Re-measurement of plasma cholesterol, TG, and HDL-C concentrations in a subset of participants with plasma available at all 3 exam cycles showed excellent correlations with original measurements over a broad range of values for each lipid fraction, at each exam cycle. For the comparison of re-measured lipids with original values from the earliest exam cycle of this series of cycles (Exam 5) correlation coefficients were 0.985 for cholesterol, 0.997 for TG, and 0.948 for HDL-C, all significant at P <0.0001 (Supplementary Figure
). The average intra-assay coefficients of variation when comparing the original measurements with these re-measurements were: for total cholesterol, 5% (examination 5), 3% (examination 6), and 3% (examination 7); for HDL-C, 7% (examination 5), 5% (examination 6), and 4% (examination 7); and for TG, 6% (examination 5), 3% (examination 6), and 4% (examination 7).
The clinical characteristics of the study sample at each of the three examinations are shown in .
Clinical Characteristics at Examination by Sex at Cycle 5, 6 and 7
Trends in Lipid Levels, Fasting Glucose, and BMI between 1991 and 2001
Total cholesterol levels did not change over the three examinations in either sex (). In contrast, HDL-C levels increased significantly over the examination cycles, also when adjusting for potential confounders including total cholesterol levels. This was accompanied by decreases of TG, which also were significant in age- and multivariable-adjusted models. These trends were essentially identical when using log-transformed TG in the models (data not shown).
Age- and Multivariable-Adjusted Mean Levels* of Lipids, Fasting Glucose and Body Mass Index By Examination Cycle
In this same time period BMI significantly increased, in age-adjusted as well as in multivariable-adjusted models (). Fasting glucose levels increased in men, but not in women.
The cumulative distribution functions of multivariable-adjusted levels of total cholesterol, HDL-C, and TG for the three examinations are shown in (men), and (women). In general, these curves show a relatively similar sized shift in the distribution of each of these lipid fractions throughout the range of measured values from the earliest exam (Offspring 5) to the most recent (Offspring 7) with the distribution of values at the intermediate exam (Offspring 6) somewhat variable in relation to the earliest and most recent exam.
Figure 1 Cumulative Distribution Functions of Multivariable-Adjusted Levels of Total Cholesterol (Panel A), High-Density Lipoprotein Cholesterol (Panel B), and Triglycerides (Panel C) at Examination 5, 6 and 7 in men not treated with lipid drugs and without cardiovascular (more ...)
Figure 2 Cumulative Distribution Functions of Multivariable-Adjusted Levels of Total Cholesterol (Panel A), High-Density Lipoprotein Cholesterol (Panel B), and Triglycerides (Panel C) at Examination 5, 6 and 7 in women not treated with lipid drugs or hormone-replacement (more ...)
In a larger sample including individuals with prevalent CVD, lipid therapy or hormone replacement therapy at any of the three examinations, there were also significant and substantial increases in HDL-C and decreases in TG over time in both men and women, in age- and multivariable-adjusted analyses (Supplementary Table 1
). BMI significantly increased over the same time period, and in contrast to the primary analysis, there were significant decreases in total cholesterol levels in this larger sample.
Trends in Proportions of Individuals in Lipid Categories
Consistent with analyses of continuous variables, both the proportions of individuals with low HDL-C and the proportions of individuals with high TG decreased significantly in age- and multivariable-adjusted analyses in both men and women (). However, the categories of LDL-C levels did not change significantly over the three examinations.
Proportions of Participants in Lipid-Related Categories* By Examination Cycle
In the larger sample including individuals with prevalent CVD, lipid therapy or hormone replacement therapy at any of the three examinations, the proportions of individuals with low HDL-C or high TG changed in the same manner as in the primary analysis (Supplementary Table 2
). Also, in this sample the proportions of individuals on prescribed lipid therapy increased significantly in age- and multivariable-adjusted analyses in both men and women, and the categories of LDL-C levels changed significantly over the three examination periods.
Conjoint Changes in HDL-C and TG between 1991 and 2001
The age-adjusted Spearman correlations between change in HDL-C and change in TG (delta HDL-C and delta TG) from examination 5 to examination 7 were −0.35 in men and −0.31 in women, (P<0.0001 for both correlations). When study participants were grouped into four categories by the pattern of change in HDL-C and TG between examinations 5 and 7, the largest category of change for both men and women, consisted of participants with an increase in HDL-C and a decrease in TG (). Furthermore, this category had the smallest increases of BMI between examination 5 and 7 in both men and women in multivariable-adjusted analyses (P<0.005 for a comparison of the group with a decrease in TG and increase in HDL-C with the three other groups, except in women with an increase in both HDL-C and TG where P=0.035).
Age- and Multivariable-Adjusted Mean Changes* of Lipids, Fasting Glucose and Body Mass Index Between Examination Cycle 5 and 7 by Pattern of Change† in High-Density Lipoprotein Cholesterol and Triglycerides
A comparison of data from diet questionnaires at examination 5 and 7 did not show any significant changes in caloric intake between the various lipid groups. Although in multivariate-adjusted analysis there was a larger but non-significant decrease in calorie intake in men with a decrease in TG and increase in HDL-C compared to an index group with both a decrease in TG and HDL-C, this change was not significant (P=0.64) and virtually no differences were apparent in women.
In secondary analyses where we repeated all models with waist circumference instead of BMI, the reciprocal trends of HDL-C and TG were almost identical as in the main analyses (data not shown).