Search

Several methods of transformation and covariate adjustment have been applied to the Collaborative Lipid Research Clinics Program Family Study data to facilitate analysis of lipid levels of examinees of different sex and age groups. After several exploratory analyses, cholesterol, high density lipoprotein cholesterol, and low density lipoprotein cholesterol were logarithmically transformed, and triglycerides were transformed by the power -0.25. Two types of covariate adjustment procedures, the regression and Z score methods, were used. A modified regression method was developed and was found to be preferable to both simple cubic regression and the Z score method on theoretical and empirical grounds. Refinements in this method to correct for change in variance with age, and for the effect of socioeconomic status, seasonality, and anthropometric measures were made. Methodological issues connected with the transformation and adjustment procedures are discussed.

Plasma total cholesterol, low-density lipoprotein (LDL), and high-density lipoprotein (HDL) measurements on 402 individuals in 62 randomly selected families from the Columbia Medical Plan population were used to select the "best" model among a series of multifactorial models using the maximum likelihood method described by Lange et al [1976]. These models included both genetic and nongenetic components of variance. The most parsimonious model for each trait was selected and examined using a goodness-of-fit statistic designed by Hopper and Mathews [1982] to test the assumptions of this technique. A simple additive genetic model was the most plausible for all three measurements, suggesting a strong role for genetic factors in determining lipid and lipoprotein levels in these data. Goodness-of-fit statistics for these models were examined and showed little evidence of deviation from the assumption of multivariate normality within pedigrees. This approach of selecting the most parsimonious model among a series of competing models and then assessing its goodness-of-fit has many applications in studying familial aggregation of quantitative traits.

Familial aggregation of plasma lipid levels among 676 black juveniles, aged less than 20 years and residing in 202 households in East Baltimore, Maryland, was assessed by the intraclass correlation coefficient of residuals from sex-specific cubic regressions on age. Significant aggregation of both plasma cholesterol (r = 0.23, p less than 0.005) and plasma triglyceride (r = 0.27, p less than 0.005) levels were found. In addition, no effect on familial aggregation was found for each of several measures of family structure: number of juvenile household members, closeness in age, all children over 9 years of age, or shared surname. Two methods of classifying juvenile household size were compared. An association between classification method and juvenile household size was demonstrated (p less than 0.005) for the distributions of households and of children. The intraclass correlation coefficients for plasma cholesterol and plasma triglyceride for households classified by number of juveniles were higher than the coefficients for the house grouped by the number of juveniles admitted to the study, when compared for each classification number, 2 through 4. Evidence for an effect on familial aggregation of the socioeconomic status was shown by a comparison of intraclass correlation coefficients from households in East Baltimore with those from white and black households from the Columbia Medical Plan which have higher levels of occupation and education. Higher correlations were shown for the groups with higher socioeconomic status, regardless of race.

Low socioeconomic status, inner city Black adults, aged 20 to 49 yr (24 males and 45 females), were randomly selected from East Baltimore, MD to study plasma lipid and lipoprotein levels. Several factors known to affect these levels also were examined: dietary intake, alcohol intake, degree of obesity (measured by body mass index), physical activity level, smoking, and hormone use. Compared to women, the men consumed 9.3 more calories/kg body weight (p less than 0.005), 273 mg more cholesterol/day (p less than 0.005), and 7% fewer calories as sucrose (p less than 0.01). The P/S ratio of both their diets was 0.5. The men also had a lower body mass index than the women (23.9 kg/m2 versus 29.0; p less than 0.001). Mean lipid and lipoprotein levels were similar in the men and women. However, the men's total cholesterol (167 mg/dl) and low density lipoprotein cholesterol (94 mg/dl) levels were lower than those of adult Blacks in other studies, while the levels of the East Baltimore women were similar to those in other studies. For women, body mass index and high density lipoprotein cholesterol were negatively correlated (p less than 0.01); triglycerides and oral contraceptive use were positively correlated (p less than 0.01). None of the factors studied explained the relatively low total cholesterol and low density lipoprotein cholesterol levels in these inner city Black adult men.

We studied familial aggregation patterns of plasma cholesterol and triglyceride concentrations in a sample of 242 families in a prepaid medical care setting. Three analytic methods of progressive sophistication--case finding, correlation, and multiple regression--were used. Case-finding analyses detected only slight association of elevated lipid levels in parent-offspring pairs. The results suggest that when probands are randomly selected from a free-living population, the value of screening families in which one member is shown to have elevated lipid levels is limited mainly to identifying "hypercholesterolemic" offspring of "hypercholesterolemic" fathers. Correlations of plasma lipid levels were significant for various pairs of first-degree relatives but negligible for spouses. Multiple regression also supported the predictive value of family members' lipid levels. Taken together, the results of these analyses provide strong evidence for familial aggregations of plasma cholesterol and lipid levels. The distinction of genetic and environmental contributions is a priority for future research.

The measurement of plasma total cholesterol as a screening test for an increased plasma level of low density lipoprotein cholesterol (greater than age- and sex-adjusted 95th percentile) with a normal plasma triglyceride level (type IIa hyperlipoproteinemia) was assessed in 1325 young subjects, aged 6 to 19 years, from seven North American populations during the years 1972-1976. When the age- and sex-specific 95th percentile for total cholesterol was used as the cutpoint for hypercholesterolemia, almost one third of young subjects with type IIa hyperlipoproteinemia were undetected (false negatives), and 40% of the test results were falsely positive. Ninety-eight per cent of all non-type IIa subjects were correctly identified. Lowering the percentile cutpoint decreased the false negative results but increased the false positive results; increasing the cutpoint had an opposite effect. Females had higher percentile cholesterol values than males; when extrinsic, arbitrary cholesterol cutpoints (175-210 mg/100 ml) for screening were used, there were fewer false negative but more false positive results in the females than in the males. When the prevalence of type IIa hyperlipoproteinemia was increased from 5 to 50%, the false positive tests decreased from 40 to 2%. A two-step screening for hypercholesterolemia did not improve efficiency. The use of plasma total cholesterol as a test for type IIa hyperlipoproteinemia in a general population results in a relatively large number of false positive and false negative tests.

From 1972 through 1975 a study of plasma lipid levels was conducted at the Columbia Medical Plan (a prepaid group practice in an upper-middle-class suburban community) in accordance with nationally standardized interview, blood-drawing and laboratory procedures of the Lipid Research Clinics Program. Data were obtained from a large group of volunteers from the plan as well as from subjects selected by random sampling from membership rolls. Of 2,591 fasting, nonpregnant adults (age greater than or equal to 20 years), 825 were volunteers and 1,766 were randomly selected subjects. Analysis of the plasma total cholesterol values indicated a possible association of volunteer status and higher educational levels with a lower plasma total cholesterol level. Age- and sex-specific comparisons confirmed this finding, although the magnitude of the differences was quite small from the standpoint of clinical risk. The data suggest that even within educational strata, self-selection for cholesterol screening was associated with a lower cholesterol level.

High density lipoprotein cholesterol data on a population-based random sample of 858 white and 72 black probands and their 3935 white and 205 black relatives were collected from nine North American clinics using a common protocol and standardized methodology. Familial associations were examined within clinics for whites and pooled across clinics for blacks. The influence of covariates and varying family size on correlations was examined using several sets of transformed and adjusted values and a variety of weighting schemes. Parent-offspring and sibling correlations were significant in most cases, but spouse correlations were not, suggesting a stronger influence of shared genes than shared environment on high density lipoprotein cholesterol. Adjustment for covariates tended to weaken the correlations, but the effect of variable family size was imperceptible. Although pairs involving pediatric offspring or siblings tended to show higher correlation than their adult counterparts, the differences were not significant. All correlations except father-daughter and brother-brother were homogeneous across clinics in whites. There was no asymmetry in parent-child correlations by the sex of the offspring, but the pooled mother-child correlation was significantly higher than father-child values, suggesting a possible maternal influence on high density lipoprotein cholesterol. No heterogeneity in correlations in high density lipoprotein cholesterol was detected between blacks and whites except for mother-son pairs.

The efficiency of screening for dyslipoproteinemias associated with hyperlipoproteinemia or with hypolipoproteinemia was examined in 8449 white examinees from the Lipid Research Clinics Prevalence Study. A two-stage lipid screening approach was used. A positive screening test for hyperlipidemia was defined as an elevated level of plasma total cholesterol or triglyceride at both visit 1 and visit 2. A positive screening test for hypolipidemia was defined as a reduced plasma total cholesterol level at both visits. The syndromes of dyslipoproteinemia were defined according to the lipoprotein pattern present at visit 2. When hyperlipidemia was used as a screening tool, generally only about one-half or less of the cases of hyperlipoproteinemia were detected (high proportion of false-negative results). Of all the hyperlipidemic participants, more than 80% had hyperlipoproteinemias (type I, IIa, IIb, III, IV, or V), and such lipid screening therefore provided a lower proportion of false-positive than false-negative results. Most of the participants (greater than 98%) without hyperlipoproteinemia were correctly identified (true negatives). The efficiency of screening for hypolipoproteinemia was in general poorer than that found for hyperlipoproteinemia. Our results were found to be related, in part, to the marked regression to the mean of plasma cholesterol and triglycerides at the extremes of the distributions and to the wide variety of dyslipoproteinemias that can be present within a given range of plasma cholesterol or triglyceride values. The results emphasize the importance of measuring plasma lipoproteins in the patient who is being evaluated for dyslipoproteinemia.

Familial associations of total cholesterol, low density lipoprotein cholesterol, triglycerides, and very low density lipoprotein cholesterol were examined in a population-based random sample of 858 white and 73 black probands and their 4,027 white and 245 black relatives from nine North American Lipid Research Clinics. Correlations among biologic relatives were highly significant for total cholesterol and low density lipoprotein cholesterol and to a lesser extent for triglycerides and very low density lipoprotein cholesterol in whites. Correlations for spouses, however, were not significant, suggesting a stronger influence of genes than shared environment in the determination of these traits. Homogeneity of familial correlations across age strata, clinics, and racial groups was examined. In general, correlations were homogeneous across age strata and clinics, and there was no asymmetry in parent-offspring correlations by the sex of the parent or offspring. Racial differences in correlations were not significant except in four of 32 comparisons, with blacks showing weaker correlations than whites in those instances.