- Email: [email protected]
Genetic factors play an important role in the pathogenesis of hyperlipidemia post-transplantation
Transplantation
Genetic Factors Play an Important Role in the Pathogenesis of Hyperlipidemia Post-Transplantation Carlos A. Aguilar-Salinas, MD, Araceli Dı´az-Polanco, MD, Eduardo Quintana, MD, Nayeli Macias, RD, Adriana Arellano, RD, Erika Ramı´rez, MD, Marı´a Luisa Ordo´n˜ez, PhD, Consuelo Vela´squez-Alva, PhD, Francisco J. Go´mez Pe´rez, MD, Josefina Alberu´, MD, and Ricardo Correa-Rotter, MD ● Background: Our purpose was to identify factors associated with hyperlipidemia post-transplantation in a Hispanic population. Methods: From 1985 to 1999, a kidney graft survival longer than 3 months occurred in 293 cases at the Instituto Nacional de la Nutricio´n. Most of the patients living in Mexico City were included (n ⴝ 83). The evaluation included a questionnaire, blood samples, and assessment of body composition and dietary habits. As many as possible first-degree relatives were studied. Results: Women had higher values of cholesterol (236 ⴞ 51 versus 215 ⴞ41; P < 0.05), low-density lipoprotein cholesterol (147 ⴞ 42 versus 131 ⴞ 34; P ⴝ 0.05), high-density lipoprotein cholesterol (57.3 ⴞ 14 versus 47.9 ⴞ 14; P ⴝ 0.002) and high-density lipoprotein-2 cholesterol. Isolated hypercholesterolemia was the most common lipid abnormality (40.9%), followed by mixed hyperlipidemia. Lipoprotein (a) greater than 30 mg/dL was found in 13 cases. Familial combined hyperlipidemia (FCHL) in the patient’s relatives was a marker for dyslipidemia (odds ratio, 7.04; 95% confidence interval, 1.2 to 59.7). These cases had a worse lipid profile. Cyclosporine-treated FCHL patients had higher lipid levels compared with the non-FCHL, cyclosporine-treated patients. The effects of cyclosporine on the lipid levels were lower, but significant, after the exclusion of the FCHL cases. Conclusion: Post-transplant dyslipidemia is determined by genetic and environmental factors. FCHL in the patient’s relatives was associated with post-transplant hyperlipidemia; an additive effect with cyclosporine was found. The evaluation of the lipid profile of relatives may be useful for the assessment of the risk of post-transplant dyslipidemia. © 2002 by the National Kidney Foundation, Inc. INDEX WORDS: Cholesterol; renal transplantation; familial combined hyperlipidemia (FCHL); cyclosporine; Mexico.
T
HE LONG-TERM MORBIDITY rates and mortality associated with renal transplantation are still high, and the most prevalent cause is cardiovascular disease.1 The incidence of new ischemic cardiac events in a 4-year period was almost 11% in patients without prior coronary heart disease and 15% in secondary prevention cases.2 These rates are almost five times higher than predicted from the tables derived from the Framingham Heart Study.3 Dyslipidemia is one of the main contributors to the progression of vascular lesions that ultimately result in a coronary event. The correction of dyslipidemia is an effective alternative for the prevention of cardiovascular events in nontransplant populations.4 The lipid abnormalities encompassed by the term dyslipidemia include hypercholesterolemia, hypertriglyceridemia, low high-density lipoprotein cholesterol (HDL-C), and combinations of these defects. In the combined results from five studies done in white subjects, 63% of 549 patients had cholesterol levels greater than 240 mg/dL. Low-density lipoprotein cholesterol (LDL-C) greater than 130 mg/dL was found in
almost 60% of cases.5 Low HDL cholesterol levels are not as common as the previous two abnormalities (approximately 12%). The prevalence of hyperlipidemia varies, and the type of immunosuppressive medications, nutritional status, dietary habits, and genetic predisposition are among the main determinants for the development of these lipid disturbances.6 Some determinants of dyslipidemia in kidney transplant patients have not been studied extensively. The data reported in white groups may not be reproducible in populations with a different genetic From the Departments of Endocrinology and Metabolism, Nephrology and Mineral Metabolism, and Transplants, Instituto Nacional de Ciencias Me´dicas y Nutricio´n Salvador Zubira´n; and Department of Healthcare, Universidad Auto´noma Metropolitana, Unidad Xochimilco, Mexico City, Mexico. Received October 26, 2001; accepted in revised form March 4, 2002. Address reprint requests to Carlos A. Aguilar-Salinas, MD, Vasco de Quiroga 15, Mexico City 14000 Mexico. E-mail: [email protected] © 2002 by the National Kidney Foundation, Inc. 0272-6386/02/4001-0021$35.00/0 doi:10.1053/ajkd.2002.33926
American Journal of Kidney Diseases, Vol 40, No 1 (July), 2002: pp 169-177
169
170
AGUILAR-SALINAS ET AL
background. The prevalence of hypertriglyceridemia and hypoalphalipoproteinemia is low in white subjects, but the genetic predisposition for having these abnormalities is strong in Hispanic and Asian groups.7 Mean cholesterol levels seem to be significantly lower in post-transplant Asian patients compared with levels found in white patients.8 The influence of pretransplant lipoprotein abnormalities, including the coexistence of a genetic form of dyslipidemia, has been partially studied.9 Finally, the clinical correlations of several lipoprotein profiles (ie, mixed hyperlipidemia or isolated hypercholesterolemia) have not been described. The purpose of this study was to describe the prevalence of lipid abnormalities and their determinants in a Mexican kidney transplant population. The evaluation included a family study for primary dyslipidemias, assessment of body composition, dietary habits, and apoE genotype. MATERIAL AND METHODS
Patients Informed consent was obtained from all participants of the study. From 1985 to 1999, 315 kidney transplants were done successfully at the Instituto Nacional de Ciencias Me´dicas y Nutricio´n Salvador Zubira´n. Among them, graft survival longer than 3 months occurred in 293 cases. Many of these patients lived outside Mexico City (70% of the population). Most of the patients living in Mexico City were reached and invited to participate. Of the total population, 28% (n ⫽ 83) was included. Each patient was evaluated by an endocrinologist who completed a questionnaire and a physical examination. A registered dietitian collected a 3-day prospective dietary registry using a previously validated format. The data were analyzed using the SCVAN program (Instituto Nacional de la Nutricio´n, Mexico City, Mexico), useful for the analysis of Mexican cooking. All patients were requested to ask as many as possible first-degree relatives to come to the clinic to have a complete lipid profile and glucose concentration done. It was possible to study at least two relatives of 51 probands (61% of the population). Blood pressure was measured to the nearest even digit with a sphygmomanometer with the subject in the supine position after a 5-minute rest. Participants removed their shoes and upper garments. Height was measured to the nearest 0.5 cm. Body weight was measured on a dailycalibrated balance and recorded to the nearest 0.1 kg. Body mass index was calculated and used as an index of overall adiposity. Body composition of the probands was measured using electric bioimpedance technology.
Methods All analytical measurements were done at the Department of Endocrinology and Lipid Metabolism of the Instituto
Nacional de Ciencias Me´dicas y Nutricio´n Salvador Zubira´n. Blood samples were obtained from the subjects after having fasted for 9 to 12 hours; they remained sitting for 5 minutes before the blood was drawn. Glucose was analyzed by the glucose-oxidase method (Boehringer, Mannheim, Germany). Serum concentrations of total cholesterol and triglycerides were determined by enzymatic methods (Boehringer, Mannheim, Germany). HDL-C was measured after precipitation of very low-density lipoproteins (VLDL) and LDL by the phosphotungstate method (Boehringer, Mannheim, Germany). LDL-C was measured by ultracentrifugation using a -quantification method.10 Briefly, 3 mL of plasma was ultracentrifuged at d 1.006 kg/L for 18 hours at 105,000 ⫻ g. The VLDL fraction, at the top, was removed; the cholesterol and HDL-C concentrations were measured in the bottom fraction. LDL-C was calculated as the difference between that in the bottom fraction and HDL-C. Intra-assay variation coefficient values for total cholesterol, triglycerides, and HDL-C were 3%, 5%, and 5%. Insulin was analyzed by enzyme-linked immunosorbent assay in the ES-33 system (Boehringer, Mannheim, Germany). The crossreactivity with proinsulin for this assay was 40%. Our laboratory followed standardization procedures according to the World Health Organization recommendations, including the use of external control sera. Apolipoproteins A-1 (apoAI) and B (apoB) and lipoprotein (a) (Lp(a)) were measured with an immunoturbidimetric method11 on a Hitachi 705 analyzer (Hitachi Science Systems LTD, Hitachinaka, Japan); the assays were done according to the manufacturer’s instructions with manufacturer-provided calibrators and antisera. All samples were kept frozen at ⫺80°C until they were analyzed; the maximum time of storage was 1 month. Cholesterol concentration of the HDL subfractions was measured using a double precipitation assay.12 Apolipoprotein E (apoE) genotype was determined by gene amplification and cleavage with HhaI as described by Hixson and Vernier.13
Definitions Overweight was defined as a body mass index between 25 and 30 kg/m2. Obesity was defined as body mass index greater than or equal to 30 kg/m2. Individuals were diagnosed as diabetics if they had a previous diagnosis of diabetes or had a fasting blood glucose value greater than or equal to 126 mg/dL (7 mmol/L) and no previous history of diabetes. Hypertension was diagnosed when the systolic blood pressure was greater than or equal to 140 mm Hg or diastolic blood pressure was greater than or equal to 90 mm Hg or when antihypertensive medications were being used. Ischemic heart disease was considered if there was a history of myocardial infarction. Smoking was defined as any tobacco consumption during the previous month. Cholesterol value greater than or equal to 200 mg/dL (5.2 mmol/L) was considered to be hypercholesterolemia. HDL-C concentration was considered to be abnormal if it was less than 35 mg/dL (0.9 mmol/L). Mixed hyperlipidemia was defined as cholesterol greater than or equal to 200 mg/dL (5.2 mmol/L) plus triglycerides greater than or equal to 200 mg/dL (2.2 mmol/L). Familial combined hyperlipidemia (FCHL) was considered to be present if besides the proband, at least two family members had two different lipid profiles (ie, isolated
DYSLIPIDEMIA POST-TRANSPLANTATION IN MEXICO
171
hypercholesterolemia, isolated hypertriglyceridemia, or mixed hyperlipidemia).14 Insulin sensitivity was estimated using the Homeostasis Model Assessment (HOMA) score based on the formula: fasting serum insulin (U/mL) ⫻ fasting plasma glucose (mmol/L)/22.5, as described by Matthews et al.15 With this method, high HOMA scores denote low insulin sensitivity.
received prednisone and azathioprine. Cyclosporine was used in 51 cases. One subject had proteinuria of 3 g/d or more. Diabetes was found in four subjects; only one had hyperglycemia before the kidney transplant.
Statistical Analysis
Lipid Profiles, Prevalence, and Associated Clinical Characteristics For both genders, cholesterol, LDL-C, and non-HDL-C were consistently higher in patients compared with their relatives (Table 2). Women had significantly higher values of cholesterol (236 ⫾ 51 versus 215 ⫾ 41; P ⬍ 0.05), LDL-C (147 ⫾ 42 versus 131 ⫾ 34; P ⫽ 0.05), HDL-C (57.3 ⫾ 14 versus 47.9 ⫾ 14; P ⫽ 0.002), and HDL-2 cholesterol (19.8 ⫾ 12.6 versus 11.7 ⫾ 7.4; P ⬍ 0.001). These concentrations were above the 90th, 80th, and 90th percentiles of the cholesterol, LDL-C, and HDL-C values reported in Mexican urban women.16 For men, the corresponding percentiles were 80th, 70th, and 80th. No differences were found between genders in triglycerides and apoB concentrations; the mean values were above the 60th and 80th percentiles. The prevalence of dyslipidemia was high in this population. Triglyceride values greater than 200 mg/dL (2.2 mmol/L) were found in 26% of cases (an additional 19.7% had levels between 150 and 199 mg/dL). Cholesterol values greater than 240 mg/dL (6.3 mmol/L) were found in 34% (an additional 29% had levels between 200 and 239 mg/dL). Finally, 25% had LDL-C levels greater than 160 mg/dL (an additional 26% had levels between 130 and 159 mg/dL), and 12% had HDL-C less than 35 mg/dL (0.9 mmol/L). Of patients, 26% had HDL-C greater than 60 mg/dL (1.6 mmol/L). Lp(a) concentrations greater than 30 mg/dL were found in 13 cases. Isolated hypercholesterolemia was the abnormal lipid profile most frequently found in this population (34 subjects [40.9%]); this abnormality was more prevalent in women (53%) than in men (29.5%). Mixed hyperlipidemia also was common (19 cases [22.8%]); this abnormality was more prevalent in men (27% versus 17%; P ⬍ 0.05). Low HDL cholesterol (⬍35 mg/dL) was found in 10 cases (12%); 8 were men (18.1%). A normal lipid profile (including HDL-C) was found in 21 cases.
The database was validated through recognition of missing values, outliers, and inconsistencies between variables. Descriptive analysis included the estimation of mean values and SDs for continuous variables. Prevalence and frequencies were expressed as percentages. Significance of the differences between the subgroups of the population were tested by one-way analysis of variance using Scheffe´ multiple comparison method. Categorical variables were compared by chi-square with Yates correction or the Fisher exact test when appropriate. Correlation coefficients were calculated for measuring the association between continuous variables. Analysis of covariance was used for gender adjustment of the variables. A stepwise multiple regression analysis was done with the lipid concentrations as dependent variables. All statistical analyses were done using the Statgraphics statistical package (STSC, Rockville, MD).
RESULTS
The study subjects were young adults who had undergone a kidney transplant 5.6 years ago (Table 1) and had a serum creatinine of 1.46 ⫾ 0.56 mg/dL (129 ⫾ 49.5 mmol/L). All patients Table 1.
Characteristics of the Patients
N Age (y) Gender (% males) Time after renal grafting (y) Serum creatinine (mg/dL) Systolic blood pressure Diastolic blood pressure Albuminuria ⬎ 1 g/L [n (%)] Hypertension [n (%)] Diabetes mellitus [n (%)] Cardiovascular acting drugs [n (%)] Diuretics Beta blocker Calcium channel blocker ACE inhibitor Statins Fibrates Immunosuppression [n (%)] Prednisone plus azathioprine Prednisone plus cyclosporine plus azathioprine Prednisone dose (mg/d) Cyclosporine dose (mg/kg/d)
83 37.2 ⫾ 12.6 53% 5.6 ⫾ 2.2 1.46 ⫾ 0.56 123.5 ⫾ 13.3 81.1 ⫾ 10.9 6 (7.2%) 61 (73.4) 4 (4.8) 4 21 43 9
(4.8) (25.3) (51.8) (10.8) 11% 4%
32 (38%) 51 (61.4%) 17.4 ⫾ 7.3 2.9 ⫾ 2.6
Abbreviation: ACE, angiotensin-converting enzyme.
172 Table 2.
AGUILAR-SALINAS ET AL Metabolic Profile of the Study Subjects and Their First-Degree Relatives Including Cases With Familial Combined Hyperlipidemia Cases
n Age Body mass index (kg/m2) Glucose (mg/dL) Cholesterol (mg/dL) Triglycerides (mg/dL) HDL cholesterol (mg/dL) LDL cholesterol (mg/dL) Non-HDL cholesterol (mg/dL) HDL-2 cholesterol Lipoprotein (a) (mg/dL) Insulin (U/mL) Mixed hyperlipidemia (n [%]) HDL cholesterol ⬍ 35 mg/dL
Relatives
Total Cases
FCHL Cases
Non-FCHL Cases
Without Family Study
FCHL Kindreds
Non-FCHL Kindreds
P Value
83 37.2 ⫾ 12.6 23.3 ⫾ 3.3 74 ⫾ 19 225 ⫾ 47.4*† 169 ⫾ 91.3*† 52 ⫾ 14.7*† 139 ⫾ 38.8† 173 ⫾ 45.8† 15.5 ⫾ 11 17.5 ⫾ 28.4 12.9 ⫾ 6 19 (22.8)† 10 (12)
14 37.9 ⫾ 13.6 24.5 ⫾ 2.1 70 ⫾ 11 266 ⫾ 51*†‡ 240 ⫾ 121*†‡ 49 ⫾ 14*† 170 ⫾ 34*†‡ 217 ⫾ 46*†‡ 14.7 ⫾ 9.3 20 ⫾ 27 10.5 ⫾ 4 8 (57)*†‡ 3 (21)
37 36.1 ⫾ 9.7 23 ⫾ 3 72 ⫾ 19 211 ⫾ 40† 158 ⫾ 87† 51 ⫾ 13† 128 ⫾ 34† 160 ⫾ 40† 15.4 ⫾ 11 16.8 ⫾ 28 13.5 ⫾ 6 4 (11) 5 (13.5)
32 37.2 ⫾ 10 23.2 ⫾ 4 74 ⫾ 20 223 ⫾ 44 149 ⫾ 64 54 ⫾ 15 138 ⫾ 31 168 ⫾ 40 15.8 ⫾ 11 17.3 ⫾ 28 13.1 ⫾ 6 7 (21) 2 (6)
47 — — — 217 ⫾ 37† 203 ⫾ 127† 38.2 ⫾ 10.7† 138 ⫾ 38† 178 ⫾ 37† — — — 15 (32)† 2 (4)
75 — — — 185 ⫾ 32 133 ⫾ 78 44.7 ⫾ 11 114 ⫾ 28 140 ⫾ 33 — — — 8 (10.6) 7 (9.3)
NS NS NS ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 NS NS NS ⬍0.001 NS
NOTE. Data expressed as mean ⫾ SD. FCHL was considered to be present if besides the proband, at least two family members had two different lipid profiles (ie, isolated hypercholesterolemia, isolated hypertriglyceridemia, or mixed hyperlipidemia). Mixed hyperlipidemia was defined as cholesterol ⱖ200 mg/dL (5.2 mmol/L) plus triglycerides ⱖ 200 mg/dL (2.2 mmol/L). Conversion factors to obtain SI units: glucose ⫽ 0.05551; cholesterol, HDL cholesterol, LDL cholesterol, non-HDL cholesterol and HDL-2 cholesterol ⫽ 0.02586; triglycerides ⫽ 0.0112. Abbreviation: NS, not significant. *P ⬍ 0.05 versus relatives from FCHL kindreds. †P ⬍ 0.05 versus relatives from non-FCHL kindreds. ‡P ⬍ 0.05 versus non-FCHL cases.
Diet and Body Composition The body mass index was similar in both genders (23.1 ⫾ 3 kg/m2 versus 23.4 ⫾ 3.6 kg/m2 for men and women). A greater percentage of women were overweight or obese (29% versus 13%; P ⬍ 0.05). Waist circumference was slightly greater in men (87.6 ⫾ 10 cm versus 84.5 ⫾ 16 cm; P ⫽ 0.2). As a mean, patients received a carbohydrate-rich, low-fat, low-cholesterol diet. Men consumed more calories than women (2,406 ⫾ 1,156 cal/d versus 2,032 ⫾ 908 cal/d; P ⫽ not significant). For men, a greater percentage of the calories came from fat (29.8% versus 24.9%; P ⬍ 0.05). This difference was explained mainly by a greater consumption of saturated (7.6% versus 6.4%; P ⫽ 0.05) and monounsaturated fat (8.4% versus 6.3%; P ⫽ 0.01). Also, the cholesterol content of the diet was higher in men (282 ⫾ 188 mg/d versus 181 ⫾ 111 mg/d; P ⫽ 0.01). No correlation was found between the components of the diet and the plasma lipid levels. The body composition was assessed in every
case. The amount of fat was significantly greater in women (17 ⫾ 1.3 kg versus 7.1 ⫾ 1.2 kg; P ⬍ 0.001). In contrast, the amount of lean mass was greater in men (57.4 ⫾ 1.1 kg versus 41.8 ⫾ 1.2 kg; P ⬍ 0.001). No correlation was found between the body composition and either cholesterol or apoB levels. In men, a strong inverse relationship was found, however, between fat mass and HDL-C (r ⫽ ⫺0.33; P ⫽ 0.05); no association was observed with the HDL-2 cholesterol. In men, an association was found between insulin (r ⫽ 0.51; P ⫽ 0.008) and HOMA (r ⫽ 0.39; P ⬍ 0.05) scores with fat mass. Genetic Factors FCHL was a strong marker for having lipid disturbances (odds ratio [OR], 7.04; 95% confidence interval [CI], 1.2 tp 59.7). The characteristics of the 14 FCHL probands were compared with the 37 patients who had relatives with nondiagnostic lipid changes (Table 2). The abnormalities of the lipid profile were more severe and common in the FCHL cases. These subjects had
DYSLIPIDEMIA POST-TRANSPLANTATION IN MEXICO Table 3.
173
Effects of Cyclosporine Treatment in Cases With or Without Familial Combined Hyperlipidemia Cyclosporine All
n 51 Age (y) 36.9 ⫾ 13.2 23.4 ⫾ 2.8 Body mass index (kg/m2) Gender (M/F) 29/22 Creatinine (mg/dL) 1.52 ⫾ 0.4 Cholesterol (mg/dL) 235 ⫾ 48* Triglycerides (mg/dL) 182 ⫾ 86* Non HDL-C (mg/dL) 186 ⫾ 44* LDL-C (mg/dL) 150 ⫾ 38* Apolipoprotein B (mg/dL) 109.2 ⫾ 27* HDL-C (mg/dL) 49 ⫾ 12* HDL-2 cholesterol (mg/dL) 14.3 ⫾ 9.7 BUN (mg/dL) 24.8 ⫾ 9* Uric acid (mg/dL) 7.6 ⫾ 1.4*
No Cyclosporine
FCHL⫹
FCHL⫺
FCHL Unknown
FCHL⫺
FCHL Unknown
P Value
13 38.3 ⫾ 11 24.8 ⫾ 2.1† 8/5 1.61 ⫾ 0.5 264 ⫾ 52† 224 ⫾ 110† 216 ⫾ 47† 171 ⫾ 34† 130 ⫾ 30† 48.3 ⫾ 14 14.5 ⫾ 12 28.4 ⫾ 9 7.7 ⫾ 1.6
23 36 ⫾ 14 22.9 ⫾ 2.9 13/10 1.47 ⫾ 0.9 216 ⫾ 42‡ 174 ⫾ 72‡ 168 ⫾ 39‡ 133 ⫾ 37‡ 97.7 ⫾ 25 47.6 ⫾ 11‡ 12.3 ⫾ 7‡ 23.1 ⫾ 9 7.7 ⫾ 1.4‡
15 36.8 ⫾ 9 23.1 ⫾ 1.5 8/7 1.51 ⫾ 0.1 239 ⫾ 47§ 158 ⫾ 96 187 ⫾ 42§ 156 ⫾ 40§ 108 ⫾ 28§ 51.7 ⫾ 12§ 17.2 ⫾ 9 24.4 ⫾ 9 7.3 ⫾ 1.6§
14 35.2 ⫾ 11 22.6 ⫾ 4 7/7 1.47 ⫾ 0.6 203 ⫾ 42§ 133 ⫾ 96 146 ⫾ 39 119 ⫾ 33 92 ⫾ 22 57.5 ⫾ 16 18.1 ⫾ 12 21 ⫾ 10 5.9 ⫾ 1.2
17 40.8 ⫾ 10 23.3 ⫾ 4 7/10 1.26 ⫾ 0.7 209 ⫾ 9 141 ⫾ 80 152 ⫾ 36 123 ⫾ 30 89 ⫾ 20 57.7 ⫾ 11 16.6 ⫾ 11 18.5 ⫾ 11 5.5 ⫾ 1.1
NS NS NS NS ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.01 NS ⬍0.05 ⬍0.001
NOTE. Data expressed as mean ⫾ SD. In the cyclosporine ⫹/FCHL⫺ group were included cases in which the family study was not diagnostic of FCHL. The FCHL unknown are cases with no family study. Differences between groups remained statistically significant after adjusting for body mass index. Conversion factors to obtain SI units: creatinine ⫽ 88.4; cholesterol, HDL cholesterol; LDL cholesterol, non-HDL cholesterol and HDL-2 cholesterol ⫽ 0.02586; triglycerides ⫽ 0.0112; insulin ⫽ 6.0; BUN ⫽ 0.357. Abbreviations: BUN, blood urea nitrogen; NS, not significant. *P ⬍ 0.05, cyclosporine-treated versus non– cyclosporine-treated cases. †P ⬍ 0.05, cyclosporine ⫹/FCHL⫹ versus cyclosporine ⫹/FCHL⫺ cases. ‡P ⬍ 0.05, cyclosporine ⫹/FCHL⫺ versus noncyclosporine/non-FCHL cases. §P ⬍ 0.05, cyclosporine ⫹/FCHL unknown versus noncyclosporine/FCHL unknown cases.
the highest lipid levels in this study group; more than half of them had a mixed hyperlipidemia. Significant differences in the lipid profile were found between the probands and their relatives, in FCHL and non-FCHL kindreds. The lipid abnormalities found in the FCHL relatives were similar to abnormalities observed in the nonFCHL transplant patients. These comparisons show the effect of renal transplantation on the lipid profile, independent of genetic factors. The apoE genotype distribution was assessed. Of subjects, 68 were e3/e3 homozygous and 15 were e3/e4 heterozygous. In agreement with the low prevalence of the e2 allele in Mexican populations,17 none of the study subjects had this genetic marker. Lipid levels were not statistically different between the e4/e3 heterozygous and the e3/e3 subjects. Two characteristics differ between these two groups, however. The e4/e3 heterozygous subjects had higher HDL-2 cholesterol levels (20.5 ⫾ 14 mg/dL versus 14.3 ⫾ 9.9 mg/dL; P ⬍ 0.05). Also, they had higher insulin concentrations (12.3 ⫾ 4.5 U/mL versus 15.4 ⫾ 10.5 U/mL; P ⬍ 0.05) and HOMA scores
(2.14 ⫾ 0.1 versus 4.18 ⫾ 0.96; P ⫽ 0.01). These data suggest the presence of a higher degree of insulin resistance in the e4/e3 heterozygous subjects, as previously reported by some authors.18 Effects of Immunosuppressive Therapy and Other Cardiovascular Drugs Prednisone dose was significantly related to HDL-C (r ⫽ 0.29; P ⬍ 0.05), HDL-2 cholesterol (r ⫽ 0.281; P ⬍ 0.05), glucose (r ⫽ 0.32; P ⬍ 0.05), and potassium levels (r ⫽ ⫺0.37; P ⬍ 0.05). Significant differences were found between patients treated with and without cyclosporine (Table 3). As expected, cyclosporine-treated patients had higher cholesterol, LDL-C, nonHDL cholesterol, apoB, and triglycerides concentrations. They also had lower HDL-C; this abnormality was explained by lower concentrations of HDL-3 cholesterol. Higher uric acid and blood urea nitrogen concentrations also were found in these cases. All but one subject with FCHL received cyclosporine. Because this association could play a role as a confounder for the analysis,
174
AGUILAR-SALINAS ET AL Table 4.
Univariate Correlations Between Blood Lipids and the Evaluated Parameters (n ⴝ 83)
Age (y) Male gender Body mass index (kg/m2) Diastolic blood pressure (mm Hg) Serum creatinine (mg/dL) Urine albumin (g/d) Pretransplant concentration Member of a FCHL kindred HOMA score Fat mass (kg) Calories (cal/d) Fat content of diet (%) Cyclosporine treatment Prednisone dosage (mg/d)
Cholesterol
Apolipoprotein B
Triglycerides
HDL-C
⫺0.059 ⫺0.215* 0.006 0.052 0.108* ⫺0.168 0.190* 0.499* ⫺0.005 0.077 0.059 ⫺0.067 0.315* 0.08
⫺0.024 ⫺0.02 0.053 0.112 0.091* 0.086 —† 0.261* ⫺0.049 0.010 0.102 ⫺0.018 0.342* 0.061
0.065 0.141 0.161 0.258* 0.057 0.258* 0.457* 0.354* ⫺0.35 0.026 0.038 ⫺0.009 0.297* 0.118
⫺0.133* ⫺0.311* ⫺0.168* ⫺0.253* ⫺0.069* ⫺0.183* 0.810* 0.073 ⫺0.016 ⫺0.048 ⫺0.112 ⫺0.149 ⫺0.24* ⫺0.293*
*P ⬍ 0.05. †No pretransplant apolipoprotein B values were available.
we analyzed separately the cyclosporine-treated patients with or without FCHL. Cases without a family study were also included as a separate group. The cyclosporine-treated/FCHL patients had remarkably higher cholesterol and apoB levels compared with the non-FCHL/cyclosporinetreated subjects. The effects of cyclosporine on the lipid profile are shown when the lipid concentrations are compared between the non-FCHL patients with or without cyclosporine treatment. The same trend was found between the patients in whom a family study was not available. The effects of other drugs on the lipid profile were analyzed. The use of angiotensin-converting enzyme inhibitors or calcium channel blockers was not related to changes in the evaluated parameters. -Blockers were associated with significantly higher triglycerides (207 ⫾ 18 mg/dL versus 148 ⫾ 14 mg/dL; P ⫽ 0.01), cholesterol (241 ⫾ 9.2 mg/dL versus 213 ⫾ 6.8 mg/dL; P ⫽ 0.01), LDL-C (152 ⫾ 7 mg/dL versus 131 ⫾ 5.7 mg/dL; P ⬍ 0.05), and apoB (115 ⫾ 5.5 mg/dL versus 95 ⫾ 4.8 mg/dL; P ⬍ 0.01). The number of patients who received diuretics (4.8%), statins (11%), or fibrates (4.8%) was too small to provide definitive conclusions. Associations Between the Evaluated Parameters and Blood Lipids As shown in Table 4, the presence of FCHL, cyclosporine treatment, and the pretransplant lipid levels were correlated with the concentration of
the apoB-containing lipoproteins. A stepwise regression model was constructed using cholesterol levels as a dependent variable (Table 5). The adjusted r2 of the model was 0.292 (P ⬍ 0.001). Gender, FCHL, serum creatinine, triglycerides, and the use of cyclosporine remained as independent predictors. We searched for factors associated with low HDL-C values. The pretransplant HDL-C values, cyclosporine treatment, and prednisone dosage were associated with lower levels of HDL-C in univariate analysis. A stepwise regression model was constructed using HDL-C levels as a Table 5. Independent Predictors of Cholesterol and HDL Cholesterol Identified by Stepwise Regression Analysis (n ⴝ 83)
Dependent Variable
Cholesterol
HDLcholesterol
Predictors
 ⫾ SE
P Value
Female gender Serum creatinine FCHL Cyclosporine Triglycerides Male gender
53.61 ⫾ 17 33.87 ⫾ 20 31.31 ⫾ 14 25.95 ⫾ 19 0.024 ⫾ 0.012 14.62 ⫾ 2.5
0.005 0.01 0.003 0.01 0.05 0.001
Serum creatinine 12.62 ⫾ 2.2 0.001 -blockers 6.45 ⫾ 2.2 0.01 Age ⫺0.21 ⫾ 0.07 0.01 Apolipoprotein B 0.16 ⫾ 0.04 0.001 NOTE. The adjusted r 2 for the cholesterol model was 0.293 (P ⬍ 0.001). The adjusted r 2 for HDL-cholesterol was 0.82 (P ⬍ 0.001).
DYSLIPIDEMIA POST-TRANSPLANTATION IN MEXICO
175
dependent variable (Table 5). The adjusted r2 of the model was highly significant (r2 ⫽ 0.82; P ⬍ 0.001). Gender, apoB, serum creatinine, age, and the use of cyclosporine remained as independent predictors.
remarked to avoid wrong extrapolations between dissimilar populations. The importance of genetic factors in posttransplant dyslipidemia is shown by the data reported here. The assessment of the lipid profile of first-degree relatives has not been included as part of the study of post renal-transplant patients. In this population, the presence of at least two relatives with an abnormal lipid profile was associated strongly with post-transplant dyslipidemia. These abnormalities are suggestive of the presence of FCHL, an atherogenic form of primary dyslipidemia. This disorder is common in Mexico.16,26,27 Its prevalence in the general population is not known, but the typical abnormalities of FCHL (hypertriglyceridemia and mixed hyperlipidemia) are among the most frequent abnormal lipid patterns found in urban Mexican adults (24.3% and 12.6%). This dyslipidemia was a strong predictor not only for an abnormal lipid profile, but also it was associated with a more severe form of dyslipidemia. High lipid levels were observed in FCHL cases during cyclosporine treatment (Table 3). The deleterious effect of cyclosporine on the lipid profile was shown in non-FCHL subjects or subjects in whom the family study was not available. The latter subjects had significantly lower lipid levels, however, compared with the cyclosporine-treated/ FCHL patients. By multivariate analysis, cyclosporine treatment and FCHL were identified as independent predictors of the cholesterol levels. These observations suggest that FCHL and cyclosporine have independent adverse effects on the lipid profile. The higher lipid levels found in cases with both abnormalities suggest an additive effect. Possible interactions between these conditions could be assessed with certainty only in a prospective randomized trial. The higher uric acid concentrations found in the cyclosporine group were not modified by the coexistence of FCHL. These data suggest that the evaluation of the lipid profile of the closest relatives may be a valuable tool for assessment of the risk of having post-transplant dyslipidemia and, perhaps, for selecting the immunosuppressive agent. Similar additive effects of FCHL with other drugs associated with an adverse lipid profile, such as sirolimus,28 must be looked for. The importance of pretransplant lipid abnormalities in the pathogenesis of post-transplant
DISCUSSION
This study shows a high prevalence of dyslipidemia in Mexican renal transplant patients. The mean lipid concentrations are above the 80th percentiles of the values found in Mexican urban adults. Important differences are found, however, when compared with reports from white populations. Compared with the levels found in Nordic19 and U.S.20 reports, the mean cholesterol concentration for Mexican post-transplant patients was 50 mg/dL lower. Also the prevalence of cholesterol levels greater than 240 mg/dL was almost half of the prevalence reported by Kasiske5 (33% versus 63%). In contrast, the mean cholesterol value found in this study is similar to that found in Asian patients.8,21 The reasons for the differences found between Mexican and white patients may be in the background lipid levels of these populations. Prevalence of hypercholesterolemia (ⱖ240 mg/dL [6.3 mmol/L]) in Mexican adults is lower than observed in every ethnic group (including Mexican Americans) included in the second and third annual National Health and Nutrition Surveys (NHANES II and III)22; Mexican Americans had the lowest prevalence of hypercholesterolemia of the racial groups included in those reports.23 The same phenomenon was observed for increased levels of Lp(a). The prevalence reported in white subjects is 35% higher (13% versus 23%) compared with the prevalence reported here; our mean Lp(a) concentration is remarkably similar to the data in Japan.21 It has been proved that the mean Lp(a) values are significantly lower in Asian and Mexican populations.17,24 The main contributor of the Lp(a) concentration is genetically determined. The distribution of the Apo(a) isoforms has not been described in Hispanic groups. Genetic factors may protect the Mexican population against hypercholesterolemia or increased Lp(a) levels even after a renal transplant. The different epidemiologic behavior of the lipid abnormalities has not been recognized systematically in the reviews of this topic25; these differences must be
176
dylispidemia is shown by the significant correlation found between the pretransplant and posttransplant concentrations of cholesterol, HDL-C, and triglycerides. These results are in accordance with previous observations9 showing that pretransplant lipoprotein disturbances predict posttransplant dyslipidemia and a worse graft function. The importance of several environmental factors also was assessed. The prevalence of obesity was close to that reported in nontransplant Mexican adults.29,30 None of the evaluated parameters of the diet was an independent predictor of plasma lipid levels. These data are in accordance with the little effect on plasma lipid levels observed after dietary advice and lifestyle modifications in post-transplant patients.31 In contrast, the graft function was a determinant of the lipid concentrations. Patients with low HDL-C had higher urine albumin concentrations. Also, by multivariate analysis, serum creatinine concentration was identified as an independent predictor of the cholesterol levels The cross-sectional nature of the study allows us only to identify associations, which do not infer or discard a causal relationship. Prospective studies are required to explore in detail the effects of diet and graft function in transplant patients. The effects of some cardiovascular drugs were assessed. As previously reported,32,33 -blockers were associated with a worse lipid profile. These actions were independent of the confounding effect of age, gender, and body mass index. The use of -blockers should be considered carefully in a post-transplant patient. Our data are in accordance with the well-known effects of prednisone and cyclosporine on the lipid profile.34-37 Prednisone dose was associated with higher HDL-C, HDL-2 cholesterol, and glucose levels. As described earlier, the use of cyclosporine in members of FCHL kindreds was associated with significantly worse lipid concentrations. The undesirable effects of cyclosporine on the lipid profile were shown in patients without FCHL. In conclusion, these data showed that the prevalence of post-transplant dyslipidemia is determined by genetic and environmental factors. The qualitative abnormalities observed on the lipid profile were similar to those reported in white subjects, but the magnitude of the abnormalities was remarkably lower in our Mexican subjects,
AGUILAR-SALINAS ET AL
reflecting the background lipoprotein pattern of the population. The presence of FCHL in the relatives is a strong marker for developing posttransplant hyperlipidemia, and its presence is associated with a more severe form of lipoprotein disturbances. This was especially true for cyclosporine-treated patients. Our data suggest that evaluation of the lipid profile of the closest relatives may be a valuable tool for the evaluation of the risk of having post-transplant dyslipidemia and perhaps for selecting an immunosuppressive agent. Other parameters, including the selection of immunosuppressive agents and -blockers and the severity of albuminuria, are important contributors to post-transplant lipid abnormalities. These results support the need for further prospective and cross-sectional studies designed to disclose the interaction of genetic factors with several acquired factors in posttransplant patients. REFERENCES 1. Kasiske B, Vazquez MA, Harmon W: Recommendations for outpatient surveillance of renal transplant recipients. J Am Soc Nephrol 11:S1-S86, 2000 2. Kasiske B: Risk factors for accelerated atherosclerosis in renal transplant recipients. Am J Med 84:985-992, 1988 3. Arkhus S, Dahl K, Wideroe TE: Cardiovascular morbidity and risk factors in renal transplant patients. Nephrol Dial Transplant 10:95-100, 1995 (suppl 1) 4. Wood D, De Backer G, Faergeman O: Prevention of coronary heart disease in clinical practice: Recommendations of the Second Joint Task Force of European and other Societies on Coronary Prevention. Atherosclerosis 140:199270, 1998 5. Kasiske B: Hyperlipidemia in patients with chronic renal disease. Am J Kidney Dis 32:S142-S156, 1998 6. Wheeler DC, Morgan R, Thomas DM: Factors influencing plasma lipid profiles including lipoprotein (a) concentrations in renal transplant recipients. Transpl Int 9:221-226, 1996 7. Mahley RW, Palaoglu KE, Atak Z: Turkish Heart Study: Lipids, lipoproteins and apolipoproteins. J Lipid Res 36:839-859, 1995 8. Li CS, Chau KF, Mak CK: Cadaveric kidney transplantation in a single center in Hong Kong. Transplant Proc 26:2012-2013, 1994 9. Dimeny E, Tufveson G, Lithell H: The influence of pretransplant lipoprotein abnormalities on the early results of renal transplantation. Eur J Clin Invest 23:572-579, 1993 10. Lipid and lipoprotein analysis, in Manual of Laboratory Operations: Lipid Research Clinics Program Report. NIH Publication 75-628. Washington, DC, U.S. Department of Health, Education and Welfare, 1974, pp 20-25 11. Siedel J, Schiefer S, Rosseneau M: Immunoturbidimetric method for routine determinations of apolipoproteins A-I, A-II and B in normo and hyperlipidemic sera: Compara-
DYSLIPIDEMIA POST-TRANSPLANTATION IN MEXICO
177
tive evaluation against immuno nephelometry. Clin Chem 34:1821-1825, 1998 12. Patsch W, Brown SA, Morriett JD, Gotto AM, Patsch JR: A dual-precipitation method evaluated for measurement of cholesterol in high-density lipoprotein subfractions HDL2 and HDL3 in human plasma. Clin Chem 35:265-270, 1989 13. Hixson JE, Vernier DT: Restriction isotyping of human apolipoprotein E by gene amplification and cleavage with HhaI. J Lipid Res 31:545-548, 1990 14. Aguilar-Salinas CA, Barrett H, Pulai J, et al: A familial combined hyperlipidemic kindred with impaired apolipoprotein B catabolism: Kinetics of apolipoprotein B during placebo and pravastatin therapy. Arterioscler Thromb Vasc Biol 17:72-82, 1997 15. Matthews D, Hosler JP, Rudenski AS: Homeostasis model assessment: Insulin resistance and B-cell function from fasting plasma glucose and insulin concentration in man. Diabetologia 28:412-419, 1985 16. Aguilar-Salinas CA, Olaiz G, Valles V, et al: High prevalence of low HDL cholesterol concentrations and mixed hyperlipidemia in a Mexican nationwide survey. J Lipid Res 42:1298-1307, 2001 17. Aguilar-Salinas CA, Talavera G, Guille´n LE, et al: The apolipoprotein E4 allele is not associated with an atherogenic lipid profile in a Native-American population following its traditional lifestyle. Atherosclerosis 142:409414, 1999 18. Valdez R, Stern M, Howard B: Apolipoprotein E polymorphism and insulin levels in a biethnic population. Diabetes Care 18:992-1000, 1995 19. Aarkhus S, Dahl K, Wideroe TE: Hyperlipidemia in renal transplant patients. J Intern Med 239:407-415, 1996 20. Chatterjee S, Chin HP, Azen S: Abnormal serum lipid patterns in primary renal allograft recipients. Surgery 82:655659, 1977 21. Sugahar S, Koyam I, Yoshikawa Y: Lipid and lipoprotein (a) in renal transplant recipients. Transplant Proc 26: 2080-2081, 1994 22. Dixon LB, Sudquist J, Winkleby M: Differences in energy, nutrient and food intakes in a US sample of Mexican American women and men: Findings from the Third National Health and Nutrition Examination Survey, 1988-1994. Am J Epidemiol 152:548-557, 2000 23. Loria C, Bush T, Carroll M: Macronutrient intakes among adult Hispanics: A comparison of Mexican Americans, Cuban Americans, and Mainland Puerto Ricans. Am J Public Health 85:684-689, 1995
24. Gaw A, Boerwinkle E, Cohen J. Comparative analysis of the apo (a) gene, apo (a) glycoprotein, and plasma concentration of Lp(a) in three ethnic groups. J Clin Invest 93:2526-2534, 1994 25. Fellstrom B: Impact and management of hyperlipidemia posttransplantation. Transplantation 70:SS51-SS57, 2000 26. Aguilar-Salinas CA, Barrett H, Schonfeld G: Metabolic modes of action of statins in the hyperlipoproteinemias. Atherosclerosis 141:203-207, 1998 27. Grundy SM, Chait A, Brunzell J: Familial combined hyperlipidemia workshop. Arteriosclerosis 7:203-207, 1987 28. Groth CG, Backman L, Morales JM: Sirolimus based therapy in human renal transplantation: Similar efficacy and different toxicity compared with cyclosporine. Transplantation 67:1036, 1999 29. Aguilar-Salinas CA, Vazquez-Chavez C, GamboaMarrufo R, et al: Prevalence of obesity, diabetes, hypertension and tobacco consumption in an urban adult Mexican population. Arch Med Res 32:446-453, 2001 30. Arroyo P, Loria A, Fernandez V, et al: Obesity in Hispanic Americans. Diabetes Care 14:691-694, 1991 31. Rao VK: Posttransplant medical complications. Surg Clin North Am 78:113-119, 1998 32. Sawicki PT, Berger M: Pharmacological treatment of diabetic patients with cardiovascular complications. J Intern Med 243:181-189, 1998 33. Troein M, Gardell B, Selander S: Guidelines and reported practice for the treatment of hypertension and hypercholesterolemia. J Intern Med 242:173-178, 1997 34. Hillbrands LB, Demacker PNM, Hoitsma AJ: The effects of cyclosporine and prednisone on serum lipids and (apo)lipoprotein levels in renal transplant recipients. J Am Soc Nephrol 5:2073-2076, 1995 35. Ponticelli C, Civati G, Tarantino A: Randomized study with cyclosporine in kidney transplantation: 10 year follow up. J Am Soc Nephrol 7:792-797, 1996 36. von Ahsen N, Helmhold M, Schutz E, Eisenhauer T, Armstrong VW, Oellerich M: Cyclosporin A trough levels correlate with serum lipoproteins and apolipoproteins: Implications for therapeutic drug monitoring of cyclosporin A. Ther Drug Monit 19:140-145, 1997 37. Fryer JP, Granger DK, Leventhal JR: Steroid related complications in the cyclosporine era. Clin Transplant 8:224229, 1994
Genetic Factors Play an Important Role in the Pathogenesis of Hyperlipidemia Post-Transplantation Carlos A. Aguilar-Salinas, MD, Araceli Dı´az-Polanco, MD, Eduardo Quintana, MD, Nayeli Macias, RD, Adriana Arellano, RD, Erika Ramı´rez, MD, Marı´a Luisa Ordo´n˜ez, PhD, Consuelo Vela´squez-Alva, PhD, Francisco J. Go´mez Pe´rez, MD, Josefina Alberu´, MD, and Ricardo Correa-Rotter, MD ● Background: Our purpose was to identify factors associated with hyperlipidemia post-transplantation in a Hispanic population. Methods: From 1985 to 1999, a kidney graft survival longer than 3 months occurred in 293 cases at the Instituto Nacional de la Nutricio´n. Most of the patients living in Mexico City were included (n ⴝ 83). The evaluation included a questionnaire, blood samples, and assessment of body composition and dietary habits. As many as possible first-degree relatives were studied. Results: Women had higher values of cholesterol (236 ⴞ 51 versus 215 ⴞ41; P < 0.05), low-density lipoprotein cholesterol (147 ⴞ 42 versus 131 ⴞ 34; P ⴝ 0.05), high-density lipoprotein cholesterol (57.3 ⴞ 14 versus 47.9 ⴞ 14; P ⴝ 0.002) and high-density lipoprotein-2 cholesterol. Isolated hypercholesterolemia was the most common lipid abnormality (40.9%), followed by mixed hyperlipidemia. Lipoprotein (a) greater than 30 mg/dL was found in 13 cases. Familial combined hyperlipidemia (FCHL) in the patient’s relatives was a marker for dyslipidemia (odds ratio, 7.04; 95% confidence interval, 1.2 to 59.7). These cases had a worse lipid profile. Cyclosporine-treated FCHL patients had higher lipid levels compared with the non-FCHL, cyclosporine-treated patients. The effects of cyclosporine on the lipid levels were lower, but significant, after the exclusion of the FCHL cases. Conclusion: Post-transplant dyslipidemia is determined by genetic and environmental factors. FCHL in the patient’s relatives was associated with post-transplant hyperlipidemia; an additive effect with cyclosporine was found. The evaluation of the lipid profile of relatives may be useful for the assessment of the risk of post-transplant dyslipidemia. © 2002 by the National Kidney Foundation, Inc. INDEX WORDS: Cholesterol; renal transplantation; familial combined hyperlipidemia (FCHL); cyclosporine; Mexico.
T
HE LONG-TERM MORBIDITY rates and mortality associated with renal transplantation are still high, and the most prevalent cause is cardiovascular disease.1 The incidence of new ischemic cardiac events in a 4-year period was almost 11% in patients without prior coronary heart disease and 15% in secondary prevention cases.2 These rates are almost five times higher than predicted from the tables derived from the Framingham Heart Study.3 Dyslipidemia is one of the main contributors to the progression of vascular lesions that ultimately result in a coronary event. The correction of dyslipidemia is an effective alternative for the prevention of cardiovascular events in nontransplant populations.4 The lipid abnormalities encompassed by the term dyslipidemia include hypercholesterolemia, hypertriglyceridemia, low high-density lipoprotein cholesterol (HDL-C), and combinations of these defects. In the combined results from five studies done in white subjects, 63% of 549 patients had cholesterol levels greater than 240 mg/dL. Low-density lipoprotein cholesterol (LDL-C) greater than 130 mg/dL was found in
almost 60% of cases.5 Low HDL cholesterol levels are not as common as the previous two abnormalities (approximately 12%). The prevalence of hyperlipidemia varies, and the type of immunosuppressive medications, nutritional status, dietary habits, and genetic predisposition are among the main determinants for the development of these lipid disturbances.6 Some determinants of dyslipidemia in kidney transplant patients have not been studied extensively. The data reported in white groups may not be reproducible in populations with a different genetic From the Departments of Endocrinology and Metabolism, Nephrology and Mineral Metabolism, and Transplants, Instituto Nacional de Ciencias Me´dicas y Nutricio´n Salvador Zubira´n; and Department of Healthcare, Universidad Auto´noma Metropolitana, Unidad Xochimilco, Mexico City, Mexico. Received October 26, 2001; accepted in revised form March 4, 2002. Address reprint requests to Carlos A. Aguilar-Salinas, MD, Vasco de Quiroga 15, Mexico City 14000 Mexico. E-mail: [email protected] © 2002 by the National Kidney Foundation, Inc. 0272-6386/02/4001-0021$35.00/0 doi:10.1053/ajkd.2002.33926
American Journal of Kidney Diseases, Vol 40, No 1 (July), 2002: pp 169-177
169
170
AGUILAR-SALINAS ET AL
background. The prevalence of hypertriglyceridemia and hypoalphalipoproteinemia is low in white subjects, but the genetic predisposition for having these abnormalities is strong in Hispanic and Asian groups.7 Mean cholesterol levels seem to be significantly lower in post-transplant Asian patients compared with levels found in white patients.8 The influence of pretransplant lipoprotein abnormalities, including the coexistence of a genetic form of dyslipidemia, has been partially studied.9 Finally, the clinical correlations of several lipoprotein profiles (ie, mixed hyperlipidemia or isolated hypercholesterolemia) have not been described. The purpose of this study was to describe the prevalence of lipid abnormalities and their determinants in a Mexican kidney transplant population. The evaluation included a family study for primary dyslipidemias, assessment of body composition, dietary habits, and apoE genotype. MATERIAL AND METHODS
Patients Informed consent was obtained from all participants of the study. From 1985 to 1999, 315 kidney transplants were done successfully at the Instituto Nacional de Ciencias Me´dicas y Nutricio´n Salvador Zubira´n. Among them, graft survival longer than 3 months occurred in 293 cases. Many of these patients lived outside Mexico City (70% of the population). Most of the patients living in Mexico City were reached and invited to participate. Of the total population, 28% (n ⫽ 83) was included. Each patient was evaluated by an endocrinologist who completed a questionnaire and a physical examination. A registered dietitian collected a 3-day prospective dietary registry using a previously validated format. The data were analyzed using the SCVAN program (Instituto Nacional de la Nutricio´n, Mexico City, Mexico), useful for the analysis of Mexican cooking. All patients were requested to ask as many as possible first-degree relatives to come to the clinic to have a complete lipid profile and glucose concentration done. It was possible to study at least two relatives of 51 probands (61% of the population). Blood pressure was measured to the nearest even digit with a sphygmomanometer with the subject in the supine position after a 5-minute rest. Participants removed their shoes and upper garments. Height was measured to the nearest 0.5 cm. Body weight was measured on a dailycalibrated balance and recorded to the nearest 0.1 kg. Body mass index was calculated and used as an index of overall adiposity. Body composition of the probands was measured using electric bioimpedance technology.
Methods All analytical measurements were done at the Department of Endocrinology and Lipid Metabolism of the Instituto
Nacional de Ciencias Me´dicas y Nutricio´n Salvador Zubira´n. Blood samples were obtained from the subjects after having fasted for 9 to 12 hours; they remained sitting for 5 minutes before the blood was drawn. Glucose was analyzed by the glucose-oxidase method (Boehringer, Mannheim, Germany). Serum concentrations of total cholesterol and triglycerides were determined by enzymatic methods (Boehringer, Mannheim, Germany). HDL-C was measured after precipitation of very low-density lipoproteins (VLDL) and LDL by the phosphotungstate method (Boehringer, Mannheim, Germany). LDL-C was measured by ultracentrifugation using a -quantification method.10 Briefly, 3 mL of plasma was ultracentrifuged at d 1.006 kg/L for 18 hours at 105,000 ⫻ g. The VLDL fraction, at the top, was removed; the cholesterol and HDL-C concentrations were measured in the bottom fraction. LDL-C was calculated as the difference between that in the bottom fraction and HDL-C. Intra-assay variation coefficient values for total cholesterol, triglycerides, and HDL-C were 3%, 5%, and 5%. Insulin was analyzed by enzyme-linked immunosorbent assay in the ES-33 system (Boehringer, Mannheim, Germany). The crossreactivity with proinsulin for this assay was 40%. Our laboratory followed standardization procedures according to the World Health Organization recommendations, including the use of external control sera. Apolipoproteins A-1 (apoAI) and B (apoB) and lipoprotein (a) (Lp(a)) were measured with an immunoturbidimetric method11 on a Hitachi 705 analyzer (Hitachi Science Systems LTD, Hitachinaka, Japan); the assays were done according to the manufacturer’s instructions with manufacturer-provided calibrators and antisera. All samples were kept frozen at ⫺80°C until they were analyzed; the maximum time of storage was 1 month. Cholesterol concentration of the HDL subfractions was measured using a double precipitation assay.12 Apolipoprotein E (apoE) genotype was determined by gene amplification and cleavage with HhaI as described by Hixson and Vernier.13
Definitions Overweight was defined as a body mass index between 25 and 30 kg/m2. Obesity was defined as body mass index greater than or equal to 30 kg/m2. Individuals were diagnosed as diabetics if they had a previous diagnosis of diabetes or had a fasting blood glucose value greater than or equal to 126 mg/dL (7 mmol/L) and no previous history of diabetes. Hypertension was diagnosed when the systolic blood pressure was greater than or equal to 140 mm Hg or diastolic blood pressure was greater than or equal to 90 mm Hg or when antihypertensive medications were being used. Ischemic heart disease was considered if there was a history of myocardial infarction. Smoking was defined as any tobacco consumption during the previous month. Cholesterol value greater than or equal to 200 mg/dL (5.2 mmol/L) was considered to be hypercholesterolemia. HDL-C concentration was considered to be abnormal if it was less than 35 mg/dL (0.9 mmol/L). Mixed hyperlipidemia was defined as cholesterol greater than or equal to 200 mg/dL (5.2 mmol/L) plus triglycerides greater than or equal to 200 mg/dL (2.2 mmol/L). Familial combined hyperlipidemia (FCHL) was considered to be present if besides the proband, at least two family members had two different lipid profiles (ie, isolated
DYSLIPIDEMIA POST-TRANSPLANTATION IN MEXICO
171
hypercholesterolemia, isolated hypertriglyceridemia, or mixed hyperlipidemia).14 Insulin sensitivity was estimated using the Homeostasis Model Assessment (HOMA) score based on the formula: fasting serum insulin (U/mL) ⫻ fasting plasma glucose (mmol/L)/22.5, as described by Matthews et al.15 With this method, high HOMA scores denote low insulin sensitivity.
received prednisone and azathioprine. Cyclosporine was used in 51 cases. One subject had proteinuria of 3 g/d or more. Diabetes was found in four subjects; only one had hyperglycemia before the kidney transplant.
Statistical Analysis
Lipid Profiles, Prevalence, and Associated Clinical Characteristics For both genders, cholesterol, LDL-C, and non-HDL-C were consistently higher in patients compared with their relatives (Table 2). Women had significantly higher values of cholesterol (236 ⫾ 51 versus 215 ⫾ 41; P ⬍ 0.05), LDL-C (147 ⫾ 42 versus 131 ⫾ 34; P ⫽ 0.05), HDL-C (57.3 ⫾ 14 versus 47.9 ⫾ 14; P ⫽ 0.002), and HDL-2 cholesterol (19.8 ⫾ 12.6 versus 11.7 ⫾ 7.4; P ⬍ 0.001). These concentrations were above the 90th, 80th, and 90th percentiles of the cholesterol, LDL-C, and HDL-C values reported in Mexican urban women.16 For men, the corresponding percentiles were 80th, 70th, and 80th. No differences were found between genders in triglycerides and apoB concentrations; the mean values were above the 60th and 80th percentiles. The prevalence of dyslipidemia was high in this population. Triglyceride values greater than 200 mg/dL (2.2 mmol/L) were found in 26% of cases (an additional 19.7% had levels between 150 and 199 mg/dL). Cholesterol values greater than 240 mg/dL (6.3 mmol/L) were found in 34% (an additional 29% had levels between 200 and 239 mg/dL). Finally, 25% had LDL-C levels greater than 160 mg/dL (an additional 26% had levels between 130 and 159 mg/dL), and 12% had HDL-C less than 35 mg/dL (0.9 mmol/L). Of patients, 26% had HDL-C greater than 60 mg/dL (1.6 mmol/L). Lp(a) concentrations greater than 30 mg/dL were found in 13 cases. Isolated hypercholesterolemia was the abnormal lipid profile most frequently found in this population (34 subjects [40.9%]); this abnormality was more prevalent in women (53%) than in men (29.5%). Mixed hyperlipidemia also was common (19 cases [22.8%]); this abnormality was more prevalent in men (27% versus 17%; P ⬍ 0.05). Low HDL cholesterol (⬍35 mg/dL) was found in 10 cases (12%); 8 were men (18.1%). A normal lipid profile (including HDL-C) was found in 21 cases.
The database was validated through recognition of missing values, outliers, and inconsistencies between variables. Descriptive analysis included the estimation of mean values and SDs for continuous variables. Prevalence and frequencies were expressed as percentages. Significance of the differences between the subgroups of the population were tested by one-way analysis of variance using Scheffe´ multiple comparison method. Categorical variables were compared by chi-square with Yates correction or the Fisher exact test when appropriate. Correlation coefficients were calculated for measuring the association between continuous variables. Analysis of covariance was used for gender adjustment of the variables. A stepwise multiple regression analysis was done with the lipid concentrations as dependent variables. All statistical analyses were done using the Statgraphics statistical package (STSC, Rockville, MD).
RESULTS
The study subjects were young adults who had undergone a kidney transplant 5.6 years ago (Table 1) and had a serum creatinine of 1.46 ⫾ 0.56 mg/dL (129 ⫾ 49.5 mmol/L). All patients Table 1.
Characteristics of the Patients
N Age (y) Gender (% males) Time after renal grafting (y) Serum creatinine (mg/dL) Systolic blood pressure Diastolic blood pressure Albuminuria ⬎ 1 g/L [n (%)] Hypertension [n (%)] Diabetes mellitus [n (%)] Cardiovascular acting drugs [n (%)] Diuretics Beta blocker Calcium channel blocker ACE inhibitor Statins Fibrates Immunosuppression [n (%)] Prednisone plus azathioprine Prednisone plus cyclosporine plus azathioprine Prednisone dose (mg/d) Cyclosporine dose (mg/kg/d)
83 37.2 ⫾ 12.6 53% 5.6 ⫾ 2.2 1.46 ⫾ 0.56 123.5 ⫾ 13.3 81.1 ⫾ 10.9 6 (7.2%) 61 (73.4) 4 (4.8) 4 21 43 9
(4.8) (25.3) (51.8) (10.8) 11% 4%
32 (38%) 51 (61.4%) 17.4 ⫾ 7.3 2.9 ⫾ 2.6
Abbreviation: ACE, angiotensin-converting enzyme.
172 Table 2.
AGUILAR-SALINAS ET AL Metabolic Profile of the Study Subjects and Their First-Degree Relatives Including Cases With Familial Combined Hyperlipidemia Cases
n Age Body mass index (kg/m2) Glucose (mg/dL) Cholesterol (mg/dL) Triglycerides (mg/dL) HDL cholesterol (mg/dL) LDL cholesterol (mg/dL) Non-HDL cholesterol (mg/dL) HDL-2 cholesterol Lipoprotein (a) (mg/dL) Insulin (U/mL) Mixed hyperlipidemia (n [%]) HDL cholesterol ⬍ 35 mg/dL
Relatives
Total Cases
FCHL Cases
Non-FCHL Cases
Without Family Study
FCHL Kindreds
Non-FCHL Kindreds
P Value
83 37.2 ⫾ 12.6 23.3 ⫾ 3.3 74 ⫾ 19 225 ⫾ 47.4*† 169 ⫾ 91.3*† 52 ⫾ 14.7*† 139 ⫾ 38.8† 173 ⫾ 45.8† 15.5 ⫾ 11 17.5 ⫾ 28.4 12.9 ⫾ 6 19 (22.8)† 10 (12)
14 37.9 ⫾ 13.6 24.5 ⫾ 2.1 70 ⫾ 11 266 ⫾ 51*†‡ 240 ⫾ 121*†‡ 49 ⫾ 14*† 170 ⫾ 34*†‡ 217 ⫾ 46*†‡ 14.7 ⫾ 9.3 20 ⫾ 27 10.5 ⫾ 4 8 (57)*†‡ 3 (21)
37 36.1 ⫾ 9.7 23 ⫾ 3 72 ⫾ 19 211 ⫾ 40† 158 ⫾ 87† 51 ⫾ 13† 128 ⫾ 34† 160 ⫾ 40† 15.4 ⫾ 11 16.8 ⫾ 28 13.5 ⫾ 6 4 (11) 5 (13.5)
32 37.2 ⫾ 10 23.2 ⫾ 4 74 ⫾ 20 223 ⫾ 44 149 ⫾ 64 54 ⫾ 15 138 ⫾ 31 168 ⫾ 40 15.8 ⫾ 11 17.3 ⫾ 28 13.1 ⫾ 6 7 (21) 2 (6)
47 — — — 217 ⫾ 37† 203 ⫾ 127† 38.2 ⫾ 10.7† 138 ⫾ 38† 178 ⫾ 37† — — — 15 (32)† 2 (4)
75 — — — 185 ⫾ 32 133 ⫾ 78 44.7 ⫾ 11 114 ⫾ 28 140 ⫾ 33 — — — 8 (10.6) 7 (9.3)
NS NS NS ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 NS NS NS ⬍0.001 NS
NOTE. Data expressed as mean ⫾ SD. FCHL was considered to be present if besides the proband, at least two family members had two different lipid profiles (ie, isolated hypercholesterolemia, isolated hypertriglyceridemia, or mixed hyperlipidemia). Mixed hyperlipidemia was defined as cholesterol ⱖ200 mg/dL (5.2 mmol/L) plus triglycerides ⱖ 200 mg/dL (2.2 mmol/L). Conversion factors to obtain SI units: glucose ⫽ 0.05551; cholesterol, HDL cholesterol, LDL cholesterol, non-HDL cholesterol and HDL-2 cholesterol ⫽ 0.02586; triglycerides ⫽ 0.0112. Abbreviation: NS, not significant. *P ⬍ 0.05 versus relatives from FCHL kindreds. †P ⬍ 0.05 versus relatives from non-FCHL kindreds. ‡P ⬍ 0.05 versus non-FCHL cases.
Diet and Body Composition The body mass index was similar in both genders (23.1 ⫾ 3 kg/m2 versus 23.4 ⫾ 3.6 kg/m2 for men and women). A greater percentage of women were overweight or obese (29% versus 13%; P ⬍ 0.05). Waist circumference was slightly greater in men (87.6 ⫾ 10 cm versus 84.5 ⫾ 16 cm; P ⫽ 0.2). As a mean, patients received a carbohydrate-rich, low-fat, low-cholesterol diet. Men consumed more calories than women (2,406 ⫾ 1,156 cal/d versus 2,032 ⫾ 908 cal/d; P ⫽ not significant). For men, a greater percentage of the calories came from fat (29.8% versus 24.9%; P ⬍ 0.05). This difference was explained mainly by a greater consumption of saturated (7.6% versus 6.4%; P ⫽ 0.05) and monounsaturated fat (8.4% versus 6.3%; P ⫽ 0.01). Also, the cholesterol content of the diet was higher in men (282 ⫾ 188 mg/d versus 181 ⫾ 111 mg/d; P ⫽ 0.01). No correlation was found between the components of the diet and the plasma lipid levels. The body composition was assessed in every
case. The amount of fat was significantly greater in women (17 ⫾ 1.3 kg versus 7.1 ⫾ 1.2 kg; P ⬍ 0.001). In contrast, the amount of lean mass was greater in men (57.4 ⫾ 1.1 kg versus 41.8 ⫾ 1.2 kg; P ⬍ 0.001). No correlation was found between the body composition and either cholesterol or apoB levels. In men, a strong inverse relationship was found, however, between fat mass and HDL-C (r ⫽ ⫺0.33; P ⫽ 0.05); no association was observed with the HDL-2 cholesterol. In men, an association was found between insulin (r ⫽ 0.51; P ⫽ 0.008) and HOMA (r ⫽ 0.39; P ⬍ 0.05) scores with fat mass. Genetic Factors FCHL was a strong marker for having lipid disturbances (odds ratio [OR], 7.04; 95% confidence interval [CI], 1.2 tp 59.7). The characteristics of the 14 FCHL probands were compared with the 37 patients who had relatives with nondiagnostic lipid changes (Table 2). The abnormalities of the lipid profile were more severe and common in the FCHL cases. These subjects had
DYSLIPIDEMIA POST-TRANSPLANTATION IN MEXICO Table 3.
173
Effects of Cyclosporine Treatment in Cases With or Without Familial Combined Hyperlipidemia Cyclosporine All
n 51 Age (y) 36.9 ⫾ 13.2 23.4 ⫾ 2.8 Body mass index (kg/m2) Gender (M/F) 29/22 Creatinine (mg/dL) 1.52 ⫾ 0.4 Cholesterol (mg/dL) 235 ⫾ 48* Triglycerides (mg/dL) 182 ⫾ 86* Non HDL-C (mg/dL) 186 ⫾ 44* LDL-C (mg/dL) 150 ⫾ 38* Apolipoprotein B (mg/dL) 109.2 ⫾ 27* HDL-C (mg/dL) 49 ⫾ 12* HDL-2 cholesterol (mg/dL) 14.3 ⫾ 9.7 BUN (mg/dL) 24.8 ⫾ 9* Uric acid (mg/dL) 7.6 ⫾ 1.4*
No Cyclosporine
FCHL⫹
FCHL⫺
FCHL Unknown
FCHL⫺
FCHL Unknown
P Value
13 38.3 ⫾ 11 24.8 ⫾ 2.1† 8/5 1.61 ⫾ 0.5 264 ⫾ 52† 224 ⫾ 110† 216 ⫾ 47† 171 ⫾ 34† 130 ⫾ 30† 48.3 ⫾ 14 14.5 ⫾ 12 28.4 ⫾ 9 7.7 ⫾ 1.6
23 36 ⫾ 14 22.9 ⫾ 2.9 13/10 1.47 ⫾ 0.9 216 ⫾ 42‡ 174 ⫾ 72‡ 168 ⫾ 39‡ 133 ⫾ 37‡ 97.7 ⫾ 25 47.6 ⫾ 11‡ 12.3 ⫾ 7‡ 23.1 ⫾ 9 7.7 ⫾ 1.4‡
15 36.8 ⫾ 9 23.1 ⫾ 1.5 8/7 1.51 ⫾ 0.1 239 ⫾ 47§ 158 ⫾ 96 187 ⫾ 42§ 156 ⫾ 40§ 108 ⫾ 28§ 51.7 ⫾ 12§ 17.2 ⫾ 9 24.4 ⫾ 9 7.3 ⫾ 1.6§
14 35.2 ⫾ 11 22.6 ⫾ 4 7/7 1.47 ⫾ 0.6 203 ⫾ 42§ 133 ⫾ 96 146 ⫾ 39 119 ⫾ 33 92 ⫾ 22 57.5 ⫾ 16 18.1 ⫾ 12 21 ⫾ 10 5.9 ⫾ 1.2
17 40.8 ⫾ 10 23.3 ⫾ 4 7/10 1.26 ⫾ 0.7 209 ⫾ 9 141 ⫾ 80 152 ⫾ 36 123 ⫾ 30 89 ⫾ 20 57.7 ⫾ 11 16.6 ⫾ 11 18.5 ⫾ 11 5.5 ⫾ 1.1
NS NS NS NS ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.01 NS ⬍0.05 ⬍0.001
NOTE. Data expressed as mean ⫾ SD. In the cyclosporine ⫹/FCHL⫺ group were included cases in which the family study was not diagnostic of FCHL. The FCHL unknown are cases with no family study. Differences between groups remained statistically significant after adjusting for body mass index. Conversion factors to obtain SI units: creatinine ⫽ 88.4; cholesterol, HDL cholesterol; LDL cholesterol, non-HDL cholesterol and HDL-2 cholesterol ⫽ 0.02586; triglycerides ⫽ 0.0112; insulin ⫽ 6.0; BUN ⫽ 0.357. Abbreviations: BUN, blood urea nitrogen; NS, not significant. *P ⬍ 0.05, cyclosporine-treated versus non– cyclosporine-treated cases. †P ⬍ 0.05, cyclosporine ⫹/FCHL⫹ versus cyclosporine ⫹/FCHL⫺ cases. ‡P ⬍ 0.05, cyclosporine ⫹/FCHL⫺ versus noncyclosporine/non-FCHL cases. §P ⬍ 0.05, cyclosporine ⫹/FCHL unknown versus noncyclosporine/FCHL unknown cases.
the highest lipid levels in this study group; more than half of them had a mixed hyperlipidemia. Significant differences in the lipid profile were found between the probands and their relatives, in FCHL and non-FCHL kindreds. The lipid abnormalities found in the FCHL relatives were similar to abnormalities observed in the nonFCHL transplant patients. These comparisons show the effect of renal transplantation on the lipid profile, independent of genetic factors. The apoE genotype distribution was assessed. Of subjects, 68 were e3/e3 homozygous and 15 were e3/e4 heterozygous. In agreement with the low prevalence of the e2 allele in Mexican populations,17 none of the study subjects had this genetic marker. Lipid levels were not statistically different between the e4/e3 heterozygous and the e3/e3 subjects. Two characteristics differ between these two groups, however. The e4/e3 heterozygous subjects had higher HDL-2 cholesterol levels (20.5 ⫾ 14 mg/dL versus 14.3 ⫾ 9.9 mg/dL; P ⬍ 0.05). Also, they had higher insulin concentrations (12.3 ⫾ 4.5 U/mL versus 15.4 ⫾ 10.5 U/mL; P ⬍ 0.05) and HOMA scores
(2.14 ⫾ 0.1 versus 4.18 ⫾ 0.96; P ⫽ 0.01). These data suggest the presence of a higher degree of insulin resistance in the e4/e3 heterozygous subjects, as previously reported by some authors.18 Effects of Immunosuppressive Therapy and Other Cardiovascular Drugs Prednisone dose was significantly related to HDL-C (r ⫽ 0.29; P ⬍ 0.05), HDL-2 cholesterol (r ⫽ 0.281; P ⬍ 0.05), glucose (r ⫽ 0.32; P ⬍ 0.05), and potassium levels (r ⫽ ⫺0.37; P ⬍ 0.05). Significant differences were found between patients treated with and without cyclosporine (Table 3). As expected, cyclosporine-treated patients had higher cholesterol, LDL-C, nonHDL cholesterol, apoB, and triglycerides concentrations. They also had lower HDL-C; this abnormality was explained by lower concentrations of HDL-3 cholesterol. Higher uric acid and blood urea nitrogen concentrations also were found in these cases. All but one subject with FCHL received cyclosporine. Because this association could play a role as a confounder for the analysis,
174
AGUILAR-SALINAS ET AL Table 4.
Univariate Correlations Between Blood Lipids and the Evaluated Parameters (n ⴝ 83)
Age (y) Male gender Body mass index (kg/m2) Diastolic blood pressure (mm Hg) Serum creatinine (mg/dL) Urine albumin (g/d) Pretransplant concentration Member of a FCHL kindred HOMA score Fat mass (kg) Calories (cal/d) Fat content of diet (%) Cyclosporine treatment Prednisone dosage (mg/d)
Cholesterol
Apolipoprotein B
Triglycerides
HDL-C
⫺0.059 ⫺0.215* 0.006 0.052 0.108* ⫺0.168 0.190* 0.499* ⫺0.005 0.077 0.059 ⫺0.067 0.315* 0.08
⫺0.024 ⫺0.02 0.053 0.112 0.091* 0.086 —† 0.261* ⫺0.049 0.010 0.102 ⫺0.018 0.342* 0.061
0.065 0.141 0.161 0.258* 0.057 0.258* 0.457* 0.354* ⫺0.35 0.026 0.038 ⫺0.009 0.297* 0.118
⫺0.133* ⫺0.311* ⫺0.168* ⫺0.253* ⫺0.069* ⫺0.183* 0.810* 0.073 ⫺0.016 ⫺0.048 ⫺0.112 ⫺0.149 ⫺0.24* ⫺0.293*
*P ⬍ 0.05. †No pretransplant apolipoprotein B values were available.
we analyzed separately the cyclosporine-treated patients with or without FCHL. Cases without a family study were also included as a separate group. The cyclosporine-treated/FCHL patients had remarkably higher cholesterol and apoB levels compared with the non-FCHL/cyclosporinetreated subjects. The effects of cyclosporine on the lipid profile are shown when the lipid concentrations are compared between the non-FCHL patients with or without cyclosporine treatment. The same trend was found between the patients in whom a family study was not available. The effects of other drugs on the lipid profile were analyzed. The use of angiotensin-converting enzyme inhibitors or calcium channel blockers was not related to changes in the evaluated parameters. -Blockers were associated with significantly higher triglycerides (207 ⫾ 18 mg/dL versus 148 ⫾ 14 mg/dL; P ⫽ 0.01), cholesterol (241 ⫾ 9.2 mg/dL versus 213 ⫾ 6.8 mg/dL; P ⫽ 0.01), LDL-C (152 ⫾ 7 mg/dL versus 131 ⫾ 5.7 mg/dL; P ⬍ 0.05), and apoB (115 ⫾ 5.5 mg/dL versus 95 ⫾ 4.8 mg/dL; P ⬍ 0.01). The number of patients who received diuretics (4.8%), statins (11%), or fibrates (4.8%) was too small to provide definitive conclusions. Associations Between the Evaluated Parameters and Blood Lipids As shown in Table 4, the presence of FCHL, cyclosporine treatment, and the pretransplant lipid levels were correlated with the concentration of
the apoB-containing lipoproteins. A stepwise regression model was constructed using cholesterol levels as a dependent variable (Table 5). The adjusted r2 of the model was 0.292 (P ⬍ 0.001). Gender, FCHL, serum creatinine, triglycerides, and the use of cyclosporine remained as independent predictors. We searched for factors associated with low HDL-C values. The pretransplant HDL-C values, cyclosporine treatment, and prednisone dosage were associated with lower levels of HDL-C in univariate analysis. A stepwise regression model was constructed using HDL-C levels as a Table 5. Independent Predictors of Cholesterol and HDL Cholesterol Identified by Stepwise Regression Analysis (n ⴝ 83)
Dependent Variable
Cholesterol
HDLcholesterol
Predictors
 ⫾ SE
P Value
Female gender Serum creatinine FCHL Cyclosporine Triglycerides Male gender
53.61 ⫾ 17 33.87 ⫾ 20 31.31 ⫾ 14 25.95 ⫾ 19 0.024 ⫾ 0.012 14.62 ⫾ 2.5
0.005 0.01 0.003 0.01 0.05 0.001
Serum creatinine 12.62 ⫾ 2.2 0.001 -blockers 6.45 ⫾ 2.2 0.01 Age ⫺0.21 ⫾ 0.07 0.01 Apolipoprotein B 0.16 ⫾ 0.04 0.001 NOTE. The adjusted r 2 for the cholesterol model was 0.293 (P ⬍ 0.001). The adjusted r 2 for HDL-cholesterol was 0.82 (P ⬍ 0.001).
DYSLIPIDEMIA POST-TRANSPLANTATION IN MEXICO
175
dependent variable (Table 5). The adjusted r2 of the model was highly significant (r2 ⫽ 0.82; P ⬍ 0.001). Gender, apoB, serum creatinine, age, and the use of cyclosporine remained as independent predictors.
remarked to avoid wrong extrapolations between dissimilar populations. The importance of genetic factors in posttransplant dyslipidemia is shown by the data reported here. The assessment of the lipid profile of first-degree relatives has not been included as part of the study of post renal-transplant patients. In this population, the presence of at least two relatives with an abnormal lipid profile was associated strongly with post-transplant dyslipidemia. These abnormalities are suggestive of the presence of FCHL, an atherogenic form of primary dyslipidemia. This disorder is common in Mexico.16,26,27 Its prevalence in the general population is not known, but the typical abnormalities of FCHL (hypertriglyceridemia and mixed hyperlipidemia) are among the most frequent abnormal lipid patterns found in urban Mexican adults (24.3% and 12.6%). This dyslipidemia was a strong predictor not only for an abnormal lipid profile, but also it was associated with a more severe form of dyslipidemia. High lipid levels were observed in FCHL cases during cyclosporine treatment (Table 3). The deleterious effect of cyclosporine on the lipid profile was shown in non-FCHL subjects or subjects in whom the family study was not available. The latter subjects had significantly lower lipid levels, however, compared with the cyclosporine-treated/ FCHL patients. By multivariate analysis, cyclosporine treatment and FCHL were identified as independent predictors of the cholesterol levels. These observations suggest that FCHL and cyclosporine have independent adverse effects on the lipid profile. The higher lipid levels found in cases with both abnormalities suggest an additive effect. Possible interactions between these conditions could be assessed with certainty only in a prospective randomized trial. The higher uric acid concentrations found in the cyclosporine group were not modified by the coexistence of FCHL. These data suggest that the evaluation of the lipid profile of the closest relatives may be a valuable tool for assessment of the risk of having post-transplant dyslipidemia and, perhaps, for selecting the immunosuppressive agent. Similar additive effects of FCHL with other drugs associated with an adverse lipid profile, such as sirolimus,28 must be looked for. The importance of pretransplant lipid abnormalities in the pathogenesis of post-transplant
DISCUSSION
This study shows a high prevalence of dyslipidemia in Mexican renal transplant patients. The mean lipid concentrations are above the 80th percentiles of the values found in Mexican urban adults. Important differences are found, however, when compared with reports from white populations. Compared with the levels found in Nordic19 and U.S.20 reports, the mean cholesterol concentration for Mexican post-transplant patients was 50 mg/dL lower. Also the prevalence of cholesterol levels greater than 240 mg/dL was almost half of the prevalence reported by Kasiske5 (33% versus 63%). In contrast, the mean cholesterol value found in this study is similar to that found in Asian patients.8,21 The reasons for the differences found between Mexican and white patients may be in the background lipid levels of these populations. Prevalence of hypercholesterolemia (ⱖ240 mg/dL [6.3 mmol/L]) in Mexican adults is lower than observed in every ethnic group (including Mexican Americans) included in the second and third annual National Health and Nutrition Surveys (NHANES II and III)22; Mexican Americans had the lowest prevalence of hypercholesterolemia of the racial groups included in those reports.23 The same phenomenon was observed for increased levels of Lp(a). The prevalence reported in white subjects is 35% higher (13% versus 23%) compared with the prevalence reported here; our mean Lp(a) concentration is remarkably similar to the data in Japan.21 It has been proved that the mean Lp(a) values are significantly lower in Asian and Mexican populations.17,24 The main contributor of the Lp(a) concentration is genetically determined. The distribution of the Apo(a) isoforms has not been described in Hispanic groups. Genetic factors may protect the Mexican population against hypercholesterolemia or increased Lp(a) levels even after a renal transplant. The different epidemiologic behavior of the lipid abnormalities has not been recognized systematically in the reviews of this topic25; these differences must be
176
dylispidemia is shown by the significant correlation found between the pretransplant and posttransplant concentrations of cholesterol, HDL-C, and triglycerides. These results are in accordance with previous observations9 showing that pretransplant lipoprotein disturbances predict posttransplant dyslipidemia and a worse graft function. The importance of several environmental factors also was assessed. The prevalence of obesity was close to that reported in nontransplant Mexican adults.29,30 None of the evaluated parameters of the diet was an independent predictor of plasma lipid levels. These data are in accordance with the little effect on plasma lipid levels observed after dietary advice and lifestyle modifications in post-transplant patients.31 In contrast, the graft function was a determinant of the lipid concentrations. Patients with low HDL-C had higher urine albumin concentrations. Also, by multivariate analysis, serum creatinine concentration was identified as an independent predictor of the cholesterol levels The cross-sectional nature of the study allows us only to identify associations, which do not infer or discard a causal relationship. Prospective studies are required to explore in detail the effects of diet and graft function in transplant patients. The effects of some cardiovascular drugs were assessed. As previously reported,32,33 -blockers were associated with a worse lipid profile. These actions were independent of the confounding effect of age, gender, and body mass index. The use of -blockers should be considered carefully in a post-transplant patient. Our data are in accordance with the well-known effects of prednisone and cyclosporine on the lipid profile.34-37 Prednisone dose was associated with higher HDL-C, HDL-2 cholesterol, and glucose levels. As described earlier, the use of cyclosporine in members of FCHL kindreds was associated with significantly worse lipid concentrations. The undesirable effects of cyclosporine on the lipid profile were shown in patients without FCHL. In conclusion, these data showed that the prevalence of post-transplant dyslipidemia is determined by genetic and environmental factors. The qualitative abnormalities observed on the lipid profile were similar to those reported in white subjects, but the magnitude of the abnormalities was remarkably lower in our Mexican subjects,
AGUILAR-SALINAS ET AL
reflecting the background lipoprotein pattern of the population. The presence of FCHL in the relatives is a strong marker for developing posttransplant hyperlipidemia, and its presence is associated with a more severe form of lipoprotein disturbances. This was especially true for cyclosporine-treated patients. Our data suggest that evaluation of the lipid profile of the closest relatives may be a valuable tool for the evaluation of the risk of having post-transplant dyslipidemia and perhaps for selecting an immunosuppressive agent. Other parameters, including the selection of immunosuppressive agents and -blockers and the severity of albuminuria, are important contributors to post-transplant lipid abnormalities. These results support the need for further prospective and cross-sectional studies designed to disclose the interaction of genetic factors with several acquired factors in posttransplant patients. REFERENCES 1. Kasiske B, Vazquez MA, Harmon W: Recommendations for outpatient surveillance of renal transplant recipients. J Am Soc Nephrol 11:S1-S86, 2000 2. Kasiske B: Risk factors for accelerated atherosclerosis in renal transplant recipients. Am J Med 84:985-992, 1988 3. Arkhus S, Dahl K, Wideroe TE: Cardiovascular morbidity and risk factors in renal transplant patients. Nephrol Dial Transplant 10:95-100, 1995 (suppl 1) 4. Wood D, De Backer G, Faergeman O: Prevention of coronary heart disease in clinical practice: Recommendations of the Second Joint Task Force of European and other Societies on Coronary Prevention. Atherosclerosis 140:199270, 1998 5. Kasiske B: Hyperlipidemia in patients with chronic renal disease. Am J Kidney Dis 32:S142-S156, 1998 6. Wheeler DC, Morgan R, Thomas DM: Factors influencing plasma lipid profiles including lipoprotein (a) concentrations in renal transplant recipients. Transpl Int 9:221-226, 1996 7. Mahley RW, Palaoglu KE, Atak Z: Turkish Heart Study: Lipids, lipoproteins and apolipoproteins. J Lipid Res 36:839-859, 1995 8. Li CS, Chau KF, Mak CK: Cadaveric kidney transplantation in a single center in Hong Kong. Transplant Proc 26:2012-2013, 1994 9. Dimeny E, Tufveson G, Lithell H: The influence of pretransplant lipoprotein abnormalities on the early results of renal transplantation. Eur J Clin Invest 23:572-579, 1993 10. Lipid and lipoprotein analysis, in Manual of Laboratory Operations: Lipid Research Clinics Program Report. NIH Publication 75-628. Washington, DC, U.S. Department of Health, Education and Welfare, 1974, pp 20-25 11. Siedel J, Schiefer S, Rosseneau M: Immunoturbidimetric method for routine determinations of apolipoproteins A-I, A-II and B in normo and hyperlipidemic sera: Compara-
DYSLIPIDEMIA POST-TRANSPLANTATION IN MEXICO
177
tive evaluation against immuno nephelometry. Clin Chem 34:1821-1825, 1998 12. Patsch W, Brown SA, Morriett JD, Gotto AM, Patsch JR: A dual-precipitation method evaluated for measurement of cholesterol in high-density lipoprotein subfractions HDL2 and HDL3 in human plasma. Clin Chem 35:265-270, 1989 13. Hixson JE, Vernier DT: Restriction isotyping of human apolipoprotein E by gene amplification and cleavage with HhaI. J Lipid Res 31:545-548, 1990 14. Aguilar-Salinas CA, Barrett H, Pulai J, et al: A familial combined hyperlipidemic kindred with impaired apolipoprotein B catabolism: Kinetics of apolipoprotein B during placebo and pravastatin therapy. Arterioscler Thromb Vasc Biol 17:72-82, 1997 15. Matthews D, Hosler JP, Rudenski AS: Homeostasis model assessment: Insulin resistance and B-cell function from fasting plasma glucose and insulin concentration in man. Diabetologia 28:412-419, 1985 16. Aguilar-Salinas CA, Olaiz G, Valles V, et al: High prevalence of low HDL cholesterol concentrations and mixed hyperlipidemia in a Mexican nationwide survey. J Lipid Res 42:1298-1307, 2001 17. Aguilar-Salinas CA, Talavera G, Guille´n LE, et al: The apolipoprotein E4 allele is not associated with an atherogenic lipid profile in a Native-American population following its traditional lifestyle. Atherosclerosis 142:409414, 1999 18. Valdez R, Stern M, Howard B: Apolipoprotein E polymorphism and insulin levels in a biethnic population. Diabetes Care 18:992-1000, 1995 19. Aarkhus S, Dahl K, Wideroe TE: Hyperlipidemia in renal transplant patients. J Intern Med 239:407-415, 1996 20. Chatterjee S, Chin HP, Azen S: Abnormal serum lipid patterns in primary renal allograft recipients. Surgery 82:655659, 1977 21. Sugahar S, Koyam I, Yoshikawa Y: Lipid and lipoprotein (a) in renal transplant recipients. Transplant Proc 26: 2080-2081, 1994 22. Dixon LB, Sudquist J, Winkleby M: Differences in energy, nutrient and food intakes in a US sample of Mexican American women and men: Findings from the Third National Health and Nutrition Examination Survey, 1988-1994. Am J Epidemiol 152:548-557, 2000 23. Loria C, Bush T, Carroll M: Macronutrient intakes among adult Hispanics: A comparison of Mexican Americans, Cuban Americans, and Mainland Puerto Ricans. Am J Public Health 85:684-689, 1995
24. Gaw A, Boerwinkle E, Cohen J. Comparative analysis of the apo (a) gene, apo (a) glycoprotein, and plasma concentration of Lp(a) in three ethnic groups. J Clin Invest 93:2526-2534, 1994 25. Fellstrom B: Impact and management of hyperlipidemia posttransplantation. Transplantation 70:SS51-SS57, 2000 26. Aguilar-Salinas CA, Barrett H, Schonfeld G: Metabolic modes of action of statins in the hyperlipoproteinemias. Atherosclerosis 141:203-207, 1998 27. Grundy SM, Chait A, Brunzell J: Familial combined hyperlipidemia workshop. Arteriosclerosis 7:203-207, 1987 28. Groth CG, Backman L, Morales JM: Sirolimus based therapy in human renal transplantation: Similar efficacy and different toxicity compared with cyclosporine. Transplantation 67:1036, 1999 29. Aguilar-Salinas CA, Vazquez-Chavez C, GamboaMarrufo R, et al: Prevalence of obesity, diabetes, hypertension and tobacco consumption in an urban adult Mexican population. Arch Med Res 32:446-453, 2001 30. Arroyo P, Loria A, Fernandez V, et al: Obesity in Hispanic Americans. Diabetes Care 14:691-694, 1991 31. Rao VK: Posttransplant medical complications. Surg Clin North Am 78:113-119, 1998 32. Sawicki PT, Berger M: Pharmacological treatment of diabetic patients with cardiovascular complications. J Intern Med 243:181-189, 1998 33. Troein M, Gardell B, Selander S: Guidelines and reported practice for the treatment of hypertension and hypercholesterolemia. J Intern Med 242:173-178, 1997 34. Hillbrands LB, Demacker PNM, Hoitsma AJ: The effects of cyclosporine and prednisone on serum lipids and (apo)lipoprotein levels in renal transplant recipients. J Am Soc Nephrol 5:2073-2076, 1995 35. Ponticelli C, Civati G, Tarantino A: Randomized study with cyclosporine in kidney transplantation: 10 year follow up. J Am Soc Nephrol 7:792-797, 1996 36. von Ahsen N, Helmhold M, Schutz E, Eisenhauer T, Armstrong VW, Oellerich M: Cyclosporin A trough levels correlate with serum lipoproteins and apolipoproteins: Implications for therapeutic drug monitoring of cyclosporin A. Ther Drug Monit 19:140-145, 1997 37. Fryer JP, Granger DK, Leventhal JR: Steroid related complications in the cyclosporine era. Clin Transplant 8:224229, 1994