Triglyceride/High-Density Lipoprotein Cholesterol Ratio: A Surrogate to Predict Insulin Resistance and Low-Density Lipoprotein Cholesterol Particle Size in Nondiabetic Patients With Schizophrenia
Objective: Insulin resistance, changes in lipid parameters, and cardiometabolic adverse events have been reported in some patients during clinical trials of antipsychotic agents. The present study examined whether the triglyceride/high-density lipoprotein (HDL) ratio can be used as a better surrogate than other conventional lipid measures (low-density lipoprotein [LDL], HDL, triglyceride) in predicting insulin resistance and LDL particle size in nondiabetic patients with schizophrenia.
Method: Outpatients 18 to 75 years old diagnosed with schizophrenia or schizoaffective disorder (DSM-IV criteria) and receiving olanzapine, risperidone, or typical antipsychotics participated in a multicenter, cross-sectional study. Fasting blood samples were obtained to determine the levels of glucose, insulin, lipids, and lipid particle size. The study was conducted from July 2001 to March 2002.
Results: In the sample of 206 patients, significant correlations were found between various lipid measures (LDL, HDL, triglyceride, and triglyceride/HDL ratio) and the homeostasis model of assessment of insulin resistance (P < .05). However, stepwise multiple regression analysis suggested that the triglyceride/HDL ratio is a stronger predictor of insulin resistance and of LDL particle size than other conventional lipoprotein measures after other potential confounding variables, including age, gender, race, family history of diabetes, body mass index, and antipsychotic agent, were taken into consideration (P < .001). Further, logistic regression analysis indicated that the triglyceride/HDL ratio and male gender predict the existence of a small LDL particle size pattern (pattern B LDL phenotype), with a sensitivity of 75.9% and a specificity of 85.4%.
Conclusions: The triglyceride/HDL ratio, a simple, readily available and inexpensive measure, can be a useful surrogate to identify those with insulin resistance as well as those with more atherogenic small LDL particles in nondiabetic patients with schizophrenia.
J Clin Psychiatry 2011;72(6):806-812
© Copyright 2010 Physicians Postgraduate Press, Inc.
Submitted: February 2, 2009; accepted November 19, 2009.
Online ahead of print: November 2, 2010 (doi:10.4088/JCP.09m05107yel).
Corresponding author: Xiaoduo Fan, MD, MS, Freedom Trail Clinic, 25 Staniford St, Boston, MA 02114 ([email protected]).
Mortality of schizophrenia patients is approximately twice that of the general population.1,2 Cardiovascular disease is responsible for as much as 50% of the excess mortality associated with schizophrenia.2 Elevated cardiovascular risk in the schizophrenia population may be attributable to numerous factors, but one source of elevated cardiovascular risk is a cluster of clinical features defining the metabolic syndrome: abdominal adiposity, atherogenic dyslipidemia, hypertension, and impaired fasting glucose or overt diabetes.3 Recent data from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) suggested that the metabolic syndrome is highly prevalent in US schizophrenia patients: 40.9% using the National Cholesterol Education Program (NCEP) criteria and 42.7% using the updated American Heart Association criteria that include a lower fasting glucose threshold of 100 mg/dL.4 In contrast, the age-adjusted prevalence of the metabolic syndrome from the third National Health and Nutrition Examination Survey (NHANES III) was 23.7% using the NCEP criteria.5
There is strong evidence of an association between insulin resistance and diabetes.6-10 Insulin resistance refers to the diminished ability of insulin to lower blood glucose, a consequent result to either decreased cellular glucose uptake or increased hepatic glucose output. When insulin resistance occurs, the body attempts to overcome the resistance by secreting more insulin, resulting in hyperinsulinemia. Insulin resistance has been reported in some patients during clinical trials of antipsychotic agents.11,12 Beyond its role in the development of type 2 diabetes, insulin resistance has been associated with cardiovascular disease.13-15
Hyperlipidemia has been reported during treatment with atypical antipsychotics. A 5-year naturalistic study of patients treated with clozapine found a nonsignificant increase in fasting total cholesterol but a significant increase in fasting triglyceride level.16 In a randomized, double-blind trial, patients treated with olanzapine and clozapine had significant increases in fasting total cholesterol level.17 In a randomized 8-week study of 112 patients with first-episode schizophrenia, fasting total cholesterol and triglyceride levels were significantly increased in those treated with clozapine or olanzapine.18 In a large-scale, matched case-control analysis with pharmacy and claims data, an increased incidence of hyperlipidemia was reported in patients with schizophrenia and mood disorders treated with clozapine, olanzapine, risperidone, quetiapine, ziprasidone, and first-generation antipsychotic agents, but not aripiprazole, compared to patients not treated with an antipsychotic agent.19 The CATIE study found a difference in association with lipid changes among the 5 antipsychotics studied. The study showed an association with greater mean increases in levels of triglycerides and total cholesterol in olanzapine-treated patients followed by quetiapine-treated patients. Patients treated with risperidone and ziprasidone experienced a mean decrease in these measures, and patients treated with perphenazine experienced a modest mean increase in levels of triglycerides and total cholesterol.20
Elevated low-density lipoprotein (LDL) cholesterol has long been considered as a major cause of cardiovascular disease.21 Low-density lipoprotein cholesterol comprises a heterogeneous spectrum of lipoprotein particles. The size of LDL particles may play distinct roles in determining its atherogenicity. The predominance of small, dense LDL particles (pattern B LDL phenotype) has been reported to be particularly atherogenic.22-24 Several mechanisms have been proposed to explain this association: small, dense LDLs have low affinity to the LDL receptor, low resistance to oxidative stress, prolonged plasma half-life, high binding affinity to surface components in the vessel wall, and efficient penetration into the intima. Despite the clinical importance of recognizing those individuals with a small LDL particle size pattern, LDL particle size determination is not universally available, is expensive, and has not been widely applied in clinical practice. Within the schizophrenia population, studies measuring LDL particle size are still rare.25
Low levels of high-density lipoprotein (HDL) cholesterol and elevated triglycerides are risk factors for insulin resistance; further, elevations of triglycerides are associated with the formation of small LDL particles.21 The simple, readily calculated measure of triglyceride/HDL cholesterol ratio has been identified as a predictor of insulin resistance26-28 and cardiovascular disease29,30 in various populations. Studies have also suggested that the triglyceride/HDL ratio is a useful marker to predict the presence of small, dense LDL particles.26,31-33
This multicenter, cross-sectional study was designed to assess metabolic health in nondiabetic patients with schizophrenia or schizoaffective disorder. The findings comparing different antipsychotic agents (olanzapine, risperidone, or typical agents) with regard to various metabolic measures including lipids have been reported elsewhere.25 The focus of the present analysis was to examine whether the triglyceride/HDL ratio can be used as a better surrogate than other conventional lipid measures (LDL, HDL, triglyceride) in predicting insulin resistance and LDL particle size in nondiabetic patients with schizophrenia.
METHOD
Subjects
Outpatients 18 to 75 years old diagnosed with schizophrenia or schizoaffective disorder (DSM-IV criteria; duration ≥ 5 years) were recruited from 28 study sites in the United States. Patients were psychiatrically stable, with no hospitalizations within the previous 3 months. Patients had been receiving olanzapine, risperidone, or typical antipsychotic agents continuously for at least 1 year and as monotherapy for the 3 months preceding the study. Exclusion criteria included unstable medical illnesses, substance dependence within the previous 3 months, known medical conditions that might affect changes in metabolic parameters, known history of diabetes or lipid disorder, and use of antidiabetic or lipid-lowering therapy or special diets to lower glucose or lipids levels. Patients on exercise programs aimed at lowering glucose or lipid levels were not excluded from the study. All procedures were approved by institutional review boards at all study sites and were conducted in compliance with the Declaration of Helsinki. The study was conducted from July 2001 to March 2002.
Procedures
After providing written informed consent, each patient underwent a physical examination and a psychiatric diagnostic evaluation, which included the Brief Psychiatric Rating Scale (BPRS) assessment. Eligible patients were admitted to an inpatient facility for an observed overnight fast. Patients received a low-fat meal and were not allowed to ingest alcoholic beverages. Patients were allowed to drink water during the fasting period. Vital signs and anthropometric measures were obtained in the beginning of the inpatient stay. Blood samples were obtained 11 hours (± 1 hour) after the completion of the meal the night before.
Blood analyses were performed at the Covance Laboratories (Princeton, New Jersey). Fasting serum levels of glucose, total cholesterol, HDL, and triglycerides were analyzed using standard enzymatic methods and an automated analyzer (Roche Modular Analyzer, Roche Diagnostics, Indianapolis, Indiana). The triglyceride/HDL ratio was derived based on fasting serum levels of triglycerides and HDL.26,27 Low-density lipoprotein levels were determined by the Direct LDL reagents (Roche Diagnostics, Indianapolis, Indiana). Coefficients of variations for inter- and intra-assay variability for glucose and lipids were < 5%. Fasting serum insulin levels were analyzed using IMMULITE 2000 (Diagnostic Products Corporation, Los Angeles, California). Coefficients of variations for inter- and intra-assay variability for serum insulin were < 9%. Fasting plasma glucose and serum insulin levels were averaged from samples drawn 15 minutes apart. The homeostasis model of assessment of insulin resistance (HOMA-IR) was calculated by the following formula: fasting plasma glucose (mg/dL) ח fasting serum insulin (μU/mL)/405.34
Low-density lipoprotein mean particle size was determined using nuclear magnetic resonance spectroscopy (LipoScience, Raleigh, North Carolina).35,36 Nuclear magnetic resonance spectroscopy capitalizes on the fact that each lipoprotein subclass particle of a given size emits its own characteristic signal, and the amplitude of the signal provides a direct measure of the lipoprotein particle concentration. Low-density lipoproteins are grouped by mean particle size as large (20.6-23.1 nm, pattern A LDL phenotype) and small (18.0-20.5 nm, pattern B LDL phenotype).
Statistical Analysis
The data were analyzed using SPSS (version 13.0; SPSS Inc, Chicago, Illinois). Descriptive statistics described demographics, clinical measures, and laboratory values. The homeostasis model of assessment of insulin resistance, triglyceride, triglyceride/HDL ratio, and LDL mean particle size were not distributed normally and were, therefore, log transformed before analysis. Pearson correlation coefficients were used to quantify relations between lipid measures and other variables. Stepwise multiple regression was used to examine whether triglyceride/HDL ratio is an independent predictor of HOMA-IR and LDL particle size when other potential confounding variables were also considered. The criteria of P = .05 for a variable to enter into the regression model and P = .10 for a variable to be removed were used in the multiple regression analysis. Further, logistic regression was used to examine predictors of pattern A versus pattern B LDL phenotypes. A receiver operating characteristic curve (ROC curve) and the area under the curve were also calculated to summarize the predictive power of the logistic model. For all statistical analyses, a P value of < .05 (2-tailed) was used to test for statistical significance.
Click figure to enlarge
RESULTS
Two hundred fifty-two patients enrolled in the study. Thirty-seven patients were missing fasting plasma glucose levels. Nine patients had fasting plasma glucose ≥ 126 mg/dL, suggesting undiagnosed diabetes,37 and were excluded from the analyses per a priori decision. Among 206 total patients included in the analysis, 74 patients were treated with olanzapine (daily dose range, 7.5-20 mg, with mean 14.6 mg); 78 patients with risperidone (daily dose range, 2-10 mg, with mean 4.5 mg); and 54 patients with typical antipsychotic agents, with haloperidol and fluphenazine most commonly used. Demographic, clinical measures and laboratory values of the study subjects are summarized in Table 1.
One major goal of this analysis was to examine the relative predictive power of different conventionally available lipid parameters for the development of insulin resistance. Figure 1 shows that all 4 lipid parameters were significantly correlated with HOMA-IR (P = .023 for LDL, P < .001 for HDL, triglyceride, and the triglyceride/HDL ratio). A stepwise multiple regression model was developed to identify relevant lipid predictors of HOMA-IR when other important potential confounding variables, including age, gender, race, family history of diabetes, body mass index (BMI), and antipsychotic agent, were also taken into consideration. Only BMI and the triglyceride/HDL ratio entered into the regression model (Table 2). No other lipid variables, including LDL, HDL, or triglyceride, entered into the regression model. Body mass index, which entered into the model as the first predictor, explained 24% variability of HOMA-IR (P < .001). The triglyceride/DL ratio predicted an additional 6% variability of HOMA-IR (P < .001). The overall regression model indicated that higher BMI and a higher triglyceride/HDL ratio predict increased HOMA-IR (P < .001).
Click figure to enlarge
Click figure to enlarge
Another goal was to examine whether or not conventionally available lipid parameters can predict LDL particle size or the existence of pattern B LDL phenotype. Figure 2 shows that 3 lipid parameters (HDL, triglyceride, the triglyceride/HDL ratio) were significantly correlated with LDL particle size (P < .001). However, there was not a significant correlation between conventional LDL measure and LDL particle size (P = .281). A stepwise multiple regression model was developed to identify relevant lipid predictors of LDL particle size. With other important potential confounding variables, including age, gender, race, family history of diabetes, BMI, and antipsychotic agent, also considered as candidate predictors, only the triglyceride/HDL ratio and gender entered into the regression model (Table 3). No other lipid variables, including LDL, HDL, or triglyceride, entered into the regression model. The triglyceride/HDL ratio entered into the model as the first predictor, explaining 57% variability of LDL particle size (P < .001). Gender predicted an additional 3% variability of LDL particle size (P = .001). The overall regression model indicated that a higher triglyceride/HDL ratio and male gender predict a smaller LDL particle size (P < .001).
Click figure to enlarge
Click figure to enlarge
Among 195 patients with LDL particle size measures, 64 (32.8%) had pattern B LDL phenotype. A logistic regression model was developed, with the triglyceride/HDL ratio and gender predicting the existence of pattern B LDL phenotype (Table 4). With a probability cutoff point of 0.5 (corresponding to a triglyceride/HDL ratio cutoff point 5.5 for male, 6.6 for female), the logistic regression model correctly classified 161 of 195 patients (82.6%), with a sensitivity of 75.9% (probability of predicting pattern B when the subject actually had pattern B) and a specificity of 85.4% (probability of predicting pattern A when the patient actually had pattern A). The model appeared to appropriately fit the data (Hosmer-Lemeshow test P = .98), indicating no reason to reject the model. The ROC curve for a range of probability cutoff points is presented in Figure 3. The total area under the ROC curve was estimated to be 0.90 (95% CI, 0.85-0.94). The area under the curve may be interpreted as the probability that the predictions and outcomes are concordant. For example, a value of 0.50 means that the predictions were no better than guessing, whereas a value of 1.0 would indicate perfect prediction.
Click figure to enlarge
Click figure to enlarge
DISCUSSION
Insulin resistance has been associated with the development of changes in metabolic parameters and cardiovascular disease.7,9 In past decades, various clinical characteristics, such as obesity, age, gender, race/ethnicity, a family history of diabetes, and lipid abnormalities, have been identified as predictors for insulin resistance.38-40 Among conventional lipid measures, plasma LDL, HDL, and triglyceride levels are associated with insulin resistance41 and cardiovascular disease.42,43 In recent years, the triglyceride/HDL ratio has emerged as a newer predictor of insulin resistance. A study of 258 nondiabetic, normotensive, overweight patients found that the triglyceride/HDL ratio was a useful surrogate to identify insulin resistance.27 Another study of 1,135 patients in good general health reported that the triglyceride/HDL ratio was the more effective predictor of insulin resistance as measured by the insulin suppression test compared to other conventional lipid measures (LDL, HDL, and triglyceride) (McLaughlin et al26). A study of 125 nondiabetic African Americans reported that the triglyceride/HDL ratio was not a reliable marker of insulin resistance as measured by frequently sampled intravenous glucose tolerance test.44 However, a study in 185 nondiabetic African Americans found a significant correlation between the triglyceride/HDL ratio and insulin sensitivity as measured by the clamp technique; the relationship was stronger in men than in women.28 Our results in nondiabetic patients with schizophrenia are consistent with most previous findings, indicating that the triglyceride/HDL ratio is a stronger predictor of insulin resistance as measured by HOMA-IR compared to other conventional lipid measures (LDL, HDL, triglyceride); and further, the predictive value of triglyceride/HDL ratio for insulin resistance is independent of age, gender, race, family history of diabetes, BMI, or antipsychotic agent.
A predominance of small, dense LDL subclass particles (pattern B) is more prevalent among men than women.45 The presence of smaller LDL particles is also associated with older age.46 In a study of 658 outpatients followed at cardiology clinics, men had significantly smaller LDL particle size than women; LDL particle size correlated with the triglyceride/HDL ratio better than with either triglyceride or HDL alone; further, no clinical variables other than the triglyceride/HDL ratio were significant in the logistic regression model analysis for the prediction of pattern B LDL phenotype; and a triglyceride/HDL ratio of ≥ 3.8 had 79% sensitivity and 81% specificity.32 In the McLaughlin et al study26 mentioned earlier, the triglyceride/HDL ratio was also the more effective predictor for LDL particle size compared to other conventional lipid measures. In a recent study33 of 146 relatively healthy Asian Indians, the triglyceride/HDL ratio correlated inversely with the LDL size; the triglyceride/HDL ratio was a better surrogate for LDL particle size than triglyceride or HDL levels; and a triglyceride/HDL ratio of ≥ 3.8 had 76% sensitivity and 93% specificity in predicting pattern B LDL phenotype. In another recent study of 150 patients without diabetes or cardiovascular disease, Tsimihodimos et al47 found that the triglyceride/HDL ratio is the more effective predictor among various clinical parameters for the presence of pattern B LDL phenotype. Consistent with previous findings, our results suggest that the triglyceride/HDL ratio is a better predictor of LDL particle size than LDL, HDL, or triglyceride in nondiabetic patients with schizophrenia; and a higher triglyceride/HDL ratio and male gender are associated with a smaller LDL particle size. In our study sample, a triglyceride/HDL ratio of 5.5 in men and 6.6 in women had 75.9% sensitivity and 85.4% specificity in predicting pattern B LDL phenotype. The sensitivity and specificity are similar to those reported in other studies. However, the optimal triglyceride/HDL ratios in predicting pattern B LDL phenotype seem to be higher in our study sample than those reported in nonpsychiatric populations.
To our knowledge, this is the first study examining the usefulness of triglyceride/HDL ratio in predicting insulin resistance and LDL particle size in the schizophrenia population. The triglyceride/HDL ratio seems to not only provide an estimate of insulin resistance but also identify patients who have an atherogenic lipoprotein profile. The triglyceride/HDL ratio, a simple, readily available, and inexpensive calculated measure, could help clinicians identify patients who may be at higher risk of metabolic disturbances and cardiovascular diseases and allow for early interventions.
The strengths of the present study include the relatively large sample size and the reliable fasting condition for blood sample collection. A major limitation is that HOMA-IR itself is a surrogate measure for insulin resistance, largely a function of fasting insulin values. While HOMA is a valid method to measure insulin resistance in clinical and epidemiologic studies, it is not as definitive as more labor-intensive measures of insulin resistance, such as the clamp technique or frequently sampled intravenous glucose tolerance test.48 The exclusion of patients with possibly undiagnosed diabetes and the lack of patients on other atypical antipsychotic agents (such as quetiapine, ziprasidone and aripiprazole) may limit the generalizability of the results to the entire schizophrenia population. Finally, any causal relationship cannot be drawn based on the cross-sectional study design. The clinical utility of our findings needs to be confirmed, and the optimal cutoff point of triglyceride/HDL ratio needs to be established in large prospective studies, especially in relation to individual antipsychotic agents.
Drug names: aripiprazole (Abilify), clozapine (Clozaril, FazaClo, and others), haloperidol (Haldol and others), olanzapine (Zyprexa), quetiapine (Seroquel), risperidone (Risperdal and others), ziprasidone (Geodon).
Author affiliation: Schizophrenia Program, Department of Psychiatry, Massachusetts General Hospital (Drs Fan, Sharma, and Henderson); Harvard Medical School (Drs Fan and Henderson), Boston; Partners Community Healthcare, Inc, Haverhill (Dr Liu), Massachusetts; and Eli Lilly and Company, Lilly Corporate Center, Indianapolis, Indiana (Drs Hoffman and Potts).
Author contribution: Dr Fan was responsible for the analysis and interpretation of the data for this article. All authors were responsible for the writing of the article.
Potential conflicts of interest: Drs Hoffmann and Potts are employees of and stock shareholders in Eli Lilly. Drs Fan, Liu, Sharma, and Henderson report no financial affiliation or other relationship relevant to the subject of this article.
Funding/support: The original study was sponsored by Eli Lilly.
Role of sponsor: Eli Lilly was responsible for the collection of the data used in this analysis.
REFERENCES
1. Brown S, Inskip H, Barraclough B. Causes of the excess mortality of schizophrenia. Br J Psychiatry. 2000;177(3):212-217. PubMed doi:10.1192/bjp.177.3.212
2. Osby U, Correia N, Brandt L, et al. Mortality and causes of death in schizophrenia in Stockholm county, Sweden. Schizophr Res. 2000;45(1-2):21-28. PubMed doi:10.1016/S0920-9964(99)00191-7
3. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA. 2001;285(19):2486-2497. PubMed doi:10.1001/jama.285.19.2486
4. McEvoy JP, Meyer JM, Goff DC, et al. Prevalence of the metabolic syndrome in patients with schizophrenia: baseline results from the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) schizophrenia trial and comparison with national estimates from NHANES III. Schizophr Res. 2005;80(1):19-32. PubMed doi:10.1016/j.schres.2005.07.014
5. Ford ES, Giles WH, Dietz WH. Prevalence of the metabolic syndrome among US adults: findings from the third National Health and Nutrition Examination Survey. JAMA. 2002;287(3):356-359. PubMed doi:10.1001/jama.287.3.356
6. Bonora E, Kiechl S, Willeit J, et al. Prevalence of insulin resistance in metabolic disorders: the Bruneck Study. Diabetes. 1998;47(10):1643-1649. PubMed doi:10.2337/diabetes.47.10.1643
7. Kitabchi AE, Temprosa M, Knowler WC, et al; Diabetes Prevention Program Research Group. Role of insulin secretion and sensitivity in the evolution of type 2 diabetes in the diabetes prevention program: effects of lifestyle intervention and metformin. Diabetes. 2005;54(8):2404-2414. PubMed doi:10.2337/diabetes.54.8.2404
8. Lillioja S, Mott DM, Spraul M, et al. Insulin resistance and insulin secretory dysfunction as precursors of non-insulin-dependent diabetes mellitus: prospective studies of Pima Indians. N Engl J Med. 1993;329(27):1988-1992. PubMed doi:10.1056/NEJM199312303292703
9. Semenkovich CF. Insulin resistance and atherosclerosis. J Clin Invest. 2006;116(7):1813-1822. PubMed doi:10.1172/JCI29024
10. Weyer C, Tataranni PA, Bogardus C, et al. Insulin resistance and insulin secretory dysfunction are independent predictors of worsening of glucose tolerance during each stage of type 2 diabetes development. Diabetes Care. 2001;24(1):89-94. PubMed doi:10.2337/diacare.24.1.89
11. Newcomer JW, Haupt DW, Fucetola R, et al. Abnormalities in glucose regulation during antipsychotic treatment of schizophrenia. Arch Gen Psychiatry. 2002;59(4):337-345. PubMed doi:10.1001/archpsyc.59.4.337
12. Henderson DC, Cagliero E, Copeland PM, et al. Glucose metabolism in patients with schizophrenia treated with atypical antipsychotic agents: a frequently sampled intravenous glucose tolerance test and minimal model analysis. Arch Gen Psychiatry. 2005;62(1):19-28. PubMed doi:10.1001/archpsyc.62.1.19
13. Grundy SM, Cleeman JI, Merz CN, et al; American Heart Association. Implications of recent clinical trials for the National Cholesterol Education Program Adult Treatment Panel III guidelines. Circulation. 2004;110(2):227-239. PubMed doi:10.1161/01.CIR.0000133317.49796.0E
14. Rader DJ. Effect of insulin resistance, dyslipidemia, and intra-abdominal adiposity on the development of cardiovascular disease and diabetes mellitus. Am J Med. 2007;120(suppl 1):S12-S18. PubMed doi:10.1016/j.amjmed.2007.01.003
15. McLaughlin T, Allison G, Abbasi F, et al. Prevalence of insulin resistance and associated cardiovascular disease risk factors among normal weight, overweight, and obese individuals. Metabolism. 2004;53(4):495-499. PubMed doi:10.1016/j.metabol.2003.10.032
16. Henderson DC, Cagliero E, Gray C, et al. Clozapine, diabetes mellitus, weight gain, and lipid abnormalities: a five-year naturalistic study. Am J Psychiatry. 2000;157(6):975-981. PubMed doi:10.1176/appi.ajp.157.6.975
17. Lindenmayer JP, Czobor P, Volavka J, et al. Changes in glucose and cholesterol levels in patients with schizophrenia treated with typical or atypical antipsychotics. Am J Psychiatry. 2003;160(2):290-296. PubMed doi:10.1176/appi.ajp.160.2.290
18. Wu RR, Zhao JP, Liu ZN, et al. Effects of typical and atypical antipsychotics on glucose-insulin homeostasis and lipid metabolism in first-episode schizophrenia. Psychopharmacology (Berl). 2006;186(4):572-578. PubMed doi:10.1007/s00213-006-0384-5
19. Olfson M, Marcus SC, Corey-Lisle P, et al. Hyperlipidemia following treatment with antipsychotic medications. Am J Psychiatry. 2006;163(10):1821-1825. PubMed doi:10.1176/appi.ajp.163.10.1821
20. Lieberman JA, Stroup TS, McEvoy JP, et al; Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) Investigators. Effectiveness of antipsychotic drugs in patients with chronic schizophrenia. N Engl J Med. 2005;353(12):1209-1223. PubMed doi:10.1056/NEJMoa051688
21. National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation. 2002;106(25):3143-3421. PubMed
22. Gardner CD, Fortmann SP, Krauss RM. Association of small low-density lipoprotein particles with the incidence of coronary artery disease in men and women. JAMA. 1996;276(11):875-881. PubMed doi:10.1001/jama.276.11.875
23. Coresh J, Kwiterovich PO Jr. Small, dense low-density lipoprotein particles and coronary heart disease risk: a clear association with uncertain implications. JAMA. 1996;276(11):914-915. PubMed doi:10.1001/jama.276.11.914
24. Lamarche B, Tchernof A, Moorjani S, et al. Small, dense low-density lipoprotein particles as a predictor of the risk of ischemic heart disease in men: prospective results from the Québec Cardiovascular Study. Circulation. 1997;95(1):69-75. PubMed
25. Hardy TA, Marquez E, Kryzhanovskaya L, et al. Cross-sectional comparison of fasting lipids in normoglycemic patients with schizophrenia during chronic treatment with olanzapine, risperidone, or typical antipsychotics. J Clin Psychopharmacol. 2006;26(4):405-408. PubMed doi:10.1097/01.jcp.0000227705.56870.06
26. McLaughlin T, Reaven G, Abbasi F, et al. Is there a simple way to identify insulin-resistant individuals at increased risk of cardiovascular disease? Am J Cardiol. 2005;96(3):399-404. PubMed doi:10.1016/j.amjcard.2005.03.085
27. McLaughlin T, Abbasi F, Cheal K, et al. Use of metabolic markers to identify overweight individuals who are insulin resistant. Ann Intern Med. 2003;139(10):802-809. PubMed
28. Stein E, Kushner H, Gidding S, et al. Plasma lipid concentrations in nondiabetic African American adults: associations with insulin resistance and the metabolic syndrome. Metabolism. 2007;56(7):954-960. PubMed doi:10.1016/j.metabol.2007.02.008
29. Jeppesen J, Hein HO, Suadicani P, et al. High triglycerides/low high-density lipoprotein cholesterol, ischemic electrocardiogram changes, and risk of ischemic heart disease. Am Heart J. 2003;145(1):103-108. PubMed doi:10.1067/mhj.2003.45
30. Gaziano JM, Hennekens CH, O’ Donnell CJ, et al. Fasting triglycerides, high-density lipoprotein, and risk of myocardial infarction. Circulation. 1997;96(8):2520-2525. PubMed
31. Maruyama C, Imamura K, Teramoto T. Assessment of LDL particle size by triglyceride/HDL-cholesterol ratio in non-diabetic, healthy subjects without prominent hyperlipidemia. J Atheroscler Thromb. 2003;10(3):186-191. PubMed
32. Hanak V, Munoz J, Teague J, et al. Accuracy of the triglyceride to high-density lipoprotein cholesterol ratio for prediction of the low-density lipoprotein phenotype B. Am J Cardiol. 2004;94(2):219-222. PubMed doi:10.1016/j.amjcard.2004.03.069
33. Bhalodkar NC, Blum S, Enas EA. Accuracy of the ratio of triglycerides to high-density lipoprotein cholesterol for predicting low-density lipoprotein cholesterol particle sizes, phenotype B, and particle concentrations among Asian Indians. Am J Cardiol. 2006;97(7):1007-1009. PubMed doi:10.1016/j.amjcard.2005.10.036
34. Hermans MP, Levy JC, Morris RJ, et al. Comparison of insulin sensitivity tests across a range of glucose tolerance from normal to diabetes. Diabetologia. 1999;42(6):678-687. PubMed doi:10.1007/s001250051215
35. Otvos JD. Measurement of lipoprotein subclass profiles by nuclear magnetic resonance spectroscopy. Clin Lab. 2002;48(3-4):171-180. PubMed
36. Otvos JD, Jeyarajah EJ, Bennett DW, et al. Development of a proton nuclear magnetic resonance spectroscopic method for determining plasma lipoprotein concentrations and subspecies distributions from a single, rapid measurement. Clin Chem. 1992;38(9):1632-1638. PubMed
37. Genuth S, Alberti KG, Bennett P, et al; Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Follow-up report on the diagnosis of diabetes mellitus. Diabetes Care. 2003;26(11):3160-3167. PubMed doi:10.2337/diacare.26.11.3160
38. Lee JM. Insulin resistance in children and adolescents. Rev Endocr Metab Disord. 2006;7(3):141-147. PubMed doi:10.1007/s11154-006-9019-8
39. Klein DJ, Aronson Friedman L, Harlan WR, et al. Obesity and the development of insulin resistance and impaired fasting glucose in black and white adolescent girls: a longitudinal study. Diabetes Care. 2004;27(2):378-383. PubMed doi:10.2337/diacare.27.2.378
40. Smith DA. Treatment of the dyslipidemia of insulin resistance. Med Clin North Am. 2007;91(6):1185-1210, x. PubMed doi:10.1016/j.mcna.2007.06.008
41. Laws A, Reaven GM. Evidence for an independent relationship between insulin resistance and fasting plasma HDL-cholesterol, triglyceride and insulin concentrations. J Intern Med. 1992;231(1):25-30. PubMed doi:10.1111/j.1365-2796.1992.tb00494.x
42. Miller GJ, Miller NE. Plasma-high-density-lipoprotein concentration and development of ischaemic heart-disease. Lancet. 1975;1(7897):16-19. PubMed doi:10.1016/S0140-6736(75)92376-4
43. Austin MA, Hokanson JE, Edwards KL. Hypertriglyceridemia as a cardiovascular risk factor. Am J Cardiol. 1998;81(4A):7B-12B. PubMed doi:10.1016/S0002-9149(98)00031-9
44. Sumner AE, Finley KB, Genovese DJ, et al. Fasting triglyceride and the triglyceride-HDL cholesterol ratio are not markers of insulin resistance in African Americans. Arch Intern Med. 2005;165(12):1395-1400. PubMed doi:10.1001/archinte.165.12.1395
45. Campos H, Blijlevens E, McNamara JR, et al. LDL particle size distribution: results from the Framingham Offspring Study. Arterioscler Thromb. 1992;12(12):1410-1419. PubMed
46. McNamara JR, Campos H, Ordovas JM, et al. Effect of gender, age, and lipid status on low density lipoprotein subfraction distribution: results from the Framingham Offspring Study. Arteriosclerosis. 1987;7(5):483-490. PubMed
47. Tsimihodimos V, Gazi I, Kostara C, et al. Plasma lipoproteins and triacylglycerol are predictors of small, dense LDL particles. Lipids. 2007;42(5):403-409. PubMed doi:10.1007/s11745-007-3050-8
48. Wallace TM, Levy JC, Matthews DR. Use and abuse of HOMA modeling. Diabetes Care. 2004;27(6):1487-1495. PubMed doi:10.2337/diacare.27.6.1487
