Data analysis of common dyslipidemias for better clinical interpretation calculations of reference or cut off levels and relative prevalence

Abstract

Dyslipidemias are characterized by abnormal concentrations of cholesterol and triglycerides in plasma lipoproteins, typically resulting from disorders in lipoprotein metabolism. In this study, we are collecting clinical biochemistry data related to dyslipidemia from hospital. The data are analyzed and interpreted clinically by dividing them into gender and age-specific groups. This partitioning helps determine (a) proportion of patients with abnormal lipid profile, (b) gender-based differences, (c) reference intervals, and (d) the distribution patterns among the various subgroups. Lipid profiles were categorized into normal, isolated increases or decreases, and combined increases or decreases of four lipid components: serum triglycerides, total cholesterol, LDL-cholesterol, and HDL- Cholesterol. Our analysis includes assessing correlations, distributions, and variance to establish reference levels and the relative prevalence of common dyslipidemias observed in this hospital population. Data analyses of 15,603 lipid profile samples were done. Out of total 15,603 samples; there were 97 male samples and 86 female samples <18 years of age. There were 369 samples between age group 18 – 25 years. Out of which, there were 210 male samples and 159 female samples. The number of samples between 26 – 50 years was 5168. Out of which, there were 3247 male samples and 1921 female samples in the age group 26 – 50 years. The number of samples in age group >50 years was 9883. Out of which, there were 5925 male samples and 3958 female samples. Approximately 84% of general population in age group between 26 – 50 years, covered in this study have abnormalities in at least one of the lipid profile parameters analyzed. Most prevalent dyslipidemia is Isolated HDL- C decrease in males and females of all age groups except 26 – 50 years. In 26 – 50 years males, the most prevalent dyslipidemia is Isolated Serum triglyceride increase. But in females, Combined LDL-C and Total Cholesterol increase is the most prevalent dyslipidemia. Triglycerides >885mg/dL is the rare form of dyslipidemia. The proportion of patients with dyslipidemia is highest in males of 26 – 50 years. The healthiest lipid profile was observed in the age group females <18years. The study employed statistical and graphical methods, including Histograms, Box-Whisker plots, and Normal Q-Q plots, to analyze and classify lipid profile parameters such as Total Cholesterol (TC), Serum Triglycerides (TG), HDL Cholesterol (HDL-C), and LDL Cholesterol (LDL–C). Histograms and Normal Q-Q plots revealed a positive skew in Total Cholesterol and Triglyceride distributions, and Box-Whisker plots helped identify outliers, which were furthercategorized into near and far outliers. These graphical tools were used not only for statistical interpretation but also to clinically differentiate common, multifactorial dyslipidemias from rare, severe monogenic forms. Correlation analysis showed that triglycerides positively correlated with glucose, Total Cholesterol, LDL–C, TSH, and BMI in adult males. While in females, Triglycerides positively associated with liver enzymes AST and ALKP, possibly due to sex-related hormonal differences. Triglyceride was inversely correlated with HDL – C in both sexes. Total Cholesterol positively correlated with LDL–C. The reference intervals for various lipid profile analytes across different gender and age groups were analysed. Due to the positive skewness of the data, the 90% reference interval calculation was deemed the most appropriate. For example, In female individuals aged 18 to 25 years, the 90% reference interval values were as: Total Cholesterol: 119.83 mg/dL (lower limit) to 235.55 mg/dL (upper limit); Serum Triglycerides: 32.75 mg/dL to 170.67 mg/dL; LDL Cholesterol: 62.08 mg/dL to 149.35 mg/dL; HDL Cholesterol: 23.59 mg/dL and 61.23 mg/dL These values reflect the range within which 90% of the population in this subgroup is expected to fall. We had collected second set of fasting, 2-hr post breakfast and 2-hr post lunch blood samples (n=312) from healthy individuals of various age groups. Samples were analysed by partitioning and by comparing fasting, 2-hr post breakfast and 2-hr post lunch lipid data. Regarding fasting versus postprandial measurements, Total Cholesterol showed no significant difference, there were Significant mean difference between fasting and post lunch Serum triglyceride in all gender-age groups. HDL-C showed no significant variation with meal timing, whereas LDL-C demonstrated significant changes in both sexes aged 26–50 years. The study underlines the importance of age- and gender-specific reference ranges and shows how graphical and statistical techniques can effectively classify and analyse lipid disorders.