Skip to main content

Long-term intra- and inter-individual biological variation of serum lipid of HIV-infected and uninfected men participating in the Los Angeles Multi-Center AIDS Cohort Study (MACS)

Abstract

Background

To assess the long-term biological coefficient of variation within individuals (CVI) and between individuals (CVG), effect of aging and cholesterol lowering drugs on blood levels of lipids in HIV-1-infected and -uninfected men.

Methods

Bloods were analyzed every six months over 17 years for total cholesterol (TC), triglycerides (TGs), high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol (LDL-C) in 140 HIV-uninfected (38–66 years old) and 90 HIV-treated infected (48–64 years old) white Caucasian men to examine CVI, CVG, and the effect of cholesterol lowering drugs (CLDs) on lipid levels, and estimated changes per year of biomarkers.

Results

With exception of HDL-C, the long term CVI compared with CVG were higher for serum levels of TC, TGs, and LDL-C in both HIV-1 infected and uninfected men not taking CLDs. Excluding results of TGs in HIV positive men, the CVI compared with CVG were lower for serum levels of TC, HDL-C, and LDL-C in both groups not taking CLDs.

There were significant (p < 0.05) differences in the median serum values of lipid biomarkers among 77 HIV negative men taking and 63 not taking CLDs. Also, with exception of HDL, there were significant (p < 0.05) differences in the median values of TC, TGs and LDL-C among 28 HIV positive men taking or not taking CLDs.

Conclusion

Long term CVI and CVG of biomarkers will be useful for monitoring antiviral therapy side effects on lipid profiles in HIV-infected men. CVI of HIV-infected men for TC, TGs, HDL, LDL were higher significantly than CVI of HIV-uninfected men. Interestingly the long term CVI were higher than CVG for the men, who were on CLDs compared to men not on CLDs. The long-term pattern of CVI and CVG of lipid markers in both HIV-infected and uninfected men on CLDs differed from their short-term pattern.

Background

Human immunodeficiency virus type 1 (HIV-1) infection, chronic inflammation and treatment with highly active antiretroviral therapy (HAART) has been associated with changes in lipid metabolism, and profiles of blood levels of total cholesterol, triglycerides, HDL-C and LDL-C [1,2,3]. These changes in the profile of blood levels of lipids is one of the risk factors for cardiovascular and cerebrovascular diseases in HIV-1 infected individuals [4, 5].

An array of sources including pre-analytical, analytical and post-analytical factors affect the outcome of measurement of laboratory biomarkers. To the extent possible, researchers should better understand and eliminate all sources of potential measurement error that can lead to biased findings so that the true biological variations can be detected.

Blood analytes can vary throughout the lifespan of each individual. Some of the variabilities are expected from biological cycles that occur within individuals, while between individual variations may account for another source of variation [6]. Some of the sources such as patient preparation, blood collection techniques, handling, transporting, centrifugation, and specimen storage are controllable, while other issues such as age, ethnicity, and biological sex are not subject to manipulation [6].

Biological changes for many biomarkers occur throughout the aging lifespan as well as in menopause in women. In addition, certain analytes have diurnal, monthly, or seasonal biological cycles or rhythms. However, cyclic variations will not be a major problem for some biomarkers [6,7,8].

The interpretation of patients’ blood test results using population reference values assumes that the biological variability of the population reference values is similar to the target population [6, 9, 10].

Certain blood biomarkers show a high degree of biological variation (BV) among healthy and unhealthy individuals as an outcome of broad homeostatic set points, even within a reference population that is regarded as ‘healthy’. This individuality between healthy populations makes it a challenging task for laboratorians to generate a reference interval for a blood analyte that represents a ‘healthy’ reference range for all individuals of a given community. Meanwhile, several individuals may have laboratory test values that are very unique and normal for them but still lie outside (5%) or inside ranges of population-based references [11,12,13].

Repeated quantification of a biomarker obtained by a longitudinal study of healthy and unhealthy subjects may be more desirable to use for evaluation of laboratory results of patients than a single measurement of an analyte associated with population-based references. In the case of repeat measurements of an analyte, the individuals will have their own baseline for the analyte and any changes in value from their baseline result may be associated with the patient’s illness and prognosis [14].

Knowledge of BV of intra-individual coefficients of variation (CVI) and inter-individual coefficients of variation (CVG) is not only essential for the longitudinal evaluation of blood biomarkers, but it is also important for calculation of desirable quality specification such as assay precision, bias, and total error of a blood biomarker [6].

This study assesses the long-term effect of HAART and cholesterol lowering drugs (CLDs) on biological coefficients of variation, mean values, and individual aging on the basic blood lipid panel of total cholesterol, triglycerides, HDL-C, and LDL-C in HIV-infected and uninfected white Caucasian men over a period of seventeen years.

In addition, we hypothesize that the long-term biological variation of blood lipids in HIV-1 infected and uninfected men on CLDs differ from that of individuals not on CLDs as well as from short term observations.

We accomplished the assessment of short- and long-term BVs by examining the biological coefficient of variation within each individual (CVI) and variation between individuals (CVG) of the lipid biomarkers in a population of HIV-infected men on HAART and HIV-uninfected men, both groups on or not on cholesterol lowering drugs.

Material and Methods

Study Participants

We investigated routine blood levels of lipid panels in HIV-infected and HIV-uninfected white Caucasian men excluding participants with cancers, viral hepatitis B or C, cytomegalovirus (CMV) infections, diabetes or kidney disease. In addition, for the minimization of inter-racial variability, other ethnicities were excluded from this study.

The HIV-uninfected group consisted of 140 white Caucasian men with age ranges between 38–66 and a mean of 51 years old at the start of the study period, all of whom were documented to be HIV-1 sero-uninfected at every study visit (six-month intervals) over the course of study. Seventy-seven out of 140 men were on cholesterol lowering drugs (statins) such as Atorvastatin, Fluvastatin, Simvastatin, Rosuvastatin, Lovastatin, Pitavastatin (n = 53) or a combination of statins with the other CLDs drugs such as Gemfibrozil, Fenofibrate, Ezetimibe, or Niacin (n = 18), and fish oil, and herbal preparations (n = 6).

The HIV-infected group included 90 white Caucasian men 37–64 years old with a mean of 48 years old who were receiving HAART. Sixty-two out of 90 men at the start of the study were on CLDs such as statins (n = 45) or combination of statins with other CLDs drugs such as Gemfibrozil, Fenofibrate, Ezetimibe or Niacin CLDs (n = 17) during this longitudinal study.

All these individuals were men who have sex with men (MSM) participating in the Los Angeles (LA) center of the Multicenter AIDS Cohort Study (MACS) who self-reported as white Caucasian [15]. The average length of follow-up between a participant’s first and last visit was 17 years.

The institutional review board (IRB) for human studies at UCLA (University of California, Los Angeles) approved the protocols.

Blood collection and laboratory assays

After informed consent, an overnight fasting blood sample was collected, into one 8 ml SST BD vacutainer tube which was obtained every six months, between 8:00 am to 12:00 pm. Serum was separated within 2- 4 h of the draw and frozen at -80 °C for batch testing. Frozen serum samples for analysis of LIPID panel were shipped to the Heinz Chemistry and Nutrition Laboratory, University Pittsburgh (Pittsburgh, PA). The lipid panel (TC, TGs, HDL-C and LDL-C) were analyzed by chemistry Analyzer AU series (Beckman Coulter, CA, USA) based on the package insert. The methods and assays precision were as follows:

Total cholesterol was measured enzymatically and the intra-assay precision (N = 100) and total run were 0.5% and 0.9% for within run and 1.1% and 1.1% CV for total runs (less than 3% CV) for two samples with mean of 115.5 mg/dL and 252.2 mg/dL respectively. Assay sensitivity was 1 mg/dL. Adult reference ranged from 125–200 mg/dL.

Triglycerides was measured enzymatically and the intra-assay precision (N = 80) and total run were 0.64%, 0.49%, and 0.51% for within run and 1.65%, 1.41%, and 1.46% CV for total runs (less than 3% CV) for three samples with a mean of 89.4 mg/dL, 191 mg/dL, and 442 mg/dL respectively. The lower detection limit was 0.31 mg/dL. Reference ranges were less than 150 mg/dL.

HDL-Cholesterol was measured enzymatically and the intra-assay precision and total (N = 80) were 0.43%, 0.59%, and 0.60% for within run and 2.64%, 2.19%, and 2.12% CV for total runs (less than 3% CV) for three samples with mean of 38.85 mg/dL, 66.39 mg/dL, and 86.33 mg/dL respectively. The lower limit of detection was 1 mg/dL with a reference limit of 40 mg/dL.

LDL-Cholesterol was calculated by the Friedewald equation. The equation is valid if the Triglycerides level is less than 400 mg/dL. The formula is: “LDL = TC- (HDL + (Triglyceride/5)”.

In the equation above TC, LDL, and HDL represent total cholesterol, LDL-C and HDL-C respectively [16]. In cases when the triglycerides level was above 400 mg/dL, the LDL was directly measured. The reference limit was130 mg/dL.

Statistical analysis

Descriptive statistics of the serum markers that comprised the cholesterol, Triglycerides, HDL-C, and LDL-C, are summarized as means and absolute ranges for both intra- and inter-individual sets of values. Linear trends were estimated using generalized estimating equations (GEE). GEE may be conceptualized as nonparametric extensions to the generalized linear model for longitudinal data [17]. Correlations among the multiple visits from each subject are estimated from data without specifying the exact form of the correlation structure. For every one-unit increase in a covariate across the population, GEE tells us how much the average response would change.

The nested analysis of variance was used for calculation of coefficients of variation of intra-individual (CVI) and inter-individuals (CVG). CVI and CVG were calculated according to the approach used by Harris and Boyd [18]. All analytes were individually tested using highly precise automated analyzers by the clinical reference laboratories.

The Index of Individuality (II) is the simple ratio of the two biological components of variation: intra-individual to inter-individuals and is calculated using the formula CVI/CVG [7]. The II, as defined by Harris [19] assesses the usefulness of population-based reference values for interpretation of laboratory tests. If the II of a given analyte is greater than 1.4, then population-based intervals are useful whereas an II below 1.4 indicates decreased utility of population-based reference intervals. Analytes with an II less than 0.6 demonstrate (paradoxically) a high degree of individuality, making individual-based reference intervals more useful [19]. The Mann–Whitney Rank sum test was used for comparing results between HIV-uninfected and HIV- infected men and closely aged matched to men on statins and not on statins.

Data analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC). Graphs were created using SigmaPlot software version 14 (Jandel Scientific, San Rafael, CA 2018).

Statistical significance was set at the alpha = 0.05 level.

Results

Components of biological variation of blood lipid biomarkers

The CVI, CVG, and index of individuality (II), along with the overall means of TC, TGs, HDL-C, and LDL-C for short term of 1 year (3 visits) and long term of 17 years (34 visits) during follow up are presented for HIV-uninfected and HIV-infected men taking or not taking lowering cholesterol drugs (CLDs) in Table 1.

Table 1 Mean values, Intra-individual (CVI) and Inter-individuals (CVG) coefficient of variation, and index of individuality (II) of blood level of basic Lipid panel of HIV-1 infected and uninfected men

As described in the methods, it is considered appropriate to use population-based reference ranges when the Index of Individuality (II) of a given analyte is greater than 1.4. The indices of individuality for both HIV-uninfected and HIV-infected men in our study for the all markers for 1 year and 17 years follow up were less than 1.4.

The short-term percentage of CVI of TC, TGs, HDL-C and LDL-C were lower than CVG for participants on or not on CLDs in both HIV-infected and -uninfected participants.

The long-term percentage of CVI of TC, TGs, and LDL-C with exception of HDL-C (lower) were higher than CVG for participants on CLDs in HIV-infected and -uninfected participants.

The long-term percentage of CVI of TC, TGs, HDL-C and LDL-C were lower than CVG for HIV-uninfected participants, and the CVI of TC, and HDL-C were lower and TGs and LDL-C were higher than CVG for HIV-infected participants not on CLDs.

In addition, regardless of CLDs, CVI of HIV-infected men for TC, TGs, HDL-C, LDL-C were significantly higher than CVI of HIV-uninfected men (p < 0.05).

Mean and minimum–maximum blood level of lipid biomarkers

The long-term mean levels of serum total cholesterol, triglycerides, HDL-C, LDL-C (A-D) and their ranges (minimum–maximum values) over 34 visits for HIV-1 uninfected men with CLDs (n = 77) without CLDs (n = 63) and HIV-1 infected men with CLDs (n = 62) without CLDs (n = 28) are shown graphically for total cholesterol in Fig. 1 A-D, triglycerides in Fig. 2 A-D, HDL-C in Fig. 3 A-D and LDL-C in Fig. 4 A-D, respectively.

Fig. 1
figure 1

Mean values (black dot) and absolute ranges (error bars) for total cholesterol for HIV-uninfected participants who were on CLDs(A) or not on CLDs(B), vs. HIV-infected participants on CLDs(C), or not on CLDs(D). *absolute maximum value of marker filled outside range of X axis

Fig. 2
figure 2

Mean values (black dot) and absolute ranges (error bar) for triglycerides for HIV-uninfected participants who were on CLDs(A) or not on CLDs(B), vs. HIV-infected participants on CLDs(C), or not on CLDs(D). *absolute maximum value of marker filled outside range of X axis

Fig. 3
figure 3

Mean values (black dot) and absolute ranges (error bar) for HDL-cholesterol for HIV-uninfected participants who were on CLDs(A) or not on CLDs(B), vs. HIV-infected participants on CLDs(C), or not on CLDs(D). *absolute maximum value of marker filled outside range of X axis

Fig. 4
figure 4

Mean values (black dot) and absolute ranges (error bar) for LDL-cholesterol for HIV-uninfected participants who were on CLDs(A) or not on CLDs(B), vs. HIV-infected participants on CLDs(C), or not on CLDs(D). *absolute maximum value of marker filled outside range of X axis

Visual inspection of those figures shows that the mean values of the same analyte can differ greatly between individuals in both HIV-uninfected and HIV-infected groups.

The mean serum levels of total cholesterol for HIV-uninfected men were 189.9, and 193.4 mg/dL and for HIV-infected men were 184.2 and 175.1 mg/dL with and without CLDs (Fig. 1 A, C and Fig. 1, B, D) respectively. The majority of mean results of individual men for both groups were within the reference ranges (125–200 mg/dL).

Similarly, the mean, minimum and maximum of serum triglycerides levels for HIV- uninfected men with or without CLDs were 120.6 and 93.1 and for HIV-infected men with or without CLDs were 188.1 and 129.3 mg/dL respectively. The majority of mean results for both groups were lower than the reference limit of < 150 mg/dL (Fig. 2, A, C and Fig. 2, B, D).

The mean, minimum and maximum of serum HDL-C levels for HIV-uninfected men with or without CLDs were 50.8 and 55.2 mg/dL and for HIV-infected men with or without CLDs were 45.1 and 47.2 mg/dL respectively. The majority of mean results for both groups were higher than the reference limit of 40 mg/dL but mean results for 17 HIV-infected men were lower than the reference limit of 40 mg/dL compared to HIV-uninfected men. The lower level of HDL-C in HIV-infected men may be due to HIV infection or HAART. (Fig. 3 A, C, and Fig. 3, B, D).

The mean, minimum and maximum of serum LDL-C levels for HIV-uninfected men were 114.3, and 118.6 mg/dL and for HIV-infected men were 105.8 and 102.3 mg/dL respectively. The majority of mean results for both HIV-infected and HIV-uninfected men were lower than the reference limit of 130 mg/dL (Fig. 4, A, C, and Fig. 4, B, D).

Magnitude of blood lipid biomarkers changes per year

In order to evaluate direction and magnitude of changes in each marker from baseline over the course of seventeen years of follow up of HIV- uninfected and HIV-infected men who were not on CLDs, generalized estimating equations in SAS software were used to estimate the average change per year in values and corresponding p-values. The longitudinal analysis over the course of seventeen years demonstrated a statistically significant (p < 0.05) increase per year for blood levels of HDL-C and a decrease per year for TGs of HIV-uninfected men shown in Table 2.

Table 2 Average change per year in lipid panels of men not on CLDs over 17 years of follow-up

Comparison of blood levels of lipid biomarkers of men on and not on CLDs

The Whitney Rank sum test was used for comparing results of blood levels of lipid biomarkers between HIV-uninfected(n = 28) and HIV- infected men (n = 28). There were significant (p =  < 0.05) differences in the median values of total cholesterol (190, 173 mg/dL), triglycerides (82, 104 mg/dL) HDL-C (53.5, 43.8 mg/dL), and LDL-C (116, 102 mg/dL) in HIV-uninfected and infected respectively.

In addition, when we compared the results of TC, TGs. HDL-C, and LDL-C of HIV-uninfected men not on (n = 63) and men on (n = 63) CLDs we found a statistically significant (p =  < 0.05) difference in the median results of TC (190, 184 mg/dl), TGs (81, 100 mg/dL), HDL-C (53.9, 49.0 mg/dL) and LDL-C (116, 110 mg/dL) respectively. For comparison analysis, data were selected based on the age of the two groups. The median results were higher for TC, HDL-C, and LDL-C for men not on CLDs and lower for result of TGs in men without CLDs.

The median value of total cholesterol, triglycerides, HDL-C, and LDL-C of 28 HIV-infected men not on CLDs were 173 mg/dL, 104 mg/dL, 43.8 mg/dL and 102 mg/dL respectively and were statistically significantly (p < 0.05) lower for total cholesterol = 184 mg/dL, triglycerides = 133 mg/dL, and LDL-C = 107 mg/dL, and not significantly different for HDL-C = 43.9 mg/dL compared with 28 HIV-infected men on CLDs.

Discussion

There have been many published studies on short-term biological variation of serum TC, TGs, HDL-C, and LDL-C [20,21,22,23]. Short- and long-term longitudinal testing of blood levels of lipids can assist physicians in distinguishing changes in the blood levels of biomarkers over time and during the natural course of an illness in prevention, prognosis, and treatment decisions making.

We longitudinally examined the mean values, CVI, CVG, and II of serum levels of TC, TGs, HDL-C, and LDL-C of HIV-infected and -uninfected men who were or were not on CLDs every six months for 1 year and 17 years (Table 1).

The short-term mean values were higher for levels of TC, TGs and LDL-C and lower for HDL-C compared with the long-term mean values for both HIV-uninfected and infected men on CLDs. Therefore, the decreases in the mean values of TC, TGs, and LDL-C and increase of HDL-C would be the beneficial consequence of the long-term use of CLDs on blood levels of lipids.

Conversely the short-term mean values were lower for TC, TGs, HDL-C and LDL-C (except for HIV infected participants) compared with the long-term mean values for both HIV-uninfected and infected men not on CLDs. This increase could be due to an ageing effect in individuals on the lipid panel in our study.

The mean values TC, TGs, HDL-C and LDL-C of our study (on or not on CLDs) differed from published studies [24, 25] The selection of white Caucasian men taking or not taking CLDs could account for the differences from other studies.

Knowing intra- and inter-individual coefficients of variation of lipid biomarkers can help the investigator to explore the usefulness of conventional population-based reference values, calculation of reference change value (RCV), and the desirable quality specifications such as assay precision, bias, and total error of blood biomarkers.

The reference intervals will be more useful when the index of individuality (II) is higher than 1.4. A low or high index of individuality of analyte means that the analytes have noticeable or slight individuality respectively [6, 26]. When II is less than 0.6, reference values are limited in detection of unusual results. In contrast when II is more than 1.4 [26]. The index individuality (II) in our lipid markers study with exception of 1 year of HIV infected men not on CLDs were between 0.6 and 1.4. Our data also suggests that an II < 1.4 indicates population-based reference values will have very little diagnostic value, while those may still be a useful tool for patient monitoring purposes. We should always be careful when using population-based reference values to interpret patients’ test results.

The CVI compared to CVG for 1 year of follow up of HIV-infected and uninfected men not on CLDs (excluding TGs) were lower (Table1) and consistent with published short-term duration of 4 weeks; CVI and CVG of the published study was for TC (8.9%, 15.2%), and for TGs (16.8%, 21.1%) respectively [20]. CVI and CVG in a Hispanic Community Health Study/Study of Latinos (HCHS/SOL) from four sites (Bronx, Chicago, Miami and San Diego) with two-time blood collection (baseline and a month apart) were for TC (6.8%, 20.0%), TGs (18.7%, 66.8%), HDL-C (6.4%, 25.7%) and LDL-C (9.7%, 27.4%) respectively [27].

CVI and CVG of the HCHS/SOL study follow the same pattern of CVI < CVG from our short-term observation of men on CLDs and for short and long-term men not on CLDs of HIV-uninfected. Similar CVI < CVG but slight differences in the value of CVI and CVG data to our study which might be this due to differences in the subject population, number of blood draws, study duration, and number of study sites (site to site variability).

In our study the value of CVI and CVG of the blood lipid of HIV-infected men were higher compared with CVI and CVG of HIV-uninfected men not on CLDs. This higher value could be due to the long duration of treatment with HAART medications and their metabolic side effects, higher incidence of opportunistic infections, diet, and/or fat redistribution [4, 24, 25].

There was a significant difference between median values of TC (190 > 173), TGs (82 < 104), HDL-C (53.5 > 43.8) and LDL-C (116 > 102) for HIV-uninfected compared to HIV-infected men not on CLDs respectively. The same observation for HIV-infected participants on HAART of a decrease or little change the serum levels of HDL-C and increase the serum level of TC, LDL-C and TGs has previously been reported [4, 25, 28].

In addition, the median values of intra-individual coefficient of variation (CVI) of blood markers (excluding LDL-C) of HIV-infected men were significantly higher compared to HIV-uninfected men.

The median CVI of blood markers in healthy and unhealthy individuals (different illnesses) of HDL-C, LDL-C, cholesterol and triglycerides of published studies (Table 3) were lower compared with the mean values of CVI of our 1, and 14 year follow up study of HIV- infected and HIV-uninfected men not on CLDs [29].

Table 3 Median CVI (%) of healthy and chronic diseases in the Ricos C. et al. study and mean CVI (%) of HIV-uninfected and infected men in our study

The CVI data in the Table 3 is lower compared with this study. The differences in CVI values from short term observations may be due to differences in gender, ethnicity, lifestyle, age distribution, socioeconomic factors, investigation duration, frequency of blood draws, medications, and chronic inflammation as reported in HIV-infected individuals.

With the exception of HDL-C the pattern of the results of long-term for CVI and CVG (CVI > CVG) of the HIV-infected and uninfected white men who were on CLDs differed from that observed in the published study CVI < CVG [13, 27, 30].This difference probably could be due to the effect of cholesterol lowering drugs on the reduction of blood levels of lipids by numerous cholesterol lowering drugs such as statins, fibrate (lowering TGs drug), bile acid sequestrants, nicotinic acid, and proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitors (LDL-C lowering drug).

Regardless of HIV status, the reduction of lipids by CLDs could be greater between individuals than within an individual. The decrease blood levels of lipids caused by CLDs are independent of the reduction caused by factors such as diet, exercise, lifestyle, smoking, alcohol consumption and stressors, which differ between individuals.

Our data of 17 years follow up of HIV-infected and uninfected men not on CLDs, shows some similarity such as lower CVI and higher CVG, to the published study [13], but not for HIV-infected and uninfected men on CLDs, which show a higher CVI and lower CVG value. This demonstrates that short-term BV of some biomarkers may not be similar to the pattern of long-term BV.

In addition, the effect of aging on the blood biomarkers and estimated average change in values per year and corresponding p-values were calculated for HIV-infected and uninfected men. There was statistically significant (p < 0.05) increase per year for HDL and decrease per year for TGs for HIV-uninfected men not on CLDs and no significant changes for HIV-infected men (Table 2).

The limitation of our study was the absence of female participants, missing data for some follow up visits of participants, and using two different methods used for measurement of serum LDL-C. In addition, 28 HIV-1 infected men were not taking CLDs compared to 63 HIV-1-uninfected men in this study.

The strengths of this study, which originated from the MACS longitudinal study population, are the large sample sizes, single-sex, and single-ethnicity population, which maximizes our ability to evaluate biological variation of 17 years follow up of CVI and CVG for HIV-infected and HIV-uninfected men on and not on CLDs. Thus, applying these observations to non-white female groups may not be appropriate.

Conclusion

Published studies of blood biomarkers demonstrated that the within individual coefficient variation (CVI) was less than the between individual coefficient of variation (CVG). In our study, for those individuals not on CLDs, with the exception of triglycerides, the results follow the previously published pattern of CVI < CVG. However, for the individuals on CLDs, with the exception of triglycerides, our data emerge in the opposite direction (CVI > CVG) to the published studies [13, 20, 30]

We believe that this is the first study which shows that within individual coefficient variation (CVI) of individuals on long-term CLDs was greater than between individual coefficient of variation (CVG). The CLDs are more likely to effect on between individual biological coefficient of variation (CVG) than within individual biological coefficient of variation (CVI).

The mean values were significantly lower for the levels of TC, HDL-C and LDL-C and were higher for triglycerides in HIV-infected men compared with HIV-uninfected men.

The median of within individual coefficients of variation (CVI) of the lipid panel (LDL-C not significant) was significantly higher in HIV- infected men compared with CVI of HIV-uninfected men.

The important “take home” messages are: 1) Data of short-term biological variation of some biomarkers in healthy and unhealthy men are not similar to the patterns of long-term biological variation. 2) Biological variation (CVI and CVG) data for healthy and unhealthy men will not be similar depending on HIV status, HIV-1 infected men had significantly higher biological variation for lipid markers compared with HIV-1 uninfected men which it may be due to HIV- infection and HAART. 4) individuals on CLDs, with exception of triglycerides, experienced different patterns of CVI > CVG.

The results of this study will be useful for predicting the results of therapy as well as for providing a guide for intra- and inter-individual coefficient variations, identifying the limitations of conventional population-based reference values, and establishing quality goals for different blood biomarkers in both healthy and diseased men. Similar studies in women, and different ethnicities are recommended.

Availability of data and materials

The datasets used for analysis for this study will be available from corresponding author on reasonable request.

Abbreviations

BV:

Biological Variation

CLIA:

Clinical Laboratory improvement Amendment

CLDs:

Cholesterol lowering drugs

CMV:

Cytomegalovirus

CVG :

Biological coefficient variation of inter-individual

CVI :

Biological coefficient variation of intra-individual

HDL-C:

High-density lipoprotein cholesterol

HIV-1:

Human immunodeficiency virus type 1

II:

Index of individuality

LDL-C:

Low-density lipoprotein cholesterol

TC:

Total cholesterol

TGs:

Triglycerides

References

  1. Ji S, Xu Y, Han D, Peng X, Lu X. Brockmeyer NH, Wu N, Changes in Lipid Indices in HIV+ Cases on HAART. Biomed Res Int. 2019;2019(5):2870647. https://doi.org/10.1155/2019/2870647.eCollection2019.PMID:30868068Syndr.2000;23:35-43 [PubMed: 10708054].

    Article  PubMed  PubMed Central  Google Scholar 

  2. Riddler SA, Li X, Chu H, Kingsley LA, Dobs A, Evans R, Palella F, Visscher B, Chmiel JS, Sharrett A. Longitudinal changes in serum lipids among HIV-infected men on highly active antiretroviral therapy. HIV Med. 2007;8(5):280–7. https://doi.org/10.1111/j.1468-1293.2007.00470.x PMID: 17561873.

    Article  CAS  PubMed  Google Scholar 

  3. Waters DD, Hsue PY. Lipid Abnormalities in Persons Living with HIV Infection. Can J Cardiol. 2019;35(3):249–59. https://doi.org/10.1016/j.cjca.2018.11.005 Epub 2018 Nov 15.

    Article  PubMed  Google Scholar 

  4. Funderburg NT, Metha NN. Lipid Abnormalities and inlammation in HIV infection. Curr HIV/AIDS Rep. 2016;13(4):218–25. https://doi.org/10.1007/s11904-016-0321-0 PMID: 27245605.

    Article  PubMed  PubMed Central  Google Scholar 

  5. Triant VA, Lee H, Hadigan C, Grinspoon SK., Increased acute myocardial infarction rates and cardiovascular risk factors among patients with human immunodeficiency virus disease., J Clin Endocrinol Metab. 2007 ;92(7):2506–12. doi: https://doi.org/10.1210/jc.2006-2190. Epub 2007 Apr 24. PMID: 17456578

  6. Fraser CG, editor. Biological Variation: From Principles to Practice. Washington, DC: AACC Press; 2001.

    Google Scholar 

  7. Costongs GM, Janson PC, Bas BM, Hermans J, Brombacher PJ, Van Wersch JW. Short-term and long-term intra-individual variations and critical differences of hematological laboratory parameters. J Clin Chem ClinBiochem. 1985;23(2):69–76 PubMed PMID: 4045388.

    CAS  Google Scholar 

  8. Garde AH, Hansen AM, Skovgaard LT, Christensen JM. Seasonal and biological variation of blood concentrations of total cholesterol, Dehydroepiandrosterone sulfate, hemoglobin A(1c), IgA, prolactin, and free testosterone in healthy women. Clin Chem. 2000;46(4):551-9.

  9. Fraser CG. Inherent biological variation and reference values. Clin Chem Lab Med. 2004;42(7):758-64. https://doi.org/10.1515/CCLM.2004.128.

  10. Ceriotti F, Hinzmann R, Panteghini M. Reference intervals: the way forward. Ann Clin Biochem. 2009;46(Pt 1):8–17. https://doi.org/10.1258/acb.2008.008170.Review PubMed PMID: 19103955.

    Article  PubMed  Google Scholar 

  11. Ricos C, Alvarez V, Cava F, García-Lario JV, Hernández A, Jiménez CV, et al. Current databases on Biological variation: pros, cons and progress Scand J Clin Lab Invest. 1999;59(7):491-500. https://doi.org/10.1080/00365519950185229.

  12. Ricos C, Carmen P, Michaela Joana, et al. Application of Biological Variation- a review, Biochemia Medica. 2009;19(3):250–9.

    Google Scholar 

  13. Aziz N, Jamieson BD, Quint JJ, Martinez-Maza O, Chow M, Detels R. Longitudinal Intra- and Inter-individual variation in T-cell subsets of HIV-infected and uninfected men participating in the LA Multi-Center AIDS Cohort Study. Medicine (Baltimore). 2019;98(41):e17525. https://doi.org/10.1097/MD.0000000000017525.

  14. Harris EK. Some theory of reference values. II. Comparison of some statistical models of intraindividual variation in blood constituents. Clin Chem. 1976;22(8):1343-50.

  15. Detels R, Jacobson L, Margolick J, Martinez-Maza O, Muñoz A, Phair J, Rinaldo C, et al. The Multicenter AIDS Cohort Study, 1983 to …… Public Health. 2012; 126:196–198. doi: https://doi.org/10.1016/j.puhe.2011.11.013. Epub 2011 Dec 27. PMID: 22206985

  16. Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18:499-502.

  17. Kung-Yee Liang and Scott Zeger. "Longitudinal data analysis using generalized linear models". Biometrika 1986, 73 (1): 13 22. doi:https://doi.org/10.1093/biomet/73.1.13

  18. Harris EK, Boyd JC. Comparison of within-subject and among-subjects variance; Statistical Bases of Reference Values in Laboratory Medicine. New York, N.Y.: Marcel Dekker, Inc.; 1995. p. 187–221.

    Book  Google Scholar 

  19. Harris EK. Effects of intra- and inter-individual variation on the appropriate use of normal ranges. Clin Chem. 1974;20(12):1535–42 PubMed PMID: 4430131.

    Article  CAS  Google Scholar 

  20. Pineda-Tenor D, Laserna-Mendieta EJ, Timón-Zapata J, Rodelgo-Jiménez L, Ramos-Corral R, Recio-Montealegre A, Reus MG. Biological variation and reference change values of common clinical chemistry and haematologic laboratory analytes in the elderly population. Clin Chem Lab Med. 2013;51(4):851–62. https://doi.org/10.1515/cCLDs-2012-0701 PMID: 23518452.

    Article  CAS  PubMed  Google Scholar 

  21. Rizi EP, Baig S, Toh SA, Loh TP, Khoo CM. Biological variation of glucose, insulin and lipids in lean, insulin-sensitive and obese, insulin-resistant Chinese males without diabetes. Pathology. 2016;48(5):510–2. https://doi.org/10.1016/j.pathol.2016.04.009 Epub 2016 Jun 14 PMID: 27311872.

    Article  CAS  PubMed  Google Scholar 

  22. Aarsand AK, Díaz-Garzón J, Fernandez-Calle P, Guerra E, Locatelli M, Bartlett WA, et al. The EuBIVAS: Within- and Between-Subject Biological Variation Data for Electrolytes, Lipids, Urea, Uric Acid, Total Protein, Total Bilirubin, Direct Bilirubin, and Glucose. Clin Chem. 2018;64(9):1380–93. https://doi.org/10.1373/clinchem.2018.288415 Epub 2018 Jun 25 PMID: 29941472.

    Article  CAS  PubMed  Google Scholar 

  23. Ortolá J, Castiñeiras MJ, Fuentes-Arderiu X. Biological variation data applied to the selection of serum lipid ratios used as risk markers of coronary heart disease. Clin Chem. 1992;38(1):56–9 PMID: 173360.

    Article  Google Scholar 

  24. Nicholaou MJ, Martinson JJ, Abraham AG, Brown TT, Hussain SK, Wolinsky SM, Kingsley LA. HAART-Associated Dyslipidemia Varies by Biogeographical Ancestry in the Multicenter AIDS Cohort Study. AIDS Res Hum Retroviruses. 2013;29(6):871–9. https://doi.org/10.1089/aid.2012.0169PMCID:PMC365339.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. Mukerji SS, Locascio JJ, Misra V, Lorenz DR, Holman A, Dutta A, at el Lipid Profiles and APOE4 Allele Impact Midlife Cognitive Decline in HIV-Infected Men on Antiretroviral Therapy Clin Infect Dis. 2016 Oct 15; 63(8): 1130–1139. Published online 2016 Jul 22. doi: https://doi.org/10.1093/cid/ciw495

  26. Petersen PH, Fraser CG, Sandberg S, Goldschmidt H. The index of individuality is often a misinterpreted quantity characteristic. Clin Chem Lab Med. 1999;37(6):655–61.

    Article  CAS  Google Scholar 

  27. Thyagarajan B, Howard AG, Durazo-Arvizu R, Eckfeldt JH, Gellman MD, Kim RS, et al. Analytical and biological variability in biomarker measurement in the Hispanic Community Health Study/Study of Latinos. Clin Chim Acta. 2016; 463:129–13.doi: https://doi.org/10.1016/j.cca.2016.10.019. Epub 2016 Oct 15. PubMed PMID: 27756543; PubMed Central PMCID: PMC5136507

  28. Riddler SA, Smit E, Cole SR, Li R, Chmiel JS, Doba A, et al. Impact of HIV infection and HAART on serum Lipids in Men. JAMA. 2003;289(22):2978-82. https://doi.org/10.1001/jama.289.22.2978.

  29. Ricós N, Iglesias N , José-Vicente García-Lario, Margarita Simón, Fernando Cava, Amparo Hernández, et al. Within-subject Biological Variation in Disease: Collated Data and Clinical Consequences, Ann Clin Biochem. 2007;44(Pt 4):343-52. https://doi.org/10.1258/000456307780945633.

  30. Aziz N, Quint JJ, Breen EC, Oishi J, Jamieson BD, Martinez-Maza O, Detels R. 30-Year Longitudinal Study of Hematological Parameters of HIV-1 Negative Men Participating in Los Angeles Multicenter AIDS Cohort Study (MACS). Lab Med. 2019;50(1):64–72. https://doi.org/10.1093/labmed/lmy044.PMID:30060104;PMCID:PMC6301185.

    Article  PubMed  Google Scholar 

Download references

Acknowledgements

We would like to thank the men who participate in the MACS, who make this and many other studies possible. Furthermore, we thank DACC with grant (U01-HL146193) for obtaining of cancers data.

Funding

The contents of this publication are solely the responsibility of the authors and do not represent the official views of the National Institutes of Health (NIH). MACS/WIHS Combined Cohort Study (MWCCS) Principal Investigator of Los Angeles CRS, funding U01-HL146333. The MWCCS is funded primarily by the National Heart, Lung, and Blood Institute (NHLBI), with additional co-funding from the Eunice Kennedy Shriver National Institute Of Child Health & Human Development (NICHD), National Institute On Aging (NIA), National Institute Of Dental & Craniofacial Research (NIDCR), National Institute Of Allergy And Infectious Diseases (NIAID), National Institute Of Neurological Disorders And Stroke (NINDS), National Institute Of Mental Health (NIMH), National Institute On Drug Abuse (NIDA), National Institute Of Nursing Research (NINR), National Cancer Institute (NCI), National Institute on Alcohol Abuse and Alcoholism (NIAAA), National Institute on Deafness and Other Communication Disorders (NIDCD), National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institute on Minority Health and Health Disparities (NIMHD), and in coordination and alignment with the research priorities of the National Institutes of Health, Office of AIDS Research (OAR). MWCCS data collection is also supported by UL1-TR000004 (UCSF CTSA), P30-AI-050409 (Atlanta CFAR), P30-AI-050410 (UNC CFAR), and P30-AI-027767 (UAB CFAR).

Author information

Authors and Affiliations

Authors

Contributions

NA* designed the experiment, drafted manuscript, and plots, DG interpreted the statistics analysis, MM editing writing, CD and RS collected and basic analysis, NA reviewed and collected health data, and RD major contributor in writing of manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Najib Aziz.

Ethics declarations

Ethics approval and consent to participants

This study was approved by the institutional review board (IRB) for human studies at the University of California Los Angeles (UCLA) and the informed consent signed by all participants.

Consent for publication

All participants signed the informed consent.

Competing interests

The authors declare of no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Aziz, N., Gjertson, D.W., Mimiaga, M.J. et al. Long-term intra- and inter-individual biological variation of serum lipid of HIV-infected and uninfected men participating in the Los Angeles Multi-Center AIDS Cohort Study (MACS). Lipids Health Dis 21, 63 (2022). https://doi.org/10.1186/s12944-022-01668-0

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12944-022-01668-0

Keywords