finding. We acknowledge the limitations to our study. Our patient population was slightly older relative to the population studied in NHANES (age 52.2 vs 46.6). This may have contributed to some of the increase in our prevalence of increased total cholesterol; however, when stratified according to age brackets, correlation between high prevalence of elevated serum cholesterol levels and MGD were still present. Statistical significance may have been achieved if our population size was larger. Our study has the inherent flaws of a retrospective design. As a result, we were unable to age-match our patient population to those of the historical controls. A larger, prospective study will be needed to better control these variables and possibly achieve statistical significance in more age groups. Our findings suggest that ophthalmologists may be instrumental in the early detection of dyslipidemia in the general population. The ability to screen for lipid abnormalities by physical examination at the slit lamp is exciting. Further prospective studies should be conducted to justify early screening of patients with meibomian gland dysfunction for dyslipidemia.
THE AUTHORS INDICATE NO FINANCIAL SUPPORT OR FINANCIAL CONFLICT OF INTEREST. INVOLVED IN STUDY CONCEPT and design (A.D., J.S., R.Y.); acquisition (A.D., J.S., B.A., R.Y.), management (R.Y.), analysis, and interpretation of data (A.D., J.S., B.A., A.J., R.Y., A.C.); and preparation (A.D., J.S., R.Y.) and review of the manuscript (all authors). Prior to data collection and review, approval was obtained from the University of Texas at Houston Committee for the Protection of Human Subjects, the authors’ institutional review board. Adam H. Dao and Jordan D. Spindle contributed equally to this work.
REFERENCES 1. Bron AJ, Benjamin L, Snibson R. Meibomian gland disease: classification and grading of lid changes. Eye 1991;5:395– 411. 2. Shields WJ, Mathers WD, Roberts J, et al. Criteria for the evaluation of lid margin in blepharitis. Invest Ophthalmol Vis Sci 1990;31:483. 3. Driver PJ, Lemp MA. Major review: meibomian gland dysfunction. Surv Ophthalmol 1996;40(5):343. 4. Brewitt H, Sistani F. Dry eye disease: the scale of the problem. Surv Ophthalmol 2001;45(Suppl 2):S199 –202. 5. Macsai M. The role of omega-3 dietary supplementation in blepharitis and meibomian gland dysfunction. Trans Am Ophthalmol Soc 2008;106:336 –356. 6. Wilson PW, D’Agostino RB, Levy D, et al. Prediction of coronary heart disease using risk factor categories. Circulation 1998;97:1837. 7. Lloyd-Jones D, Adams R, Carnethon M, et al. Heart disease and stroke statistics—2009 update: a report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation 2009;119:e21–181.
VOL. 150, NO. 3
8. Ford E, Li C, Zhao G, Pearson WS, Mokdad AH. Hypertriglyceridemia and its pharmacologic treatment among US adults. Arch Intern Med 2009;169(6):572–578. 9. Foulks GN, Bron AJ. Meibomian gland dysfunction: a clinical scheme for description, diagnosis, classification, and grading. Ocul Surf 2003;1(3):107–126. 10. Butovich IA, Millar TJ, Ham BM. Understanding and analyzing meibomian lipids—a review. Curr Eye Res 2008; 33(5):405– 420. 11. Butovich IA. The meibomian puzzle: combining pieces together. Prog Retin Eye Res 2009;28(6):483– 498. 12. Krenzer KL, Dana MR, Ullman MD, et al. Effect of androgen deficiency on the human meibomian gland and ocular surface. Clin Endocrinol Metab 2000;85(12):4874 – 4882. 13. Morozova RP, Zhaboiedov HD, Kirieiev VV, Nikolenko IA. The normal tear fluid and decreased tearing in patients with Sjogren’s disease and Sjogren’s syndrome II. The lipid content. Ukr Biokhim Zh 1996;68:82– 87. 14. Van Haeringen NJ, Glasius E. Cholesterol in human tear fluid. Exp Eye Res 1975;20:271–274.
MEIBOMIAN GLAND DYSFUNCTION
Biosketch Adam H. Dao, MD, graduated from the University of Texas Medical School at Houston in May 2010. He will be starting an Ophthalmology Residency at the University of South Carolina in 2011.
VOL. 150, NO. 3
MEIBOMIAN GLAND DYSFUNCTION