Electronic health record reveals community-level cardiometabolic health benefits associated with 10 years of community-based participatory research

BACKGROUND

While a major goal of community-based participatory research (CBPR) is to improve community health; it is unclear how to measure longstanding success of CBPR.

OBJECTIVE

We sought to determine the impact of ongoing CBPR on cardiometabolic health of participating communities, including in people not directly participating in research.

METHODS

We used linear mixed-effects modelling with electronic medical records from 2002 to 2012 from the Yukon-Kuskokwim Health Corporation, which provides health care to all Alaska Native people in southwestern Alaska, to compare rates of change in cardiometabolic risk factors between communities that did and did not participate in ongoing CBPR beginning in 2003.

RESULTS

We analysed 1,262,035 medical records from 12,402 individuals from 10 study and 38 control communities. Blood pressure declined faster in study than in control communities: systolic blood pressure (0.04 mmHg/year; 95% confidence interval [CI]: 0.01, 0.08); diastolic blood pressure (DBP) (0.07 mmHg/year; 95% CI: 0.04, 0.09). Body mass index increased 0.04 units/year faster in study communities than in control communities (95% CI: 0.03, 0.05). More study visits were associated with faster reduction of DBP and triglyceride levels in study communities.

CONCLUSIONS

Ongoing CBPR may improve overall cardiometabolic health in communities, perhaps by increasing engagement in health and advocacy.

Associations between diet and cardiometabolic risk among Yup'ik Alaska Native people using food frequency questionnaire dietary patterns

BACKGROUND AND AIMS

In previous analyses, we identified three dietary patterns from food frequency questionnaire data among a sample of Yup'ik Alaska Native people living in Southwest Alaska: a "subsistence foods" dietary pattern and two market-based dietary patterns "processed foods" and "fruits and vegetables". In this analysis, we aimed to characterize the association between the dietary patterns and cardiometabolic (CM) risk factors (lipids, blood pressure, glucose, adiposity).

METHODS AND RESULTS

We used multilevel linear regression to estimate the mean of each CM risk factor, comparing participants in the 4th to the 1st quartile of each dietary pattern (n = 637). Models were adjusted for age, sex, past smoking, current smoking, and physical activity. Mean log triglyceride levels were significantly higher among participants in the 4th compared to the 1st quartile of the processed foods dietary pattern (β = 0.11). Mean HbA1c percent was significantly lower (β = -0.08) and mean diastolic blood pressure (DBP) mm Hg was significantly higher (β = 2.87) among participants in the 4th compared to the 1st quartile of the fruits and vegetables dietary pattern. Finally, mean log triglyceride levels and mean DBP mm Hg were significantly lower among participants in the 4th compared to the 1st quartile of the subsistence foods dietary pattern (β = -0.10 and β = -3.99 respectively).

CONCLUSIONS

We found increased CM risk, as reflected by increased triglycerides, associated with eating a greater frequency of processed foods, and reduced CM risk, as reflected by lower triglycerides and DBP, associated with eating a greater frequency of subsistence foods.

Sex differences in obesity prevalence and cardiometabolic factors among Western Alaska Native people

BACKGROUND AND AIMS

Obesity is associated with increased risks of cardiovascular disease, type 2 diabetes, and other chronic diseases. Prevalence estimates for metabolic disorders are well documented in many populations, but Alaska Native groups are understudied. The Western Alaska Tribal Collaborative for Health Study combines data from three Alaska Native study cohorts to assess differences in obesity prevalence and associations with cardiometabolic risk factors by sex.

METHODS AND RESULTS

Analyses were based upon a sample of 3985 adult Yup'ik and Inupiat participants with a mean age of 40 years. Prevalence of obesity and metabolic risk factors was assessed according to nationally recognized guidelines. Regression analysis was used to evaluate the association between obesity and cardiometabolic risk factors, including lipids, blood pressure and glucose. The prevalence of obesity (BMI ≥ 30) was significantly higher in women (40%) than men (20%). Only 18.6% of men had a waist circumference (WC) > 102 cm, while 58% of women had a WC > 88 cm (p < 0.001). Women had higher mean HDL-C and triglyceride levels compared to men, while systolic and diastolic blood pressure, LDL-C, and glucose means were higher in men than in women. In multivariate analyses, BMI and WC were significantly associated with all of the cardiometabolic risk factors, although these associations were more pronounced in men than women.

CONCLUSION

The high prevalence of obesity and central adiposity among AN women is an important public health concern. Differences in associations between obesity and cardiometabolic risk factors by sex warrants further investigation to develop effective intervention programs.

Association of gestational diabetes mellitus and low-density lipoprotein (LDL) particle size

A predominance of small, dense low-density lipoproteins (LDL) is characteristic of the dyslipidemic state seen in type 2 diabetes. However, no study has investigated the association in gestational diabetes mellitus (GDM), which is pathophysiologically similar to type 2 diabetes. We hypothesized that LDL particle size is reduced in GDM cases compared with controls. Gradient gel electrophoresis was used to characterize LDL subclass phenotypes in non-fasting intrapartum plasma from 105 GDM cases and 96 controls. All participants were free of pre-existing diabetes or hypertension. The authors used logistic regression to estimate odds ratios (OR) and 95 % confidence intervals (CI) adjusted for confounders. Women with this phenotype had a significant 4.9-fold (95 % CI: 1.1-23.2) increased risk of GDM compared with those with the large, buoyant phenotype. The magnitude of this association was attenuated when plasma triglyceride and other confounders were included in the model (OR=4.2, 95 % CI: 0.5-39.5). Mean LDL particle size in GDM cases was smaller compared with controls (270.1 vs. 272.7A, p=0.003). The OR of GDM risk was 1.8 (95 % CI: 0.9-3.3) for every 10-A reduction in LDL particle size. Large prospective studies are needed to evaluate the association between smaller LDL particle size in early pregnancy with subsequent GDM risk.

Functional variants of the lipoprotein lipase gene and the risk of preeclampsia among non-Hispanic Caucasian women

Hypertriglyceridemia is an important pathophysiologic feature of preeclampsia, a common vascular disorder of pregnancy. Three well-documented functional variants (N291S, S447X, and D9N) of the lipoprotein lipase gene were related to hypertriglyceridemia. Results from the only two studies concerning the relationship between these variants and preeclampsia risk have been inconsistent. We investigated this relationship in a case-control study including 144 preeclamptic and 290 normotensive pregnant women (all non-Hispanic Caucasians). We estimated odds ratios (OR) and 95% confidence intervals (CI) adjusted for maternal age, pre-pregnancy body mass index, and parity. After adjusting for covariates, women with the 291 N/S or S/S genotype had significantly increased risk of preeclampsia (OR 6.9, 95% CI 1.9-25.4) compared with women with the common 291N/N genotype. The 447 S/X or X/X genotype was not significantly associated with preeclampsia risk. The frequency of the 9N variant allele was 1.8% in controls, while this allele was not observed among cases. Haplotype 9D/291S/447S was strongly associated with higher risk of preeclampsia as compared with the most common haplotype 9D/291N/447S (adjusted OR 6.6, 95% CI 1.7-25.0). Results from our study support the thesis that abnormal lipid metabolism is important in the pathogenesis of preeclampsia.

Heritability of C-reactive protein and association with apolipoprotein E genotypes in Japanese Americans

Numerous studies have demonstrated that increased C-reactive protein (CRP) levels predict coronary heart disease, stroke, peripheral vascular disease, and diabetes, and are associated with features of the metabolic syndrome. Only three previous studies have investigated the heritability of CRP levels, primarily in samples of Caucasian families. The purpose of the present study was to estimate the magnitude of genetic influences on CRP levels, and to examine potential associations between variation in the APOE gene and CRP levels, using a sample of 562 individual Japanese Americans from 68 extended kindreds. In general, correlation coefficients between first-degree relatives for CRP were approximately 0.2, and spouse correlations did not differ from zero, consistent with genetic influences. Heritability estimates were approximately 0.3 (p < 0.01), even with adjustment for factors known to influence CRP levels. A significant relationship was seen between unadjusted CRP levels and APOE genotypes (p = 0.02), with the highest mean CRP level among epsilon2 carriers (1.20 mg/L), and nearly the same mean levels among epsilon3/epsilon3 subjects and epsilon4 carriers (0.72 and 0.74 mg/L, respectively). However, this relationship was diminished with adjustment for covariates (p = 0.07). These results demonstrate the presence of both genetic and environmental effects on CRP levels among Asian Americans, and additional studies are needed to determine if the APOE gene contributes to these genetic influences.

Opportunities for public health genetics trainees: results of an employer/workplace survey

OBJECTIVE

To conduct the first employer/workplace survey identifying employment opportunities for graduates of programs with training in public health genetics in the USA, and to determine whether employment opportunities will increase in coming years.

METHODS

Six public health genetics training competencies were developed. A survey about workplace and employment opportunities was then conducted with mailings to (1) departments in schools of public health and departments of preventive medicine, (2) local and regional public health officials, (3) insurance companies and health management organizations (HMOs), and (4) biotechnology and pharmaceutical companies.

RESULTS

A total of 196 surveys were returned among 1,464 that were mailed. Response rates varied from 5.8 to 46.5% among the target groups. The percent of responding organizations currently employing individuals with skills in genetics ranged from 20 to 62%. The percent currently employing individuals with skills in public health ranged from 39 to 96%. Training opportunities such as internships or practicum experiences are reported for one-third of respondents. For all of the competencies, approximately half of survey respondents who rated the competency important or very important already employ individuals with public health genetics skills. Similarly, at least a quarter of survey respondents who rated the competency important or very important plan to hire individuals with that skill in the next 5 years. Overall, approximately 40% of those surveyed are planning to hire individuals with competencies in public health genetics in the next 5 years.

CONCLUSION

Employment opportunities already exist and new positions are becoming available in schools of public health and departments of preventive medicine, departments of public health, insurance companies and HMOs for professionals with public health genetics training. Based on our survey findings, skills and training in public health genetics are important in the workplace.

Genetic analysis of plant disease resistance pathways

Plant disease resistance (R) genes are introduced into high yielding crop varieties to improve resistance to agronomically important pathogens. The R gene-encoded proteins are recognitionally specific, interacting directly or indirectly with corresponding pathogen avirulence (avr) determinants, and are therefore under strong diversifying selection pressure to evolve new recognition capabilities. Genetic analyses in different plant species have also revealed more broadly recruited resistance signalling genes that provide further targets for manipulation in crop improvement strategies. Understanding the processes that regulate both plant-pathogen recognition and the induction of appropriate defences should provide fresh perspectives in combating plant disease. Many recent studies have utilized the model plant, Arabidopsis thaliana. Here, mutational screens have identified genes that are required for R gene function and for restriction of pathogen growth in compatible plant-pathogen interactions. Genetic analyses of these plant mutants suggest that whilst signalling pathways are conditioned by particular R protein structural types they are also influenced by pathogen lifestyle. Two Arabidopsis defence signalling genes, EDS1 and PAD4, are required for the accumulation of salicylic acid, a phenolic molecule required for systemic immunity. The cloning, molecular and biochemical characterization of these components suggests processes that may be important in their disease resistance signalling roles.

Lipoprotein and apolipoprotein abnormalities in familial combined hyperlipidemia: a 20-year prospective study

In order to characterize the lipoprotein abnormalities in familial combined hyperlipidemia (FCHL) and to describe factors associated with the stability of the FCHL phenotype during 20-year follow-up, 287 individuals from 48 families with FCHL originally identified in the early 1970s (baseline) were studied. Hyperlipidemia was defined as lipid-lowering medication use, or > or =age- and sex-specific 90th percentile for triglycerides or cholesterol. Triglyceride, cholesterol and medical history data were obtained at baseline and 20-year follow-up. Additional follow-up measures included HDL-C, LDL-C, LDL particle size, lipoprotein(a), apolipoprotein (apo) A-I, apoB, and apoE polymorphism. Longitudinally, two-thirds of relatives were consistently normolipidemic or hyperlipidemic, and one third were discordant for hyperlipidemic status at baseline and 20-year follow-up. Individuals with hyperlipidemia at baseline and/or follow-up had higher apoB levels than those with consistently normal lipids (P<0.05), whereas small LDL size was associated with concurrent hyperlipidemia. Among individuals who were normolipidemic at baseline, the following variables were independently associated with development of hyperlipidemia over 20 years: older age at baseline, male sex, greater increase in BMI during follow-up, and apoE alleles epsilon 2 or epsilon 4. In conclusion, apoB is associated with hyperlipidemia and apoE polymorphism is associated with later onset of hyperlipidemia in FCHL.

Lipoprotein(a) as a risk factor for maternal cardiovascular disease mortality in kindreds with familial combined hyperlipidemia or familial hypertriglyceridemia

Most but not all epidemiologic studies have shown that lipoprotein(a) [Lp(a)] is a risk factor for cardiovascular disease (CVD). Lp(a) levels are also strongly genetically influenced. The purpose of this study was to evaluate the association between Lp(a) levels in adult offspring and parental CVD mortality in 61 kindreds with familial forms of hyperlipidemia. The study sample consisted of offspring-parent pairs in which offspring had fasting Lp(a) measurements and parents had 20-year vital status data and standardized cause-of-death classification if deceased. Linear regression analyses, using a robust variance estimator, were performed separately for 241 offspring with known maternal history (114 mothers) and 194 offspring with known paternal history (93 fathers). Maternal history of CVD mortality was significantly (p=0.004) associated with 2.4-fold higher median Lp(a) levels in offspring compared with those with no maternal history, independent of diabetes, lipid-lowering medications and hormone use. No association was observed between paternal CVD mortality and offspring Lp(a) levels (p=0.505). Adjusting for apolipoprotein(a) kringle 4 number did not alter these parent-specific associations. In conclusion, Lp(a) levels in offspring may be associated with maternal but not paternal history of CVD mortality. This parent-specific finding needs to be confirmed in other samples of high-risk families.

Low density lipoprotein particle size and risk of early-onset myocardial infarction in women

Previous studies of middle-aged men have shown a univariate association between low density lipoprotein (LDL) particle diameter (size) and coronary heart disease, but this association has yet to be examined in younger women. Using a subsample from a population-based case-control study of women living in western Washington State, the authors examined the association between LDL particle size and risk of early-onset myocardial infarction (MI) in 1992-1995. Gradient gel electrophoresis was used to characterize LDL subclasses in nonfasting blood samples from 72 MI cases and 159 controls aged 20-44 years. Mean LDL particle size in cases was significantly smaller compared with controls (26.4 vs. 26.9 nm, p < 0.001), with an odds ratio of 2.3 (p < 0.0001) for a 1-nm smaller LDL particle size. These results were independent of age, menopausal status, smoking, diabetes, hypertension, and LDL cholesterol (odds ratios = 1.9-2.3 for a 1-nm smaller LDL particle size, all p < 0.02) but were not independent of body mass index, high density lipoprotein cholesterol, or triglyceride (odds ratios = 1.4, 1.4, and 1.1, respectively; all p > 0.05). Therefore, in age-adjusted analyses, smaller LDL particle size was associated with MI in young women, but the risk was attenuated after adjustments for metabolic factors related to both LDL particle size and MI.

Family history as a risk factor for early onset myocardial infarction in young women

BACKGROUND

The relation between a family history of heart attack and the occurrence of early myocardial infarction (MI) has not been studied extensively in women. In addition, whether recognized and newly-identified coronary heart disease (CHD) risk factors account for the familial aggregation of these events remains unknown. We therefore examined these questions in a population-based case-control study among female 18- to 44-year-old residents of western Washington State.

METHODS AND RESULTS

The patients consisted of 107 women with first acute MI, and the control subjects comprised 526 women similar in age identified from the community and without a history of recognized clinical coronary heart disease or stroke. Trained interviewers used a structured questionnaire to elicit a detailed history of heart attack in first-degree relatives. Information about other known MI risk factors was collected and biochemical measurements performed, and common polymorphisms in various candidate genes were determined. The rate of MI among first-degree relatives of MI cases was twice as high as among first-degree relatives of controls (relative risk, 1.96; 95% confidence interval (CI), 1.46-2.48); this association was present for each familial relationship. Sibling history of MI but not parental history was associated with MI, after controlling for established CHD risk factors. In a subsample of subjects with blood measurements, further adjustment for lipids, lipoproteins and specific genetic risk factors slightly reduced the association with sibling MI history (from odds ratio (OR), 5.17; 95% CI, 1.93-13.85 to OR, 3.97; 95% CI, 0.92-17.17).

CONCLUSION

Family history of MI is positively associated with the risk of early MI in women. While the association with parental history of MI is mediated through the clustering of other common risk factors, the association of sibling history of MI with early-onset MI in young women is only partially explained by the clustering of established and newly-identified risk factors.

The interface of genetics and public health: research and educational challenges

As the target date for the sequencing of the human genome approaches, there is growing recognition that public health practice, research, and education will be impacted by new genetic technologies and information and that a multidisciplinary approach is required. Research in the emerging field of public health genetics encompasses a broad range of disciplines and will increasingly involve the interactions among the investigators in these fields. An overview of these areas of research is provided, with illustrative examples. Education in public health genetics needs to address a variety of audiences, including public health graduate students and practitioners, students from related disciplines, and health care professionals. Two new graduate programs at the Universities of Michigan and Washington and training opportunities for public health professionals are described. These educational efforts must be ongoing so that the potential of genetic technology and information can be appropriately used to benefit the health of all.

Lipoprotein candidate genes for multivariate factors of the insulin resistance syndrome: a sib-pair linkage analysis in women twins

The insulin resistance syndrome (IRS) is characterized by a combination of interrelated coronary heart disease risk factors, including low high-density lipoprotein cholesterol (HDLC) levels, obesity and increases in triglyceride (TG), systolic and diastolic blood pressure (BP), small low-density lipoprotein particles (LDL-size), and fasting and postload plasma insulin and glucose. Using factor analysis, we previously identified multivariate factors based on data from women participating in the Kaiser Permanente Women Twins Study: 1) Weight/Fat, 2) Insulin/Glucose, 3) Lipids, and 4) BP. The purpose of this study is to evaluate evidence for genetic linkage between the multivariate factors and candidate genes. Quantitative sib-pair analysis based on the factor scores with markers for 9 candidate genes was carried out based on data from 126 pairs of dizygotic (DZ) women twins from the second exam of the Kaiser Permanente Women Twins study. Suggestive evidence for linkage was found for the Weight/fat factor and the Apo E gene (p = 0.01), and stronger evidence for linkage with the Lipid factor and the cholesterol ester transfer protein (p = 0.002) gene. Therefore, the CETP gene appears to influence covariation in LDL size, TG, and HDL, and may account for a portion of the well-established statistical and metabolic associations observed between these risk factors.

Low-density lipoprotein particle size, triglycerides, and high-density lipoprotein cholesterol as risk factors for coronary heart disease in older Japanese-American men

Decreased low-density lipoprotein (LDL) particle size is associated with coronary heart disease (CHD) risk among middle-aged Caucasian populations, and has been consistently correlated with increased plasma levels of triglyceride and decreased levels of high-density lipoprotein (HDL) cholesterol. This study examines whether these risk factors predict CHD among older Japanese-American men. With use of the Honolulu Heart Program Lipoprotein Exam 3 (1980 to 1982) as baseline, and 12-year follow-up for CHD events, a nested, case-control study was designed. One hundred forty-five incident CHD cases were identified and matched to 2 controls each. LDL particle diameter (size) was determined by gradient gel electrophoresis. A 10-angstrom (A) decrease in LDL size at baseline was associated with increased risk of incident CHD (relative risk 1.28, 95% confidence interval 1.01 to 1.63). After adjustment for baseline risk factors, the LDL size association was no longer statistically significant (relative risk 1.13, 95% confidence interval 0.86 to 1.49). When principal components analysis was used to define a composite variable for LDL size, triglycerides, and HDL cholesterol, this component predicted CHD independent of smoking, alcohol consumption, physical activity, body mass index, hypertension, diabetes, and beta-blocker use (p <0.01). Therefore, this prospective analysis of data from older, Japanese-American men demonstrated that decreased LDL size is a univariate predictor of incident CHD, and that a composite risk factor of LDL size, triglyceride, and HDL cholesterol was a risk factor for CHD independent of other risk factors.

Sib-pair linkage analysis of longitudinal changes in lipoprotein risk factors and lipase genes in women twins

Based on longitudinal twin data in women, we have previously demonstrated a genetic influence on changes in lipoprotein risk factors, blood pressure measurements, and body mass index over a decade. The present study examined the linkage between changes in lipoprotein variables and candidate genes encoding the hormone-sensitive lipase (HSL), hepatic lipase (HL), and lipoprotein lipase (LPL). The sample consisted of 126 dizygotic (DZ) pairs of women twins who participated in the two examinations of the Kaiser Permanente Women Twins Study, performed a decade apart. Using quantitative sib-pair linkage analysis, a linkage was demonstrated between the locus for hormone-sensitive lipase and age-adjusted changes in plasma triglyceride (P = 0.015), which became more significant after adjustment for environmental factors and the exam-1 level (P = 0.005). There was also evidence suggesting linkage between the locus for hepatic lipase and changes in triglyceride (P = 0.023), but no linkage was detected for lipoprotein lipase and changes of lipid levels with time. These findings suggest that variation at these candidate gene loci may underlie a portion of the intraindividual variations in these coronary heart disease (CHD) risk factors, and that studies to identify the functional variants could provide new insights into genetic susceptibility to cardiovascular disease.

Heritability of factors of the insulin resistance syndrome in women twins

The insulin resistance syndrome (IRS) is characterized by a combination of interrelated coronary heart disease (CHD) risk factors, including low high-density lipoprotein cholesterol (HDL-C) levels, obesity and increases in triglyceride (TG), blood pressure, small low-density lipoprotein particles (LDL), and both fasting and postload plasma insulin and glucose. Using factor analysis, we previously identified 3 uncorrelated factors that explained 66% of the variance among these variables, based on data from women participating in examination 2 of the Kaiser Permanente Women Twins Study in Oakland, CA during 1989-1990. The factors were interpreted as: 1) body mass/fat distribution, 2) insulin/glucose, and 3) lipids: TG, HDL-C, LDL peak particle diameter. In this analysis, heritability of each of the factors was estimated based on data from 140 monozygotic and 96 dizygotic pairs of non-diabetic women twins. Heritability estimates were calculated using the classical approach, the analysis of variance (ANOVA) approach, and the maximum likelihood approach. For the body mass/fat distribution factor heritability estimates suggest moderate genetic influences; 0.61 (P < 0.001), 0.14 (P > 0.05), and 0.71 (P < 0.001), respectively. The insulin/glucose factor appeared to be highly heritable, with estimates of 0.87, 0.92, and 0.57 (all P < 0.001), respectively. The heritability estimates for the lipid factor were moderate and consistent across methods: 0.25 (P < 0.10), 0.32 (P < 0.05), and 0.30 (P < 0.05), respectively. These results are consistent with genetic influences on each of the 3 "factors," and suggest that both genetic and environmental effects are involved in the clustering of IRS risk factors.

Cardiovascular disease mortality in familial forms of hypertriglyceridemia: A 20-year prospective study

BACKGROUND

Familial combined hyperlipidemia (FCHL) and familial hypertriglyceridemia (FHTG) are 2 of the most common familial forms of hyperlipidemia. There is a paucity of prospective data concerning the risk of cardiovascular disease (CVD) in such families. The purposes of this study were to estimate 20-year total and CVD mortality risk among relatives in these families and to evaluate plasma triglyceride as a predictor of death.

METHODS AND RESULTS

The study was based on lipid and medical history data from 101 families ascertained in 2 studies conducted in the early 1970s. Vital status and cause of death was determined during 1993 to 1997 for 685 family members, including first-degree relatives of the probands and spouse control subjects. Compared with spouse control subjects, 20-year CVD mortality risk was increased among siblings and offspring in FCHL (relative risk 1.7, P=0.02) after adjustment for baseline covariates. In FHTG families, the relative risk was also 1.7 but was not statistically significant (P=0.39). Baseline triglyceride was associated with increased CVD mortality risk independent of total cholesterol among relatives in FHTG families (relative risk 2.7, P=0.02) but not in FCHL families (relative risk 1.5, P=0.16) after adjustment for baseline covariates.

CONCLUSIONS

This prospective study establishes that relatives in FCHL families are at increased risk for CVD mortality and illustrates the need for effective prevention strategies in this group. Baseline triglyceride level predicted subsequent CVD mortality among relatives in FHTG families, adding to the growing evidence for the importance of hypertriglyceridemia as a risk factor for CVD.

Inheritance of LDL peak particle diameter: results from a segregation analysis in Israeli families

Genetic and environmental determinants of LDL peak particle diameter (LDL-PPD) were investigated in a sample of 80 kindreds residing in kibbutz settlements in Israel. The sample included 182 males and 191 females ages 15-93 years. LDL-PPD levels were first adjusted for variability in sex and age. Commingling analysis demonstrated that a mixture of two normal distributions fit the adjusted LDL-PPD levels better than did a single normal distribution. Complex segregation analysis was first applied to these sex and age adjusted data but was not conclusive. However, when the regression model for sex and age allowed coefficients to be ousiotype (class) specific, the mixed environmental model was rejected while a major Mendelian model was not. These results suggest that the particular genotypes determined by the major gene, which are associated with different phenotypic variances, are likely to be more realistic, and that this analytic approach can contribute to improving our understanding of the genetics of LDL particle size.

Linkage of the cholesteryl ester transfer protein (CETP) gene to LDL particle size: use of a novel tetranucleotide repeat within the CETP promoter

BACKGROUND

A preponderance of small, dense LDL particles, elevated levels of plasma triglycerides (TG), and low levels of HDL characterize the atherogenic lipoprotein phenotype, which is associated with increased coronary artery disease (CAD) risk. Genetic and environmental factors influence LDL size, cholesteryl ester transfer protein (CETP) being one of the candidate genes. CETP mediates the transfer of cholesteryl ester from HDL to apolipoprotein (apo) B-containing lipoproteins in exchange for TG, promoting reverse cholesterol transfer and remodeling of lipoprotein particles.

METHODS AND RESULTS

We have identified a tetranucleotide repeat (fragment sizes from 324 to 464 bp; heterozygosity index = 0.74) within the CETP promoter and used it in quantitative sib-pair linkage analysis in 119 female dizygotic (DZ) twins. Linkage was found to LDL size (P<0.001), TG (P<0.005), and plasma apoB (P = 0.02). The distribution of the tetranucleotide repeats was bimodal, and there was strong allelic association of the "short" alleles with the B2 allele of CETP TaqIB polymorphic site (P<0.001).

CONCLUSIONS

This report of linkage of the CETP gene to LDL particle size adds to the list of candidate genes linked to LDL size, supporting the hypothesis of multigenic determination of LDL size heterogeneity. Whether this promoter variation is itself functional or is a marker for a functional site in the CETP gene remains to be determined.

Triglyceride, small, dense low-density lipoprotein, and the atherogenic lipoprotein phenotype

This review provides an overview of the recent data evaluating triglyceride and low-density lipoprotein (LDL) size, two highly interrelated, genetically influenced, risk factors for cardiovascular disease (CVD). An examination of new epidemiologic studies continues to demonstrate that plasma triglyceride levels predict CVD. The first prospective study of the familial forms of hypertriglyceridemia has shown that relatives in familial-combined hyperlipidemia families are at increased risk for CVD mortality and that triglyceride levels predicted 20-year, CVD mortality among relatives in familial hypertriglyceridemia families. A meta-analysis of three, large-scale, prospective studies in men, and the first study to examine the correlation of LDL particle size distribution and vascular changes measured by B-mode ultrasound, add to growing evidence that small, dense LDL is atherogenic. Quantitative genetic analysis has recently shown substantial pleiotropic (common) genetic effects on triglyceride and LDL size. At least part of this may be explained by variation at the cholesterol ester transfer protein locus on chromosome 16, possibly through its role in reverse cholesterol transport. Taken together, these data provide new insights into the importance of triglyceride and LDL particle size for understanding genetic susceptibility to cardiovascular disease and its prevention.

LDL size and risk of coronary heart disease in elderly men and women

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 A, 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 A(1), 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.

Pleiotropic genetic effects on LDL size, plasma triglyceride, and HDL cholesterol in families

The interrelationships among low density lipoprotein (LDL) particle size, plasma triglyceride (TG), and high density lipoprotein cholesterol (HDL-C) are well established and may involve underlying genetic influences. This study evaluated common genetic effects on LDL size, TG, and HDL-C by using data from 85 kindreds participating in the Genetic Epidemiology of Hypertriglyceridemia (GET) Study. A multivariate, maximum likelihood-based approach to quantitative genetic analysis was used to estimate the additive effects of shared genes and shared, unmeasured nongenetic factors on variation in LDL size and in plasma levels of TG and HDL-C. A significant (P<0.001) proportion of the variance in each trait was attributable to the additive effects of genes. Maximum-likelihood estimates of heritability were 0.34 for LDL size, 0.41 for TG, and 0.54 for HDL-C. Significant (P<0.001) additive genetic correlations (rho(G)), indicative of the shared additive effects of genes on pairs of traits, were estimated between all 3 trait pairs: for LDL size and TG rho(G)=-0.87, for LDL size and HDL-C rho(G)=0.65, and for HDL-C and TG rho(G)=-0.54. A similar pattern of significant environmental correlations between the 3 trait pairs was also observed. These results suggest that a large proportion of the well-documented correlations in LDL size, TG, and HDL-C are likely attributable to the influence of the same gene(s) in these families. That is, the gene(s) that may contribute to decreases in LDL size also contribute significantly to higher plasma levels of TG and lower plasma levels of HDL-C. These relationships may be useful in identifying genes responsible for the associations between these phenotypes and susceptibility to cardiovascular disease in these families.

Epidemiology of hypertriglyceridemia and cardiovascular disease

Epidemiologic studies provide increasing evidence that hypertriglyceridemia is a risk factor for cardiovascular disease. A meta-analysis of 17 population-based studies of triglyceride levels and cardiovascular disease identified a 76% increase in cardiovascular disease risk in women and a 31% increase in men associated with a 1 mmol/L increase in triglyceride levels. Additional epidemiologic studies have shown that plasma triglyceride levels and low-density lipoprotein particle size predict subsequent coronary artery disease. Taken together, the existing epidemiologic data may help identify a group of patients who may benefit from interventions aimed at decreasing triglyceride levels associated with an increased risk of cardiovascular disease.

Linkage of low-density lipoprotein size to the lipoprotein lipase gene in heterozygous lipoprotein lipase deficiency

Small low-density lipoprotein (LDL) particles are a genetically influenced coronary disease risk factor. Lipoprotein lipase (LpL) is a rate-limiting enzyme in the formation of LDL particles. The current study examined genetic linkage of LDL particle size to the LpL gene in five families with structural mutations in the LpL gene. LDL particle size was smaller among the heterozygous subjects, compared with controls. Among heterozygous subjects, 44% were classified as affected by LDL subclass phenotype B, compared with 8% of normal family members. Plasma triglyceride levels were significantly higher, and high-density lipoprotein cholesterol (HDL-C) levels were lower, in heterozygous subjects, compared with normal subjects, after age and sex adjustment. A highly significant LOD score of 6.24 at straight theta=0 was obtained for linkage of LDL particle size to the LpL gene, after adjustment of LDL particle size for within-genotype variance resulting from triglyceride and HDL-C. Failure to adjust for this variance led to only a modest positive LOD score of 1.54 at straight theta=0. Classifying small LDL particles as a qualitative trait (LDL subclass phenotype B) provided only suggestive evidence for linkage to the LpL gene (LOD=1. 65 at straight theta=0). Thus, use of the quantitative trait adjusted for within-genotype variance, resulting from physiologic covariates, was crucial for detection of significant evidence of linkage in this study. These results indicate that heterozygous LpL deficiency may be one cause of small LDL particles and may provide a potential mechanism for the increase in coronary disease seen in heterozygous LpL deficiency. This study also demonstrates a successful strategy of genotypic specific adjustment of complex traits in mapping a quantitative trait locus.

Linkage analysis of candidate genes and the small, dense low-density lipoprotein phenotype

There is accumulating evidence for the importance of small, dense low-density lipoprotein (LDL), the defining feature of the atherogenic lipoprotein phenotype, as a risk factor for coronary heart disease. Although both family studies and twin studies have demonstrated genetic influences on this phenotype, the specific gene(s) involved remain to be identified. The purpose of this study was to determine whether there was evidence for genetic linkage between small, dense LDL (LDL subclass phenotype B), as determined by gradient gel electrophoresis, and selected candidate genes known to be involved in lipid metabolism. The linkage analyses were based on a sample of 19 families, including 142 individual family members, using a lod score linkage analysis approach. Nine candidate genes were examined, including loci for manganese superoxide dismutase (Mn SOD2), apolipoproteins CIII, AII, and apo CII, lipoprotein lipase, hepatic lipase, microsomal triglyceride transport protein, the insulin receptor and the LDL receptor. The analyses did not provide significant evidence for genetic linkage between markers for any of these genes and LDL subclass phenotype B, nor did it confirm previous reports of linkage between the LDL receptor gene and LDL subclass phenotype B. Using three closely linked markers for the Mn SOD2 locus excluded close linkage between this candidate gene region and LDL subclass phenotype B. These findings demonstrate the complexity of genetically mapping risk factor phenotypes, and emphasize the necessity of identifying new genetic loci, other than known candidate genes, involved in susceptibility to atherosclerosis.

Plasma triglyceride as a risk factor for cardiovascular disease

OBJECTIVES

To determine whether cardiovascular disease can be predicted from plasma triglyceride levels independent of high density lipoprotein (HDL) cholesterol levels.

DATA SOURCES AND STUDY SELECTION

Seventeen prospective population-based studies, ranging in publication date from 1965 to 1994, were selected based on uniform criteria. Among the studies, 16 were conducted in men, comprising data for more than 46,000 men, with an average follow-up period of more than eight years. The data from women came from five studies, including nearly 11,000 women with an 11-year average follow-up.

DATA EXTRACTION

The meta-analysis protocol comprised the following steps: estimating the relative risk of triglyceride for each individual study; standardizing the relative risks to a 1 mmol/l increase; estimating the variance and weighting relative risks by the inverse of the variance, and estimating the summary relative risk by averaging the weighted relative risks.

DATA SYNTHESIS

Among studies of men, relative risks ranged from 1.1 to 2.0, and the summary relative risk was 1.32 (P < 0.05). All five studies of women were statistically significant, with relative risks ranging from 1.7 to 2.0. The summary relative risk was 1.76, higher than that for men, indicating a 76% increase in risk for women. For the studies among men in which HDL cholesterol data were available, multivariate adjusted relative risk was attenuated to 1.14 but was still statistically significant. Although only two studies of women included HDL cholesterol adjustments, the summary relative risk for triglyceride remained statistically significant, with a value of 1.37.

CONCLUSIONS

When all available data are taken into account, increased triglyceride level is a risk factor for cardiovascular disease, independent of HDL cholesterol level. Clinical trials are urgently needed to determine whether lowering plasma triglyceride levels reduces subsequent risk of cardiovascular disease.

Factors of the insulin resistance syndrome in nondiabetic and diabetic elderly Japanese-American men

Factor analysis has previously identified four independent factors that characterize the insulin resistance syndrome in women, interpreted as 1) weight/waist, 2) lipids, 3) insulin/glucose, and 4) systolic and diastolic blood pressure. Because it is not known whether similar factors emerge for men, or for diabetics, factor analysis was used to investigate the clustering of features characterizing the insulin resistance syndrome using data from 3,159 elderly (71-93 years) Japanese-American men participating in the fourth examination of the Honolulu Heart Program during 1991-1993. Consistent with previous results, factor analysis reduced eight risk factors (insulin, glucose, systolic blood pressure, diastolic blood pressure, triglycerides, high-density lipoprotein cholesterol, weight, and waist circumference) to four uncorrelated factors that explained 78.2% and 74.7% of the variance in nondiabetics (n = 2,760) and diabetics (n = 399), respectively. These factors were interpreted as 1) weight/waist, 2) blood pressure, 3) lipids, and 4) insulin/glucose. Modest differences in the associations between fasting insulin and factors 1, 3, and 4 were noted for diabetics. These consistently identified composite factors may represent markers for underlying pathophysiologic mechanisms of the insulin resistance syndrome and risk of non-insulin-dependent diabetes mellitus.

Hypertriglyceridemia as a cardiovascular risk factor

To determine the relation between plasma triglyceride levels and the risk of incident cardiovascular disease, the semiquantitative techniques of meta-analysis were applied to 17 population-based prospective studies of triglyceride and cardiovascular disease. Sixteen of these studies represented 2,445 events among 46,413 Caucasian men followed for an average period of 8.4 years, and 5 studies represented 439 events among 10,864 Caucasian women followed for an average period of 11.4 years. Univariate relative risk (RR) estimates for incident cardiovascular disease associated with a 1-mmol/L increase in triglyceride was 1.07-1.98 in men, with a summary RR of 1.32 (95% confidence interval [CI]: 1.26-1.39), indicating a 32% increase in disease risk associated with increased triglyceride. In the studies involving women, individual RR estimates for triglyceride were 1.69-2.05, with a summary RR of 1.76 (95% CI: 1.50-2.07), indicating a 76% increase in disease risk associated with increased triglyceride. After adjustment for high-density lipoprotein cholesterol and other risk factors, these risks were decreased to 14% in men and 37% in women but remained statistically significant. Three recent prospective epidemiologic studies have also shown that plasma triglyceride and low-density lipoprotein particle size predict subsequent coronary artery disease in Caucasian populations. Taken together, these studies demonstrate the importance of triglyceride levels as a risk factor for cardiovascular disease.

Candidate-gene studies of the atherogenic lipoprotein phenotype: a sib-pair linkage analysis of DZ women twins

There is a growing body of evidence supporting the roles of small, dense LDL and plasma triglyceride (TG), both features of the atherogenic lipoprotein phenotype, as risk factors for coronary heart disease. Although family studies and twin studies have demonstrated genetic influences on these risk factors, the specific genes involved remain to be determined definitively. The purpose of this study was to investigate genetic linkage between LDL size, TG, and related atherogenic lipoproteins and candidate genes known to be involved in lipid metabolism. The linkage analysis was based on a sample of 126 DZ women twin pairs, which avoids the potentially confounding effects of both age and gender, by use of a quantitative sib-pair linkage-analysis approach. Eight candidate genes were examined, including those for microsomal TG-transfer protein (MTP), hepatic lipase, hormone-sensitive lipase, apolipoprotein (apo) B, apo CIII, apo E, insulin receptor, and LDL receptor. The analysis suggested genetic linkage between markers for the apo B gene and LDL size, plasma levels of TG, of HDL cholesterol, and of apo B, all features of the atherogenic lipoprotein phenotype. Furthermore, evidence for linkage was maintained when the analysis was limited to women with a major LDL-subclass diameter >255 A, indicating that the apo B gene may influence LDL heterogeneity in the intermediate-to-large size range. In addition, linkage was found between the MTP gene and TG, among all the women. These findings add to the growing evidence for genetic influences on the atherogenic lipoprotein phenotype and its role in genetic susceptibility to atherosclerosis.

Evidence against linkage of familial combined hyperlipidemia to the apolipoprotein AI-CIII-AIV gene complex

Familial combined hyperlipidemia (FCHL) was originally described as a disorder characterized by elevated levels of either plasma cholesterol or triglyceride (TG) or both in members ofthe same family. More recent studies have indicated that apolipoprotein B levels (apoB) are also elevated in these individuals. Although a dominant mode of inheritance was originally proposed, recent studies have questioned this simple mode of inheritance, and the genetic basis of the disorder has eluded investigators. A study that reported evidence that FCHL is linked to the apolipoprotein AI-CIII-AIV region on chromosome 11 is therefore of interest. We have attempted to replicate this finding in three large, well-characterized FCHL kindreds by using a highly polymorphic marker in the apoCIII gene. Using the same definitions and parameters as were used in the initial report, we obtained strong evidence against linkage of FCHL to the apolipoprotein AI-CIII-AIV region on chromosome 11 (combined lod score of -7.87 at 0% recombination). Two other models, one based on total cholesterol (TC) levels alone and one based on the joint distribution of TC and apoB levels, also gave evidence against linkage of FCHL to this region (lod scores at 0% recombination of -8.95 and -2.58, respectively). An additional regression-based linkage analysis also gave no support for the existence of a locus in this region that influences these lipid levels in these pedigrees. Explanations for the differences in results between these studies include genetic heterogeneity, differences in clinical phenotype used to select the pedigrees, and ascertainment bias.

Family history as a risk factor for primary cardiac arrest

BACKGROUND

The hypothesis that a family history of myocardial infarction (MI) or primary cardiac arrest (PCA) is an independent risk factor for primary cardiac arrest was examined in a population-based case-control study. In addition, we investigated whether recognized risk factors account for the familial aggregation of these cardiovascular events.

METHODS AND RESULTS

PCA cases, 25 to 74 years old, attended by paramedics during the period 1988 to 1994 and population-based control subjects matched for age and sex were identified from the community by random digit dialing. All subjects were free of recognized clinical heart disease and major comorbidity. A detailed history of MI and PCA in first-degree relatives was collected in interviews with the spouses of case and control subjects by trained interviewers using a standardized questionnaire. For each familial relationship, there was a higher rate of MI or primary cardiac arrest (MI/PCA) in relatives of case compared with relatives of control subjects. Overall, the rate of MI/PCA among first-degree relatives of cardiac arrest patients was almost 50% higher than that in first-degree relatives of control subjects (rate ratio [RR]=1.46; 95% CI=1.23 to 1.72). In a multivariate logistic model, family history of MI/PCA was associated with PCA (RR=1.57; 95% CI=1.27 to 1.95) even after adjustment for other common risk factors.

CONCLUSIONS

Family history of MI or PCA is positively associated with the risk of primary cardiac arrest. This association is mostly independent of familial aggregation of other common risk factors.

A double-blind, randomized, controlled trial of the effects of two eggs per day in moderately hypercholesterolemic and combined hyperlipidemic subjects taught the NCEP step I diet

OBJECTIVE

We investigated the effect of egg feeding in 161 hypercholesterolemic (HC) or combined hyperlipidemic (CHL) free-living subjects taught the NCEP Step I Diet.

METHODS

Subjects had LDL-C between 3.36 and 4.91 mmol/L (130 and 190 mg/dL). HC subjects had triglyceride concentrations < 75th %-ile, CHL > or = 75th %-ile. Six weeks after instituting the Step I Diet, subjects were randomized to two eggs or egg substitute daily for 12 weeks.

RESULTS

Mean dietary intake (9-day food records) was within NCEP guidelines and 131 subjects finished with stable weight and adherence > 80%. Placebo group lipoprotein lipids were unchanged at study end for both HC (n = 35) and CHL (n = 21) subjects. Egg-fed HC subjects (n = 44) increased LDL-C nonsignificantly, 0.07 mmol/L (3 mg/dL) (p = 0.49). Egg-fed CHL subjects (n = 31) increased LDL-C 0.31 mmol/L (12 mg/dL) (p < 0.001). HDL-C increased significantly in both HC and CHL groups, 0.10 and 0.08 mmol/L (4 and 3 mg/dL, p = 0.003 and 0.02), respectively. HC and CHL subjects did not differ by apo E phenotype distribution. Postprandial triglyceride-rich lipoproteins and LDL subclass phenotype were unaffected by egg feeding in subsets of subjects.

CONCLUSIONS

CHL subjects ingesting a Step I Diet in a free-living setting are sensitive to egg feeding and should benefit from dietary cholesterol restriction. The limited LDL-C rise in HC subjects resembles that seen in egg-fed normocholesterolemic subjects ingesting a Step I Diet, but requires confirmation.

Genetic influences on changes in body mass index: a longitudinal analysis of women twins

Numerous studies have demonstrated genetic influences on body fat, but there also may be genetic effects on its intraindividual variation over time. This study examined changes in body mass index (BMI) using longitudinal data from two examinations of the Kaiser Permanente Women Twins Study, performed a decade apart. The analysis included 630 women, 185 monozygotic and 130 dizygotic twin pairs, with average ages of 41 years and 51 years at the two examinations, respectively. Age-adjusted heritability estimates for the change in BMI over the decade ranged from 0.57 to 0.86 (all p < or = 0.001) using three different statistical analysis approaches, indicating that at least half and possibly as much as 85%+, of the variance in the change in BMI is attributable to genetic influences under a polygenic model. These estimates remained statistically significant after adjusting for environmental factors (ranging from 0.57 to 0.78) and with additional adjustment for BMI at baseline (ranging from 0.41 to 0.79), although dizygotic intraclass correlations were low after these adjustments. Thus, in addition to known environmental and behavioral influences, these results provide evidence for genetic influences on changes in BMI over a decade in women.

Triacylglycerol and coronary heart disease

Based on meta-analysis of prospective studies from the epidemiological literature, TAG is a risk factor for CVD, independent of HDL-cholesterol. The RR values were 1.3 and 1.8 for a 1 mmol/l increase in TAG among men and women respectively. Adjustment for HDL-cholesterol and other risk factors attenuated these estimates, but they remained statistically significant. Recent prospective findings from the Stanford Five City Project (Gardner et al. 1996) and the Physicians' Health Study (Stampfer et al. 1996) further demonstrate that TAG and LDL size are highly inter-related risk factors for CHD. Quantitative genetic analysis from large-scale family studies show that these correlations reflect common genetic influences that may be important for understanding genetic susceptibility to CHD.

Evidence that the apolipoprotein E-genotype effects on lipid levels can change with age in males: a longitudinal analysis

We previously reported that change, with age, in plasma levels of total cholesterol (TC) and LDL cholesterol (LDL-C) differed between apolipoprotein E (APOE) genotypes epsilon 3 epsilon 3 and epsilon 3 epsilon 4, in a sample of 77 older, unrelated males. By use of a larger sample from that cohort, followed longitudinally during 1969-87, the change in TC and in LDL-C, between the epsilon 3 epsilon 3 and epsilon 3 epsilon 4 APOE genotypes, over three exams, was reanalyzed. Additionally, the change in triglycerides (TG) and in HDL-cholesterol (HDL-C), between the epsilon 3 epsilon 3 and epsilon 3 epsilon 4 APOE genotypes-as well as the differences between the epsilon 3 epsilon 3 and epsilon 3 epsilon 2 genotypes, for TC, LDL-C, TG, and HDL-C-were contrasted over the three exams. At exam 1 TG was higher in the epsilon 3 epsilon 4 group than in the epsilon 3 epsilon 3 group (mean age 48 years), and at exams 2 and exam 3 (mean ages 58 and 63 years, respectively) it was similar (P = .009 for the exam-by-genotype-interaction effect in the repeated-measures analysis). A similar trend was seen for TC (P = .03), yet previously detected LDL-C effects were not apparent (P = .46). Those with the epsilon 3 epsilon 2 genotype had higher TG and lower LDL-C and TC at each exam than were seen in those with the epsilon 3 epsilon 3 genotype, although the differences in the values were not always statistically significant. Differences in TC, LDL-C, and TG, between the epsilon 3 epsilon 2-genotype and epsilon 3 epsilon 3-genotype groups, did not significantly change over the three exams. HDL-C levels were relatively stable over the exams; however, the exam-by-genotype interaction was significant for the epsilon 3 epsilon 2 genotype versus the epsilon 3 epsilon 3 genotype (P = .02). The epsilon 4 allele effects on TG and TC changed between longitudinal exams and may be age dependent. Changes, with age, in the effect of the epsilon 3 epsilon 4 genotype on lipids may impact the risk of developing atherosclerotic disease.

Heritability of longitudinal changes in coronary-heart-disease risk factors in women twins

Numerous studies have demonstrated genetic influences on levels of coronary heart disease (CHD) risk factors, but there also may be genetic effects on the intraindividual variation in these risk factors over time. Changes in risk factors are likely to reflect genetic-environmental interactions and may have important implications for understanding CHD risk. The present study examines the heritability of changes in CHD risk factors, using data from the two examinations by the Kaiser Permanente Women Twins Study, performed a decade apart. The sample consisted of 348 pairs of women twins who participated in both examinations, including 203 MZ pairs and 145 DZ pairs. Average ages at the two examinations were 41 and 51 years, respectively. By means of three different statistical analytic approaches, moderate heritability estimates were demonstrated for changes in LDL cholesterol (h2 = .25-.36) and in HDL cholesterol (h2 = .23-.58), some of which were statistically significant. Although small to moderate heritability estimates were found for systolic blood pressure (.18-.37; P < .05 for some estimates), no genetic influence on changes in diastolic blood pressure was detected. Based on longitudinal twin data in women, this study demonstrates a genetic influence on changes in both lipoprotein risk factors and systolic blood pressure over a decade. In addition to environmental factors, which clearly are operating, the effect of various "variability genes" may be acting independently of the genetic influences on the absolute levels of these risk factors. Both mapping the gene(s) underlying intraindividual variations in these CHD risk factors and understanding their function(s) could lead to targeted intervention strategies to reduce CHD risk among genetically susceptible individuals.

Plasma triglyceride level is a risk factor for cardiovascular disease independent of high-density lipoprotein cholesterol level: a meta-analysis of population-based prospective studies

OBJECTIVES

Despite nearly 40 years of research, the role of plasma triglyceride as a risk factor for cardiovascular disease remains elusive. The objectives of the present study were to quantify the magnitude of the association between triglyceride and cardiovascular disease in the general population, and to determine whether this relationship is independent of high-density lipoprotein (HDL) cholesterol, using the semi-quantitative techniques of metaanalysis.

METHODS AND DESIGN

Seventeen studies were selected for the analysis based on published reports of population-based, prospective studies, including 46413 men and 10864 women. To insure comparability, only studies reporting the association between fasting triglyceride levels and incident cardiovascular endpoints were included. Using standard meta-analysis calculations, relative risks (RR) and 95% confidence intervals (CI) were calculated and standardized with respect to a 1 mmol/l increase in triglyceride. Multivariable-adjusted RRs were determined for the six studies in men and two studies in women that reported adjustments for HDL cholesterol.

RESULTS

For men and women, the univariate RRs for triglyceride were 1.32 (95% Cl 1.26-1.39) and 1.76 (95% Cl 1.50-2.07), respectively, indicating an approximately 30% increased risk in men and a 75% increase in women. Adjustment of HDL cholesterol and other risk factors attenuated these RRs to 1.14 (95% Cl 1.05-1.28) and 1.37 (95% Cl 1.13-1.66), respectively, which were still statistically significant values.

CONCLUSION

Based on combined data from prospective studies, triglyceride is a risk factor for cardiovascular disease for both men and women in the general population, independent of HDL cholesterol. These finding demonstrate the necessity for clinical trials to evaluate whether lowering plasma triglyceride decreases the risk of cardiovascular disease.

Genetic epidemiology of dyslipidaemia and atherosclerosis

The clinical relevance of the heterogeneity in the size and density of low-density lipoprotein (LDL) particles is now widely recognized. The evidence from epidemiological studies, family studies and twins studies demonstrates that small, dense LDL (LDL subclass phenotype B) is a common, genetically influenced risk factor for coronary heart disease (CHD). Several atherogenic mechanisms have been proposed to explain the association of small, dense LDL with CHD, including evidence that small, dense LDL is an integral feature of the insulin resistance syndrome. Furthermore, a recent study in elderly Finnish men and women has shown that phenotype B prospectively predicts non-insulin-dependent diabetes mellitus (NIDDM). In addition, ongoing studies of large Japanese-American kindreds will provide valuable data for evaluating small, dense LDL as a marker for genetic susceptibility to both CHD and NIDDM in a high-risk ethnic group.

Guidelines of the National Heart, Lung, and Blood Institute Working Group on Blood Drawing, Processing, and Storage for Genetic Studies

The guidelines presented here are intended for epidemiologic investigators who wish to store blood samples for genetic studies, either by extracting DNA directly from white blood cells or from immortalized cell lines. Recommended procedures for blood drawing and for processing samples are described. Protocols for freezing and storage of both white blood cells and extracted DNA are provided. Although the extraction of DNA and immortalization of cell lines should be conducted in collaboration with an experienced laboratory, a summary of the available methods with appropriate references is given, and a method for the extraction of DNA from clotted blood is referenced. In addition, criteria for selecting study subjects for whom immortalized cell lines are preferable to merely extracting DNA are presented. Finally, the use of alternative sources of genetic material, including cheek swabs and dried blood spots, is described briefly.

Compositional differences of LDL particles in normal subjects with LDL subclass phenotype A and LDL subclass phenotype B

A predominance of small LDL particles (subclass phenotype B), as determined by gradient-gel electrophoresis is found among patients with myocardial infarction. Despite physical differences in phenotype A and B particles, differences in lipid composition of particles in these phenotypes have yet to be reported in an unselected population of males and females. The present study used lipid/apoB ratios to analyze the amount of lipid per LDL particle, isolated by density-gradient ultracentrifugation, in 70 healthy subjects. Relative to apoB, the LDL particles from phenotype B subjects were found to contain less free cholesterol (0.391 +/- 0.05 versus 0.465 +/- 0.05; mean +/- SD; P < .001), phospholipid (1.26 +/- 0.2 versus 1.43 +/- 0.2; P < .001), and cholesteryl ester (1.97 +/- 0.1 versus 2.11 +/- 0.2; P < .001) than particles from phenotype A subjects. The amount of triglyceride per LDL particle did not differ between the two phenotypes (0.410 +/- 0.1 versus 0.406 +/- 0.1; P = NS) despite higher plasma triglyceride levels in the phenotype B subjects. LDL size and buoyancy were positively correlated with particle free cholesterol, phospholipid, and cholesteryl ester but not with particle triglyceride. These data suggest that the physical properties of LDL from subjects with phenotype A and B reflect their lipid composition. The compositional differences between LDL particles of the two phenotypes may provide new insight into the increased risk of myocardial infarction in subjects with small, dense LDL.

Small, dense low density lipoproteins, the insulin resistance syndrome and noninsulin-dependent diabetes

The role of small, dense LDL as a risk factor for coronary heart disease is becoming well established. Several atherogenic mechanisms have been proposed to explain this relationship, including the association of small, dense LDL with the insulin resistance syndrome. A variety of studies, including large epidemiologic investigations, have now demonstrated that a predominance of small, dense LDL is one of the constellation of interrelated risk factors that characterize the insulin resistance syndrome. Recent results using a multivariate factor analysis approach have also shown that this set of risk factors can be simplified into three or four uncorrelated components that may provide new insights into the mechanisms underlying the syndrome and its association with disease. Furthermore, a number of cross-sectional studies and one prospective study have shown that small, dense LDL is a risk factor for noninsulin-dependent diabetes itself, and that this association may also be attributable to the insulin resistance syndrome.

Plasma triglyceride level is a risk factor for cardiovascular disease independent of high-density lipoprotein cholesterol level: a meta-analysis of population-based prospective studies

OBJECTIVES

Despite nearly 40 years of research, the role of plasma triglyceride as a risk factor for cardiovascular disease remains elusive. The objectives of the present study were to quantify the magnitude of the association between triglyceride and cardiovascular disease in the general population, and to determine whether this relationship is independent of high-density lipoprotein (HDL) cholesterol, using the semi-quantitative techniques of metaanalysis.

METHODS AND DESIGN

Seventeen studies were selected for the analysis based on published reports of population-based, prospective studies, including 46413 men and 10864 women. To insure comparability, only studies reporting the association between fasting triglyceride levels and incident cardiovascular endpoints were included. Using standard meta-analysis calculations, relative risks (RR) and 95% confidence intervals (CI) were calculated and standardized with respect to a 1 mmol/l increase in triglyceride. Multivariable-adjusted RRs were determined for the six studies in men and two studies in women that reported adjustments for HDL cholesterol.

RESULTS

For men and women, the univariate RRs for triglyceride were 1.32 (95% Cl 1.26-1.39) and 1.76 (95% Cl 1.50-2.07), respectively, indicating an approximately 30% increased risk in men and a 75% increase in women. Adjustment of HDL cholesterol and other risk factors attenuated these RRs to 1.14 (95% Cl 1.05-1.28) and 1.37 (95% Cl 1.13-1.66), respectively, which were still statistically significant values.

CONCLUSION

Based on combined data from prospective studies, triglyceride is a risk factor for cardiovascular disease for both men and women in the general population, independent of HDL cholesterol. These finding demonstrate the necessity for clinical trials to evaluate whether lowering plasma triglyceride decreases the risk of cardiovascular disease.

Genetic and environmental influences on insulin levels and the insulin resistance syndrome: an analysis of women twins

Multiple factors may determine insulin resistance and the insulin resistance syndrome. The contributions of genes and environment to the distribution of fasting insulin levels and to the associations of fasting insulin with elements of the syndrome were evaluated in the second examination of the Kaiser Permanente Women Twins Study (Oakland, California, 1989-1990). Subjects included 556 white women (165 monozygous twin pairs, 113 dizygous pairs; 455 women with normal glucose tolerance, 75 with impaired glucose tolerance, and 26 with non-insulin-dependent diabetes by World Health Organization criteria). The intraclass correlation coefficients for log fasting insulin for monozygous and dizygous twin pairs were 0.64 and 0.40, respectively. After adjustment for age, behavioral factors, and body mass index, the estimated classic heritability was 0.53 (p = 0.003). Commingling analysis of fasting insulin indicated the presence of four distributions (p < 0.001), consistent with at least one, and perhaps two, genes influencing this trait. In an unmatched multiple regression model among women from monozygous twin pairs only, log fasting insulin was independently associated with body mass index (p < 0.0001), waist/hip ratio (p = 0.02), and glucose intolerance (p = 0.04), but not with triglycerides, high density lipoprotein cholesterol, or hypertension. After removal of genetic influences by analysis of monozygous intrapair differences, only body mass index (p < 0.0001) remained independently related to fasting insulin. The authors conclude that, in addition to significant genetic influences on fasting insulin, environmental or behavioral factors (particularly nongenetic variation in obesity) are important determinants of fasting insulin and the insulin resistance syndrome.

Deletion of the entire NF1 gene detected by the FISH: four deletion patients associated with severe manifestations

Genetic analysis of NF1 has indicated a wide diversity of mutations, including chromosome rearrangements, deletions, insertions, duplications, and point mutations. Recently, five severely affected individuals have been found by Kayes et Al. [1994] to have deletions encompassing the entire gene. These deletions were detected by quantitative Southern analysis. To simplify deletion detection, we have employed fluorescence in situ hybridization (FISH) using intragenic probes. Thirteen unrelated individuals with NF1 have been studied. Among six with severe manifestations, four have been found to have deletions detected by probes cFF13, cFB5D, cP5, yA43A9, yA113D7 and yD8F4. All four deletions patients have severe developmental delay, minor and major anomalies (including one with bilateral iris colobomas), and multiple cutaneous neurofibromas or plexiform neurofibromas which were present before age 5 years. FISH provides a simple and rapid means of identification of NF1 gene deletions and will allow more rigorous testing of the hypothesis that such deletions are associated with severe manifestations.