This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Mykkänen, L. Articles by Austin, M. A. Search for Related Content PubMed PubMed Citation Articles by Mykkänen, L. Articles by Austin, M. A. Pubmed/NCBI databases Substance via MeSH Related Collections Pathophysiology Risk Factors Epidemiology Lipid and lipoprotein metabolism

(Arteriosclerosis, Thrombosis, and Vascular Biology. 1999;19:2742.)
© 1999 American Heart Association, Inc.

Atherosclerosis and Lipoproteins

LDL Size and Risk of Coronary Heart Disease in Elderly Men and Women

Leena Mykkänen; Johanna Kuusisto; Steven M. Haffner; Markku Laakso; Melissa A. Austin

From the Division of Clinical Epidemiology (L.M., S.M.H.), Department of Medicine, University of Texas Health Science Center at San Antonio; the Department of Medicine (L.M., J.K., M.L.), University of Kuopio, Kuopio, Finland; and the Department of Epidemiology (M.A.A.), School of Public Health and Community Medicine, University of Washington, Seattle.

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
Abstract—A predominance of small, dense, low density lipoprotein (LDL) particles has consistently been associated with coronary heart disease (CHD) in young and middle-aged subjects in cross-sectional studies. Recently, 3 prospective, case-control studies showed that decreased LDL size is a predictor of CHD in middle-aged subjects. However, it is not known whether decreased LDL size is mainly associated with premature CHD or whether it continues to play a role in CHD risk at older ages also. We performed a prospective, nested case-control study in 86 subjects (58 nondiabetic and 28 type 2 diabetic) aged 65 to 74 years who were free of myocardial infarction at baseline and who then had a myocardial infarction or CHD death during a 3.5-year follow-up (cases) and in 172 controls matched for sex and diabetes status but who remained free of CHD during follow-up. LDL particle size determined by gradient gel electrophoresis (268.2±0.9 versus 268.5±0.7 Å, P=0.782) and the proportion of subjects with LDL subclass phenotype B (20.9 versus 21.5, P=0.914) were similar among cases and controls. Furthermore, diastolic blood pressure, total cholesterol, high density lipoprotein cholesterol, triglycerides, apolipoprotein A₁, fasting glucose, fasting insulin, waist-to-hip ratio, and body mass index were not associated with CHD risk. However, smoking and increased systolic blood pressure, apolipoprotein B levels, and the total cholesterol–high density lipoprotein cholesterol ratio were significant predictors of CHD events both in univariate and multivariate analyses. Our findings indicate that LDL size is not a predictor of CHD events in elderly white subjects after controlling for diabetes status.

Key Words: lipoproteins • LDL • coronary heart disease • epidemiology • prospective study

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
An increased concentration of LDL cholesterol is a major risk factor for coronary heart disease (CHD).^{1 2} There is, however, considerable heterogeneity in the size, density, and composition of LDL particles.^{3 4} With the use of a diversity of laboratory techniques, variation in LDL particle characteristics, including decreased particle diameter and increased density, have consistently been associated with the risk of CHD in young and middle-aged subjects in cross-sectional studies.^{5 6 7 8 9}

Recently, 3 prospective, case-control studies showed that decreased LDL size, as characterized by gradient gel electrophoresis, is a predictor of CHD in middle-aged subjects.^{10 11 12} In the Stanford Five-City Project,¹⁰ LDL size was associated with CHD risk independently of triglyceride, HDL cholesterol, and non-HDL cholesterol levels, but not after adjusting for the ratio of total cholesterol to HDL cholesterol. In the Physicians’ Health Study,¹¹ LDL particle diameter was also associated with the risk of myocardial infarction (MI), but not after adjustment for triglyceride level. Results of the Québec Cardiovascular Study¹² were, however, not as straightforward as those in the other 2 studies. The mean LDL size was not different in men who developed CHD compared with men who were healthy, but the frequency distribution of LDL particle sizes was shifted toward smaller particles in men who developed CHD.¹² Men in the first tertile of LDL size distribution had an increased risk of CHD compared with those in the highest tertile, independently of triglyceride and HDL cholesterol levels.

Little is known about LDL size and CHD risk in the elderly. Previously, we have shown that a predominance of small LDL particles is a risk factor for the future development of type 2 diabetes in elderly subjects, which might imply that decreased LDL size contributes to the risk of CHD in prediabetics.¹³ It is not known whether decreased LDL size is mainly associated with premature CHD or whether it continues to play a role in CHD risk at older ages also. Therefore, we investigated the relationship of LDL size and LDL subclass phenotype to the risk of CHD over a 3.5-year follow-up in a nested case-control sample of Finnish white men and women 65 to 74 years old. Because our previous results showed that type 2 diabetes is a strong risk factor for CHD in elderly subjects¹⁴ and that decreased LDL particle size predicts type 2 diabetes in the elderly,¹³ we matched the cases and controls for diabetes status. Previous studies on the association of LDL size with CHD have not controlled for the effect of diabetes.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Study Population and Follow-Up
This prospective analysis is based on a nested case-control subsample from a well-characterized cohort of elderly white subjects from Kuopio, Finland.^{14 15 16} The original cohort, studied at baseline during 1986 through 1988, consisted of 1298 randomly selected men and women aged 65 to 74 years from the Kuopio population.^{14 15} The follow-up visit was conducted during 1990 and 1991. Among the 1190 subjects living at follow-up, the response rate was 88%, for a total of 1054 participants. The mean follow-up period for the participants was 3.5 years (42±4 months).¹⁴ For nonparticipants, the follow-up period was defined as the time between baseline study and June 30, 1991, and deaths and cardiovascular events during this period were recorded.¹⁴ The study was approved by the ethics committee of the Kuopio University Hospital, and all subjects gave informed consent.

Baseline Data Collection
Body mass index (BMI, kg/m²), waist-to-hip ratio, supine blood pressure, and smoking were recorded as previously described.¹⁴ Previously diagnosed diabetes was considered to be present if the diagnosis had been made earlier by a physician.¹⁵ The World Health Organization plasma glucose criteria for diabetes were used in the classification of subjects without previously diagnosed diabetes.¹⁷

Diagnosis of CHD at Baseline
A conventional 12-lead resting ECG was recorded, and the classification of all ECGs was made according to the Minnesota code.¹⁸ Previous MI was defined as present if a subject had had a possible or definite MI verified at the hospital by WHO criteria¹⁹ before the baseline examination or if there was a major Q-wave (Minnesota code 1.1 or 1.2) on ECG at baseline.

Diagnosis of Incident CHD Events
A conventional 12-lead ECG was taken at the follow-up study visit, and ECG findings were classified according to the Minnesota code.¹⁸ Medical records of participants who reported hospitalization because of symptoms suggestive of MI during the follow-up and medical records of all nonparticipants and those who died during the follow-up were reviewed. FINMONICA AMI Register Study Group criteria^{19 20} for verified and possible MI, based on chest pain symptoms, ECG changes, and enzyme determinations, were used in the ascertainment of a new MI. All death certificates of those who died during follow-up were reviewed. Hospital and autopsy records were used in the final classification of the causes of death. All deaths were coded according to the 9th revision of the International Classification of Diseases (ICD9).

A CHD death was defined as a death resulting from CHD (ICD9 codes 410 through 414). For participants, a new nonfatal MI during follow-up was defined as follows: (1) a definite or possible MI verified at the hospital by FINMONICA criteria or (2) a new major Q-QS change on the ECG (progression from no Minnesota Q-QS code to 1.1 or 1.2, or from 1.3 to 1.1). For nonparticipants, a new nonfatal MI was defined as a definite or possible MI verified at the hospital by FINMONICA criteria. All CHD events included CHD deaths and nonfatal MIs. If a subject had >1 CHD event during follow-up, only 1 CHD event was included in the statistical analyses.

Pairing Procedures
For this nested case-control design, only the 1144 subjects without a previous MI at baseline were eligible. A total of 58 nondiabetic and 28 type 2 diabetic subjects either had a nonfatal MI or died of CHD during follow-up. The 86 incident cases of CHD included 67 first events of MI and 19 CHD deaths. For this analysis, each subject with incident CHD (case subjects) was matched with 2 control subjects with respect to sex and diabetes.

Laboratory Methods at Baseline
Blood samples were taken between 7:30 and 9:30 AM after a 12-hour fast. All subjects, except for those receiving insulin, underwent a 2-hour (75-g) oral glucose tolerance test. Plasma glucose was determined by the glucose oxidase method (Glucose Auto & Stat HGA-1120 analyzer, Daiichi). Plasma insulin was determined by a double-antibody, solid-phase radioimmunoassay (Phadeseph insulin RIA 100, Pharmacia Diagnostics AB), which also detects proinsulin and proinsulin conversion products with considerable sensitivity. The interassay coefficient of variation for the insulin assay was <8.7%.

Serum lipid and lipoprotein levels were determined from fresh samples drawn after a 12-hour overnight fast. Serum HDL cholesterol was determined after precipitation of LDL and VLDL with dextran sulfate and MgCl₂.²¹ Commercial enzymatic methods were used in the determination of cholesterol and triglycerides as described before.¹⁴ Serum apolipoprotein (apo) B and apoA₁ were determined from samples stored at -70°C by a commercial immunochemical method (Kone Diagnostics) that was based on the measurement of immunoprecipitation at 340 nm.²² Apolipoprotein measurements were standardized with a calibrator. The interassay coefficient of variation was <3.8% for total cholesterol, <4.9% for HDL cholesterol, <2.5% for triglyceride, <5.2% for apoB, and <4.7% for apoA₁.

Characterization of LDL Subclasses
Serum samples used for characterizing LDL subclasses were also collected at baseline after a 12-hour fast and stored at -70°C for an average of 6.4 years (range 67 to 92 months). They were transported to Seattle on dry ice and were never thawed until gradient gel electrophoresis procedures were performed. Several studies have suggested that freezing samples does not alter their LDL phenotype.^{23 24 25} LDL subclasses were characterized by use of 2% to 14% polyacrylamide gels and electrophoresis procedures in Dr Austin’s laboratory as previously described.¹³ Two gels were electrophoresed for each case-control triplet: 1 using whole serum and another using isolated LDL as previously described.¹³ A set of high-molecular-weight standards was also run on each gel and was used to construct a quadratic calibration curve for estimating the diameter of LDL subclasses as previously described.²⁶ In addition, 2 quality control samples with well-characterized LDL subclass phenotypes were run on each gel. Each gel lane was scanned with a QuickScan densitometer (Helena Laboratories) interfaced with a personal computer via an analog-to-digital converter and the Hoefer GS365 data system.

Gradient gel electrophoresis data were processed with software developed in Dr Austin’s laboratory (Gel Scan Calibration Program) as previously described.¹³ The estimated diameter for the major peak in each scan was called the peak particle diameter and was used as a continuous variable in statistical analyses. Then 3 evaluators independently classified the LDL subclass phenotype as phenotype A, B, or I (intermediate) as previously described.¹³ For statistical analyses requiring a dichotomous variable, phenotype I subjects were grouped with those with phenotype B.⁵

The results presented here are based on analyses of LDL size determined from whole-serum samples. Results were essentially similar when LDL size was determined from isolated LDL samples (data not shown). Correlation between LDL particle sizes determined from whole serum and isolated LDL samples was 0.91 (P<0.001).

Statistical Analysis
Data analyses were performed with SPSS/PC+ (SPSS Inc) and SAS (SAS Institute) statistical software. Comparisons of mean values between groups at baseline were made with Student’s t test. Analyses of categorical variables were performed with a ² test. Conditional logistic regression analysis and calculation of relative risk were used to assess the association between risk factors at baseline and incident CHD events. The risk of CHD associated with each continuous variable was standardized as the relative risk of CHD for a 1-SD increase in the concentration of the variable. Interaction terms involving smoking, systolic blood pressure, and apoB were also evaluated in the multivariate conditional logistic regression analysis, but no significant (P>0.280) interactions were detected. Because the distributions of insulin and triglyceride were skewed, the natural logarithms of these variables were used in all calculations. However, these variables are reported in their natural units in the tables.

Power Calculations
We calculated the power of the present study to detect differences in LDL particle size and proportion of LDL phenotype B between controls and CHD cases by using a 2-sided test for a 2-sample normal distribution with equal variances²⁷ and a 2-sided test for a 2-sample binomial distribution,²⁸ respectively. The minimum detectable difference in LDL particle size was 3.4 Å (268.5 versus 265.1 Å) for 172 control subjects and 86 CHD cases with 80% power (=0.05). The minimum detectable difference in the proportion of subjects with LDL phenotype B was 16.8 percentage points (21.5% versus 38.3%) with 80% power (=0.05).

	Results

Top Abstract Introduction Methods Results Discussion References

 
Baseline characteristics of the study population are shown in Table 1. Mean age of the subjects was 69 years. Nondiabetic and diabetic subjects who had a CHD event during follow-up did not differ from the corresponding control subjects in regard to age, BMI, waist-to-hip ratio, waist circumference, fasting plasma glucose levels, or fasting plasma insulin levels at baseline. Furthermore, the proportion of subjects with impaired glucose tolerance (fasting venous plasma glucose <7.8 mmol/L and 2-hour venous plasma glucose 7.8 to 11.1 mmol/L) was similar among nondiabetic cases and controls. Duration of diabetes was similar in diabetic subjects who had a CHD event compared with diabetic control subjects. Thus, cases and controls of this study were very well matched for glucose tolerance status.

View this table: [in this window] [in a new window]   Table 1. Baseline Characteristics of the Study Population

Table 2 shows baseline LDL particle size and the proportion of subjects with LDL subclass phenotype B in nondiabetic and diabetic CHD cases and controls. There were no significant differences in LDL particle size or the proportion of subjects with phenotype B between nondiabetic or diabetic CHD cases and controls. Diabetic subjects had a significantly smaller LDL particle size (263.6 versus 270.7 Å, P<0.001) and a higher prevalence of LDL phenotype B (39.3% versus 12.6%, P<0.001) than did nondiabetic subjects. Even after adjustment for sex and triglyceride concentration (266.1 versus 269.5 Å, P<0.001) or for sex and triglyceride and HDL cholesterol concentrations (266.4 versus 269.4 Å, P=0.001), diabetic subjects continued to have a smaller LDL particle size compared with nondiabetic subjects. Further comparisons between cases and controls were done after pooling nondiabetic and diabetic subjects.

View this table: [in this window] [in a new window]   Table 2. Baseline LDL Particle Size and Proportion of Subjects With LDL Subclass Phenotype B in Nondiabetic and Diabetic CHD Cases and Controls

Table 3 shows baseline cardiovascular risk factors in all cases and controls. Smoking was twice as common among subjects who had a subsequent CHD event compared with controls. Cases also had a higher systolic blood pressure and apoB levels at baseline compared with controls. However, subjects who had a CHD event did not differ from controls in regard to baseline levels of diastolic blood pressure, total cholesterol, HDL cholesterol, total cholesterol–HDL cholesterol ratio, triglycerides, or apoA₁. Altogether, 34% of the cases and controls were taking antihypertensive medication. LDL particle size and the proportion of subjects with LDL subclass phenotype B were similar among cases and controls. As shown in the Figure, the frequency distribution of LDL particle size was also similar in CHD cases and controls (² P=0.607).

View this table: [in this window] [in a new window]   Table 3. Baseline Cardiovascular Risk Factors

View larger version (23K): [in this window] [in a new window]   Figure 1. Frequency distributions of LDL particle size in CHD cases (solid bars) and controls (open bars). The frequency distributions were not different between the 2 groups (2 P=0.607).

The magnitude of the associations between risk factors and CHD was estimated by calculating univariate relative risks by using conditional logistic regression analysis (Table 4). Statistically significant risk factors for CHD were smoking and increased levels of systolic blood pressure and apoB. Smoking was related to a 2.6-fold risk of CHD. A 1-SD increase in baseline systolic blood pressure (24 mm Hg) or apoB level (0.28 g/L) was associated with a 40% increased risk of CHD. Furthermore, a 1-SD increase in the total cholesterol–HDL cholesterol ratio (1.77 units) was related to a 28% increase in CHD risk (P=0.068), and a 1-SD increase in HDL cholesterol level (0.31 mmol/L) was related to a 20% decrease in CHD risk (P=0.091), but these associations did not reach statistical significance. An increase in BMI or triglyceride level (logarithmically transformed) was also related to a 20% increase in relative risk for CHD, although neither of these associations was statistically significant. Age, waist-to-hip ratio, waist circumference, fasting glucose, fasting insulin, diastolic blood pressure, total cholesterol, apoA₁, LDL size, and LDL subclass phenotype were not related to CHD risk.

View this table: [in this window] [in a new window]   Table 4. Standardized1 Univariate Relative Risks for Coronary Heart Disease Events by Conditional Logistic Regression Analyses

We also investigated the relationship between LDL subclass phenotypes and cardiovascular risk factors after pooling cases and controls (Table 5). Subjects with LDL subclass phenotype B had a significantly higher waist-to-hip ratio, lower HDL cholesterol levels, a higher total cholesterol–HDL cholesterol ratio, and higher triglyceride, apoB, fasting glucose, and fasting insulin levels compared with subjects with phenotype A. Moreover, the proportion of subjects with type 2 diabetes was 2.4-fold greater in subjects with phenotype B compared with subjects with phenotype A. Thus, LDL subclass phenotype B was associated with multiple adverse changes in cardiovascular risk factors in these elderly subjects, although it was not related to the risk of CHD in the case-control comparison.

View this table: [in this window] [in a new window]   Table 5. Relationship of LDL Subclass Phenotype to Cardiovascular Risk Factors

Finally, we performed a stepwise conditional logistic regression analysis to investigate which risk factors were independently associated with CHD risk (Table 6). Independent variables tested in the model included age, current smoking, systolic blood pressure, total cholesterol–HDL cholesterol ratio, log triglyceride, apoB, and LDL size. In this multivariate analysis, smoking, total cholesterol–HDL cholesterol ratio, and systolic blood pressure were significant independent risk factors for CHD. When total cholesterol and HDL cholesterol concentrations were tested as independent variables instead of their ratio in a stepwise conditional logistic regression analysis, smoking, systolic blood pressure, and apoB were significant independent risk factors for CHD (data not shown).

View this table: [in this window] [in a new window]   Table 6. Significant Predictors of CHD Events in Stepwise Conditional Logistic Regression

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
The present study shows that LDL size is not a significant predictor of CHD in elderly white subjects after controlling for diabetes status. This finding is in contrast to previous studies in middle-aged subjects.^{10 11 12} This lack of association was confirmed by using both LDL particle size as a continuous variable and the dichotomous LDL subclass phenotype.

What could be the explanation for these apparently conflicting results? First, survival bias could explain the lack of association between LDL size and CHD in the present study. Because our study population was 65 to 74 years old, they represented survivors of their cohort. If small LDL size is mainly related to premature CHD death, then the association between LDL size and CHD would be expected to be weaker in this age group. Furthermore, the proportion of subjects with LDL phenotype B in our study population was relatively low compared with the Québec Cardiovascular Study¹² and the Physician Health Study¹¹ populations. This may support the survival bias hypothesis. On the other hand, the relatively low prevalence of LDL phenotype B in our study population also reflects the LDL phenotype distribution in the Finnish population. We have previously shown²⁹ that the mean LDL size in healthy, middle-aged, Finnish men is relatively large (26.9 nm). However, Finnish subjects are not unique in this sense, since the population in the Stanford Five-City Project¹⁰ also had an equally large mean LDL size. Ethnic and lifestyle factors might contribute to the differences in LDL sizes as well as to the differences in the risk of CHD associated with decreased LDL size in various populations. The relatively small proportion of subjects with LDL phenotype B tends to decrease the power of the present study to detect associations with this phenotype when a dichotomous approach is used. This is less of an issue when LDL size is represented as a continuous variable.

Second, diabetes status is a major confounding factor in the association between LDL size and CHD. Type 2 diabetes is associated both with small, dense LDL^{30 31 32 33} and with an increased risk of CHD events³⁴ in middle-aged subjects. The association between diabetes and small LDL size has consistently been found to be independent of triglyceride concentration.^{31 32 33} Because our previous results showed that type 2 diabetes is a strong risk factor for CHD in elderly subjects also¹⁴ and that decreased LDL particle size predicts type 2 diabetes in the elderly,¹³ we matched the cases and controls for diabetes in the present study. Furthermore, the proportion of subjects with impaired glucose tolerance was similar among nondiabetic cases and controls. Thus, we fully controlled for the effect of glucose tolerance status in the present study. In contrast, in 2 of the previous studies,^{11 12} the prevalence of type 2 diabetes was markedly higher among CHD cases: 15% versus 1% in the Québec Cardiovascular Study¹² and 6% versus 4% in the Physicians’ Health Study.¹¹ The report of the Stanford Five-City Project¹⁰ did not have data on diabetes status among cases and controls. The report of the Québec Cardiovascular Study¹² revealed that exclusion of diabetic individuals did not affect the relationship between dense LDL phenotype and the risk of CHD, although these results were not presented in detail. However, in the Québec Cardiovascular Study¹² as in the Physician Health Study,¹¹ the diagnosis of diabetes was based on medical history and was not verified by fasting plasma glucose measurements or an oral glucose tolerance test. Thus, the prevalence of diabetes was probably an underestimation of the true rate. Therefore, it is possible that in the previous studies, the relation between LDL size and CHD was partly due to type 2 diabetes.

This hypothesis is supported by the finding that the proportion of subjects with type 2 diabetes was 2.4-fold greater among subjects with LDL phenotype B compared with those with LDL phenotype A in the present study. Moreover, diabetes was associated with smaller LDL size, independent of triglyceride and HDL cholesterol concentrations. Even though LDL size was not related to CHD risk in the case-control comparison here, subjects with LDL subclass phenotype B had more adverse cardiovascular risk factors compared with subjects with phenotype A. Thus, small, dense LDL is not a benign lipoprotein phenotype, even in older age.

This case-control study showed that smoking and increased levels of systolic blood pressure and the total cholesterol–HDL cholesterol ratio were independent risk factors for CHD among elderly subjects without a previous MI. These results are in line with our previous findings in the whole study cohort,¹⁴ with the exception that in the whole study cohort the lipid risk factor was HDL cholesterol. Type 2 diabetes, male sex, and previous MI were also independent predictors of CHD in the whole study cohort.¹⁴ In the present study, however, cases and controls were matched for type 2 diabetes and sex. Previous studies have shown that elevated total and LDL cholesterol concentrations continue to be associated with CHD risk in the elderly.^{35 36 37 38} However, HDL cholesterol and the total cholesterol–HDL cholesterol ratio have been stronger risk factors for CHD in older age than is total cholesterol.^{35 36 39} Blood pressure, particularly systolic blood pressure, has been shown to be a potent risk factor for CHD in older age.⁴⁰ Although the role of smoking in CHD risk in the elderly has been controversial,³⁶ recent studies have shown that current smoking is an independent risk factor for cardiovascular mortality among subjects aged over 65 years.^{41 42} Thus, results of the present study and previous studies indicate that the classic cardiovascular risk factors are still important in the elderly.

A limitation of the present study is that there were only 86 subjects who developed CHD during follow-up. This is partly due to the relatively short follow-up of 3.5 years. In the previous nested case-control studies on LDL size and CHD, the follow-up period was 5 to 7 years, and the number of cases was 103 to 266.^{10 11 12} Thus, our study has lower power compared with the previous studies. Our study had 80% power (=0.05) to detect a minimum of a 3.4-Å difference in mean LDL particle size between the controls and CHD cases. In the Physicians’ Health Study,¹¹ the difference in the mean LDL size between controls and cases was 3 Å and in the Stanford Five-City Project,¹⁰ the difference was 5.1 Å (5.2 Å in men and 4.8 Å in women). Thus, our study had enough power to detect differences in LDL particle size of a similar magnitude to those seen in previous studies in middle-aged subjects. Our power to detect differences in the proportion of subjects with LDL phenotype B (dichotomous approach) was lower. We had enough power to detect a 16.8 percentage-point difference in the proportion of LDL phenotype B. This is not too different from the previously found 14 to 15.5 percentage-point difference in the proportion of subjects with LDL phenotype B between controls and CHD cases in middle-aged subjects.^{10 12} If LDL size were truly associated with CHD risk in elderly white subjects matched for sex and diabetes in the present study, one would have expected to see at least some difference in LDL size between cases and controls. However, LDL size and the proportion of subjects with LDL phenotype B were identical among cases and controls. Furthermore, the frequency distribution of LDL particle size was similar in CHD cases and controls. In contrast, smoking, systolic blood pressure, apoB level, and the total cholesterol–HDL cholesterol ratio were significantly associated with CHD risk both in univariate and multivariate analyses. In the Stanford Five-City Project, which is the only previous study in which cases and controls were not matched for smoking, LDL size and smoking were equally strong risk factors for CHD.¹⁰ Therefore, we believe that our results are not due to low power of the study but can be explained by examining this relationship in the elderly and by controlling for the effect of diabetes.

In conclusion, LDL size was found not to be a predictor of CHD in elderly white subjects matched for sex and type 2 diabetes. In this cohort, smoking and elevated systolic blood pressure, apoB levels, and the total cholesterol–HDL cholesterol ratio were independent predictors of CHD events. Our results support the hypothesis that decreased LDL size is mainly associated with premature CHD. Furthermore, it is possible that reduced LDL size contributes to the risk of CHD in elderly subjects with diabetes. Additional data from larger, population-based studies are needed to confirm that LDL particle size is not a major risk factor in the elderly.

	Acknowledgments

 
This work was performed during Dr Austin’s tenure as an Established Investigator of the American Heart Association. The authors would like to thank Cherry Tamblyn and Carries Nelson for LDL subclass determinations and Tom Fraser and Andy Lovie for data management and programming assistance.

	Footnotes

 
Reprint requests and correspondence to Leena Mykkänen, MD, Department of Medicine, University of Kuopio, PO Box 1777, FIN-70211 Kuopio, Finland.

Received January 13, 1999; accepted April 14, 1999.

	References

Top Abstract Introduction Methods Results Discussion References

 

1. Kannel W, Castelli WP, Gordon T, McNamara PM. Serum cholesterol, lipoproteins and the risk of coronary heart disease: the Framingham Study. Ann Intern Med. 1971;74:1–12.
2. Stamler J, Wentworth D, Neaton JD, for the MRFIT Research Group. Is relationship between serum cholesterol and risk of premature death from coronary heart disease continuous and graded? findings in 356 222 primary screenees of the Multiple Risk Factor Intervention Trial (MRFIT). JAMA. 1986;256:2823–2828.[Abstract]
3. Krauss RM, Burke DJ. Identification of multiple subclasses of plasma low density lipoproteins in normal humans. J Lipid Res. 1981;23:236–244.
4. Swinkels DW, Hak-Lemmers HLM, Demacker PNM. Single spin density gradient ultracentrifugation method for the detection and isolation of light and heavy low density lipoprotein subfractions. J Lipid Res. 1987;28:1233–1239.[Abstract]
5. Austin MA, Breslow JL, Hennekens CH, Buring JL, Willett WC, Krauss RM. Low density lipoprotein subclass pattern and risk of myocardial infarction. JAMA. 1988;260:1917–1921.[Abstract]
6. Crouse JR, Parks JS, Schey HM, Kahl FT. Studies of low density lipoprotein molecular weight in human beings with coronary artery disease. J Lipid Res. 1985;26:566–574.[Abstract]
7. Campos H, Genest JJ, Blijlevens E, McNamara JR, Jenner JL, Ordovas JM, Wilson PWF, Schaefer EJ. Low density lipoprotein particle size and coronary artery disease. Arterioscler Thromb. 1992;12:187–195.[Abstract/Free Full Text]
8. Tornvall P, Karpe F, Carlson LA, Hamsten A. Relationships of low density lipoprotein subfractions to angiographically defined coronary heart disease in young survivors of myocardial infarction. Atherosclerosis. 1991;90:67–80.[Medline] [Order article via Infotrieve]
9. Coresh J, Kwiterovich PO Jr, Smith HH, Bachorik PS. Association of plasma triglyceride concentration and LDL particle diameter, density, and chemical composition with premature coronary artery disease in men and women. J Lipid Res. 1993;34:1687–1697.[Abstract]
10. 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;276:875–881.[Abstract]
11. Stampfer MJ, Krauss RM, Ma J, Blanche PJ, Holl LG, Sacks FM, Hennekens CH. A prospective study of triglyceride level, low-density lipoprotein particle diameter, and risk of myocardial infarction. JAMA. 1996;276:882–888.[Abstract]
12. Lamarche B, Tchernof A, Moorjani S, Cantin B, Dagenais GR, Lupien PJ, Després J-P. 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;95:69–75.[Abstract/Free Full Text]
13. Austin MA, Mykkänen L, Kuusisto J, Edwards KL, Nelson C, Haffner SM, Pyörälä K, Laakso M. Prospective study of small LDLs as a risk factor for non-insulin dependent diabetes mellitus in elderly men and women. Circulation. 1995;92:1770–1778.[Abstract/Free Full Text]
14. Kuusisto J, Mykkänen L, Pyörälä K, Laakso M. NIDDM and its metabolic control predict coronary heart disease in elderly subjects. Diabetes. 1994;43:960–967.[Abstract]
15. Mykkänen L, Laakso M, Uusitupa M, Pyörälä K. Prevalence of diabetes and impaired glucose tolerance in elderly subjects and their association with obesity and family history of diabetes. Diabetes Care. 1990;13:1099–1105.[Abstract]
16. Mykkänen L, Laakso M, Penttilä I, Pyörälä K. Asymptomatic hyperglycemia and cardiovascular risk factors in the elderly. Atherosclerosis. 1991;88:153–161.[Medline] [Order article via Infotrieve]
17. World Health Organization. Diabetes Mellitus: Report of a WHO Study Group. Geneva, Switzerland: World Health Organization; 1985. Technical report series No. 727.
18. Prineas RJ, Crow RS, Blackburn H. The Minnesota Code Manual of Electrocardiographic Findings: Standards and Procedures for Measurement and Classification. Boston, Mass: Wright; 1982.
19. World Health Organization. MONICA Manual: CVD/MNC. Geneva, Switzerland: World Health Organization; 1990.
20. Tuomilehto J, Arstila M, Kaarsalo E, Kankaanpää J, Ketonen M, Kuulasmaa K, Lehto S, Miettinen H, Mustaniemi H, Palomäki P, Puska P, Pyörälä K, Salomaa V, Torppa J, Vuorenmaa T. Acute myocardial infarction (AMI) in Finland: baseline data from the FINMONICA AMI Register in 1983–85. Eur Heart J. 1992;13:577–587.[Abstract/Free Full Text]
21. Penttilä I, Voutilainen E, Laitinen P, Juutilainen P. Comparison of different analytic and precipitation methods for direct estimation of serum high-density lipoprotein cholesterol. Scand J Clin Lab Invest. 1981;41:353–360.[Medline] [Order article via Infotrieve]
22. Fruchart J-C, Kora I, Cachera C, Clavey V, Duthilleul P, Moschetto Y. Simultaneous measurements of plasma apolipoproteins A-1 and B by electroimmunoassay. Clin Chem. 1982;28:59–62.[Abstract/Free Full Text]
23. McNamara JR, Campos H, Ordovas JM, Wilson PWF, Schaefer EJ. Gradient gel electrophoretic analysis of low density lipoproteins. Am Biotech Lab. 1988;6:28–33.
24. Dormans TPJ, Swinkels DW, De Graaf J, Hendriks JCM, Stalenhoef AFP, Demacker PNM. Single-spin density-gradient ultracentrifugation vs. gradient gel electrophoresis: two methods for detecting low-density-lipoprotein heterogeneity compared. Clin Chem. 1991;37:853–858.[Abstract/Free Full Text]
25. Hokanson JE, Austin MA, Zambon A, Brunzell JD. Plasma triglyceride and LDL heterogeneity in familial combined hyperlipidemia. Arterioscler Thromb. 1993;13:427–434.[Abstract/Free Full Text]
26. Krauss RM, Burke DJ. Identification of multiple subclasses of plasma low-density lipoprotein in normal humans. J Lipid Res. 1982;23:97–104.[Abstract]
27. Brown BW, Brauner C, Chan A, Gutierrez D, Herson J, Lovato J, Polsley J, Russell K. STPLAN. version 4.1. Houston, Tex: Department of Biomathematics, University of Texas, MD Anderson Cancer Center; 1996:101–106.
28. Brown BW, Brauner C, Chan A, Gutierrez D, Herson J, Lovato J, Polsley J, Russell K. STPLAN. version 4.1. Houston, Tex: Department of Biomathematics, University of Texas, MD Anderson Cancer Center; 1996:31–34.
29. Mykkänen L, Haffner SM, Rainwater DL, Karhapää P, Miettinen H, Laakso M. Relationship of LDL size to insulin sensitivity in normoglycemic men. Arterioscler Thromb Vasc Biol. 1997;17:1447–1453.[Abstract/Free Full Text]
30. Selby JV, Austin MA, Newman B, Zhang D, Quesenberry CP Jr, Mayer EJ, Kraus RM. LDL subclass phenotypes and the insulin resistance syndrome in women. Circulation. 1993;88:381–387.[Abstract/Free Full Text]
31. Feingold KR, Grunfeld C, Pang M, Coerrler W, Krauss RM. LDL subclass phenotypes and triglyceride in non–insulin-dependent diabetes. Arterioscler Thromb. 1992;12:1496–1502.[Abstract/Free Full Text]
32. Haffner SM, Mykkänen L, Stern MP, Paidi M, Howard BV. Greater effect of diabetes on LDL size in women than in men. Diabetes Care. 1994;17:1164–1171.[Abstract]
33. Gray RS, Robbins DC, Wang W, Yeh JL, Fabsitz RR, Cowan LD, Welty TK, Lee ET, Krauss RM, Howard BV. Relation of LDL size to the insulin resistance syndrome and coronary heart disease in American Indians: the Strong Heart Study. Arterioscler Thromb Vasc Biol. 1997;17:2713–2720.[Abstract/Free Full Text]
34. Laakso M, Lehto S. Epidemiology of macrovascular disease in diabetes. Diabetes Rev. 1997;5:294–315.
35. Gordon T, Castelli WP, Hjortland MC, Kannel WB, Dawber TR. High density lipoprotein as a protective factor against coronary heart disease: the Framingham Study. Am J Med. 1977;62:707–714.[Medline] [Order article via Infotrieve]
36. Castelli WP, Wilson PWF, Levy D, Anderson K. Cardiovascular risk factors in the elderly. Am J Cardiol.. 1989;63:12H–19H.[Medline] [Order article via Infotrieve]
37. Benfante R, Reed D. Is elevated serum cholesterol level a risk factor for coronary heart disease in the elderly? JAMA.. 1990;263:393–396.[Abstract]
38. Zimetbaum P, Frishman WH, Ooi WL, Derman MP, Aronson M, Gidez LI, Eder HA. Plasma lipids and lipoproteins and the incidence of cardiovascular disease in the very elderly. Arterioscler Thromb. 1992;12:416–423.[Abstract/Free Full Text]
39. Corti M-C, Guralnik JM, Salive ME, Harris T, Field TS, Wallace RB, Berkman LF, Seeman TE, Glynn RJ, Hennekens CH, Havlik RJ. HDL cholesterol predicts coronary heart disease mortality in older persons. JAMA. 1995;274:539–544.[Abstract]
40. Vokonas PS, Kannel WB, Cupples LA. Epidemiology and risk of hypertension in the elderly: the Framingham Study. J Hypertens. 1988;6(suppl 1):S3–S9.
41. Jajich CL, Ostfeld AM, Freeman DH. Smoking and coronary heart disease mortality in the elderly. JAMA. 1984;252:2831–2834.[Abstract]
42. LaCroix AZ, Lang J, Scherr P, Wallace RB, Cornoni-Huntley J, Berkman L, Curb JD, Evans D, Hennekens CH. Smoking and mortality among older men and women in three communities. N Engl J Med. 1991;324:1619–1625.[Abstract]

This article has been cited by other articles:

				K. J. Mukamal, R. H. Mackey, L. H. Kuller, R. P. Tracy, R. A. Kronmal, M. A. Mittleman, and D. S. Siscovick Alcohol Consumption and Lipoprotein Subclasses in Older Adults J. Clin. Endocrinol. Metab., July 1, 2007; 92(7): 2559 - 2566. [Abstract] [Full Text] [PDF]

				H.-h. Chen, B. D. Hosken, M. Huang, J. W. Gaubatz, C. L. Myers, R. D. Macfarlane, H. J. Pownall, and C.-y. Yang Electronegative LDLs from familial hypercholesterolemic patients are physicochemically heterogeneous but uniformly proapoptotic J. Lipid Res., January 1, 2007; 48(1): 177 - 184. [Abstract] [Full Text] [PDF]

				M. Rizzo and K. Berneis Low-density lipoprotein size and cardiovascular risk assessment QJM, January 1, 2006; 99(1): 1 - 14. [Abstract] [Full Text] [PDF]

				V. Bittner Perspectives on Dyslipidemia and Coronary Heart Disease in Women J. Am. Coll. Cardiol., November 1, 2005; 46(9): 1628 - 1635. [Abstract] [Full Text] [PDF]

				R. M. Krauss Dietary and Genetic Probes of Atherogenic Dyslipidemia Arterioscler. Thromb. Vasc. Biol., November 1, 2005; 25(11): 2265 - 2272. [Abstract] [Full Text] [PDF]

				E. M. Stuveling, S. J. L. Bakker, H. L. Hillege, P. E. de Jong, R. O. B. Gans, and D. de Zeeuw Biochemical risk markers: a novel area for better prediction of renal risk? Nephrol. Dial. Transplant., March 1, 2005; 20(3): 497 - 508. [Full Text] [PDF]

				A. C. St-Pierre, B. Cantin, G. R. Dagenais, P. Mauriege, P.-M. Bernard, J.-P. Despres, and B. Lamarche Low-Density Lipoprotein Subfractions and the Long-Term Risk of Ischemic Heart Disease in Men: 13-Year Follow-Up Data From the Quebec Cardiovascular Study Arterioscler. Thromb. Vasc. Biol., March 1, 2005; 25(3): 553 - 559. [Abstract] [Full Text] [PDF]

				G. Boden and M. Laakso Lipids and Glucose in Type 2 Diabetes: What is the cause and effect? Diabetes Care, September 1, 2004; 27(9): 2253 - 2259. [Full Text] [PDF]

				Y. Bosse, L. Perusse, and M.-C. Vohl Genetics of LDL particle heterogeneity: from genetic epidemiology to DNA-based variations J. Lipid Res., June 1, 2004; 45(6): 1008 - 1026. [Abstract] [Full Text] [PDF]

				M. A. Austin, K. L. Edwards, S. A. Monks, K. M. Koprowicz, J. D. Brunzell, A. G. Motulsky, M. C. Mahaney, and J. E. Hixson Genome-wide scan for quantitative trait loci influencing LDL size and plasma triglyceride in familial hypertriglyceridemia J. Lipid Res., November 1, 2003; 44(11): 2161 - 2168. [Abstract] [Full Text] [PDF]

				F. M. Sacks and H. Campos Low-Density Lipoprotein Size and Cardiovascular Disease: A Reappraisal J. Clin. Endocrinol. Metab., October 1, 2003; 88(10): 4525 - 4532. [Full Text] [PDF]

				References Circulation, December 17, 2002; 106(25): 3373 - 3421. [Full Text]

				L. Kuller, A. Arnold, R. Tracy, J. Otvos, G. Burke, B. Psaty, D. Siscovick, D. S. Freedman, and R. Kronmal Nuclear Magnetic Resonance Spectroscopy of Lipoproteins and Risk of Coronary Heart Disease in the Cardiovascular Health Study Arterioscler. Thromb. Vasc. Biol., July 1, 2002; 22(7): 1175 - 1180. [Abstract] [Full Text] [PDF]

				J. Vakkilainen, M. Jauhiainen, K. Ylitalo, I. O. Nuotio, J. S. A. Viikari, C. Ehnholm, and M.-R. Taskinen LDL particle size in familial combined hyperlipidemia: effects of serum lipids, lipoprotein-modifying enzymes, and lipid transfer proteins J. Lipid Res., April 1, 2002; 43(4): 598 - 603. [Abstract] [Full Text] [PDF]

				M. Kratz, E. Gulbahce, A. von Eckardstein, P. Cullen, A. Cignarella, G. Assmann, and U. Wahrburg Dietary Mono- and Polyunsaturated Fatty Acids Similarly Affect LDL Size in Healthy Men and Women J. Nutr., April 1, 2002; 132(4): 715 - 718. [Abstract] [Full Text] [PDF]

				S. Miremadi, A. Sniderman, and J. Frohlich Can Measurement of Serum Apolipoprotein B Replace the Lipid Profile Monitoring of Patients with Lipoprotein Disorders? Clin. Chem., March 1, 2002; 48(3): 484 - 488. [Abstract] [Full Text] [PDF]

				A. C. St-Pierre, I. L. Ruel, B. Cantin, G. R. Dagenais, P.-M. Bernard, J.-P. Despres, and B. Lamarche Comparison of Various Electrophoretic Characteristics of LDL Particles and Their Relationship to the Risk of Ischemic Heart Disease Circulation, November 6, 2001; 104(19): 2295 - 2299. [Abstract] [Full Text] [PDF]

				H. Campos, L. A. Moye, S. P. Glasser, M. J. Stampfer, and F. M. Sacks Low-Density Lipoprotein Size, Pravastatin Treatment, and Coronary Events JAMA, September 26, 2001; 286(12): 1468 - 1474. [Abstract] [Full Text] [PDF]

				A. D. Sniderman, T. Scantlebury, and K. Cianflone Hypertriglyceridemic HyperapoB: The Unappreciated Atherogenic Dyslipoproteinemia in Type 2 Diabetes Mellitus Ann Intern Med, September 18, 2001; 135(6): 447 - 459. [Abstract] [Full Text] [PDF]