Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Jul 1.
Published in final edited form as: Kidney Int. 2016 May 7;90(1):172–180. doi: 10.1016/j.kint.2016.02.031

Impaired postprandial lipemic response in chronic kidney disease

Jeffrey M Saland 1, Lisa M Satlin 1, Jeanna Zalsos-Johnson 1, Serge Cremers 2,3, Henry N Ginsberg 2,3
PMCID: PMC4912421  NIHMSID: NIHMS785171  PMID: 27162092

Abstract

Dyslipidemia in chronic kidney disease (CKD) is usually characterized by hypertriglyceridemia. Here we studied post-prandial lipemia in children and young adults to determine if an increasing degree of CKD results in a proportional increase in triglyceride and chylomicron concentration. Secondary goals were to determine if subnephrotic proteinuria, apolipoprotein (apo)C-III and insulin resistance modify the CKD effect. Eighteen fasting participants (mean age of 15 years, mean glomerular filtration rate (GFR) of 50 ml/min/1.73m2) underwent a postprandial challenge with a high fat milkshake. Triglycerides, apoB-48, insulin, and other markers were measured before and 2, 4, 6, and 8 hours afterward. Response was assessed by the incremental area under the curve of triglycerides and of apoB-48. The primary hypothesis was tested by correlation to estimated GFR. Significantly, for every 10 ml/min/1.73m2 lower estimated GFR, the incremental area under the curve of triglycerides was 17% greater while that of apoB-48 was 16% greater. Univariate analyses also showed that the incremental area under the curve of triglycerides and apoB-48 were significantly associated with subnephrotic proteinuria, apoC-III, and insulin resistance. In multivariate analysis, CKD and insulin resistance were independently associated with increased area under the curve and were each linked to increased levels of apoC-III. Thus, postprandial triglyceride and chylomicron plasma excursions are increased in direct proportion to the degree of CKD. Independent effects are associated with subclinical insulin resistance and increased apoC-III is linked to both CKD and insulin resistance.

Keywords: chronic kidney disease, lipids, nutrition, insulin resistance, pediatric nephrology

Introduction

Dyslipidemia is frequent among individuals with chronic kidney disease (CKD), most commonly manifesting as hypertriglyceridemia with reduced high density lipoprotein (HDL) cholesterol concentrations. CKD in adults is often associated with other diseases and co-morbidities that affect lipid metabolism. Among children with CKD, primary renal disease is predominant yet the same dyslipidemia is seen, supporting the idea that it arises directly as a consequence of renal dysfunction.(1) While multiple CV risk factors are present in the CKD population and the deleterious effect of each is difficult to isolate, multiple lines of evidence support a pathological role for dyslipidemia in this setting.(2) CKD is associated with a many fold increased risk of cardiovascular (CV) events and death attributable to CV disease, justifying its study.(3)

The mechanism by which renal disease leads to dyslipidemia remains an active area of inquiry. Hypertriglyceridemia could result from two broad processes: 1) increased secretion of triglycerides (TG) as a component of very low density lipoproteins (VLDL) or chylomicrons (CM) from the liver or intestine, respectively or 2) decreased utilization or catabolism of VLDL and chylomicron TG. The latter mechanism is generally considered the more relevant one in CKD.(4, 5) Past investigations have separated renal insufficient patients and those with nephrotic syndrome; the latter routinely producing extreme dyslipidemia.

Most circulating TG is carried by apolipoprotein-B (apoB) containing lipoproteins. The liver synthesizes and secretes TG in VLDL, which each contain a single molecule of apoB-100. VLDL are metabolized to intermediate and low-density lipoproteins (IDL and LDL) during circulation. CM carry freshly absorbed dietary TG into the circulation following meals. Synthesized by the intestine and containing a single apoB-48 molecule, each CM delivers dietary TG to target tissues where lipolysis releases fatty acids for use as energy or storage. Lipoprotein lipase (LPL) is one of the principal lipases involved; apoC-III inhibits the activity of LPL, slowing TG catabolism. As TG is removed from its core, the CM shrinks to a less buoyant “remnant” particle that is taken up by the liver where residual CM TG is either oxidized or recycled in VLDL.(4)

Even in the fasting state, CM remnants are increased among individuals with severe CKD, suggesting a defect in clearance.(6, 7) Many of the catabolic mechanisms by which CM TG are cleared following a meal parallel the clearance of VLDL TG.(8) While the in vivo clearance of CM TG cannot be assessed independently from that of VLDL TG, the relative importance of hepatic synthesis is smaller in the postprandial state compared to fasting. Furthermore, since production of CM after a meal is closely related to the amount of TG ingested, a “fat tolerance test” is well suited to test the hypothesis that CKD impairs the clearance of CM and their TG, and to explore potential mediators of this CKD effect. Two prior studies of 9 and 21 subjects respectively, demonstrated defective post-prandial clearance of TG and CM remnants among adult dialysis-dependent patients.(9, 10) Impaired post-prandial TG clearance was also shown in a group of 14 adults with less severe CKD, in whom insulin resistance inversely proportional to GFR was also demonstrated. (11) While elevated apoC-III is a consistent finding in all hypertriglyceridemic states, recent studies have linked delayed VLDL catabolism to increased apoC-III levels and delayed apoC-III catabolism to CKD.(12, 13) To our knowledge, no prior study has investigated these abnormalities of triglyceride metabolism in children, or (in any age group) whether these abnormalities are proportional to the degree of CKD. Finally, we do not believe any prior investigation has measured the extent, if any, to which proteinuria contributes to these associations in non-nephrotic patients.

The purpose of this study was to investigate the effects of CKD on CM and TG metabolism following a single high fat meal in a pediatric and young adult group and determine if such effects vary with the degree of renal impairment. We hypothesized that CKD results in decreased clearance of CM and TG in proportion to the severity of reduction in renal function. We investigated the relevance of sub-nephrotic proteinuria and other potential mediators of impaired TG metabolism including apoC-III, insulin resistance, free fatty acid (FFA) concentrations, and markers of inflammation, as well as anthropomorphic measures of adiposity. We also determined whether this “fat tolerance test” provides measures that are more sensitive to the effects of CKD as compared to fasting measures.

Results

Cohort characteristics

The baseline characteristics and fasting lab values for the study group are shown in Table 1. All 3 subjects with a urine protein/creatinine ratio (uPCR) 2–3 mg/mg had congenital structural CKD, albumin levels of 4 g/dL or greater, and none manifested edema. The group was mostly teenagers with two participants under age 10 (ages 5 and 9), and one over age 20 (age 25). The HOmeostasis Model Assessment of Insulin Resistance Index (HOMA-IR) and body mass index (BMI) Z scores were elevated, though waist circumference was average. No subjects demonstrated fasting glucose levels > 100 mg/dl or postprandial glucose levels > 110 mg/dl. For descriptive purposes, supplemental figures 1–4 show post-prandial levels of TG, insulin, free fatty acids, and HDL-C.

Table 1. Study population characteristics and baseline measures.

Characteristics and fasting laboratory measures of the study group.

Variation is described by either mean ± standard deviation or median with (interquartile range, IQR)

Normal ranges or recommended (rec) levels are provided where available (uln: upper limit of normal)

Fasting Measures Normal Range
Age (years) 15 ± 5
GFR (ml/min/1.73m2) 53 ± 30 > 70 by equation
urine protein/creatinine
(mg/mg)		 0.8 ± 1.0 < 0.2
Serum Albumin (g/dl) 4.5 ± 0.5 3 – 5
BMI (kg/m2) 22 (18–29) (age/sex dependent)
BMI Z score 1.07 ± 1.28 0 ± 2
Waist Circumference Z score −0.05 ± 0.71 0 ± 2
Insulin (μIU/ml) 12.79 ± 6.5 5 – 15 (age dependent)
HOMA-IR 2.40 (1.7–3.8) < 2.5 (puberty dependent)
Total TG (mg/dl) - median
(IQR)		 121 (90–175) < 130 (age dependent)
Total Cholesterol (mg/dl) 176 ± 53 < 200
HDL-C (mg/dl) 42 ± 7 > 40 (guideline)
Non-HDL-C (mg/dl) 132 ± 51 < 130 (guideline)
LDL-C (mg/dl) 81 ± 30 < 100 (guideline)
VLDL-TG (mg/dl) - median
(IQR)		 58 (42–98) -
ApoA1 (mg/dl) 133 ± 21 104–225
ApoB (mg/dl) 84 ± 27 60–133
ApoB-48 (ng/L) 11.3 ± 5.3 1.4 – 9
ApoC-III (mg/dl) 12.4 ± 4.7 < ~9 (age dependent)
Lp(a) (mg/dl) - median (IQR) 17 (6–59) uln 14.6 M / 17.6 F
TNF-alpha (pg/mL) 2.5 ± 1.3 0.55–2.82
IL-6 (pg/mL) 1.5 ± 1.1 0.45–9.96
Free Fatty Acids (mMol/L) 0.6 ± 0.5 0.1–0.6
Open in a new tab

Association between GFR and AUCs of TG, apoB-48

Figures 1a and 1b demonstrate the associations between GFR and both total and incremental (above fasting level) postprandial TG clearance. Figures 1c and 1d demonstrate the same associations for CM clearance as measured by apoB-48 AUC (individual p values shown in figures). Table 2 provides estimates of effect size: for every 10 ml/min/1.73m2 lower GFR, the iAUC of TG was 17% (95% CI: 4–33%, p<0.05) greater and the iAUC of apoB-48 16% (95% CI: 1–33%, p<0.05) greater. Multivariable models confirmed these effect sizes and significance.

Figure 1.

Figure 1

Open in a new tab

Univariate associations for GFR: 1a) with plasma TG AUC, 1b) plasma TG iAUC, 1c) plasma apoB-48 AUC and 1d) plasma apoB-48 iAUC. Lower GFR was associated with greater AUC for each, indicating impaired clearance of TG and chylomicron remnants. Incremental AUC adjusts for the baseline associations that are also present.

Table 2. Determinants of post-prandial lipid excursions.

Magnitude of effect of GFR, HOMA-IR, fasting ApoC-III, and proteinuria on plasma excursions (iAUC) of TG and of chylomicrons (ApoB48). Unadjusted refers to univariate effects. Adjusted (1) refers to the model with both GFR and HOMA-IR, and Adjusted (2) refers to the model with GFR, HOMA-IR, and fasting Apo-CIII. In the last model, Apo-C-III itself was not a significant determinant.

Magnitude of Effects on TG Excursion (unadjusted) Adjusted(1) Adjusted(2)
Variable Change iAUC of TG 95% CI Change iAUC of TG 95% CI Change iAUC of TG 95% CI
GFR: each 10 ml/min/1.73m2 lower 17% higher (p < 0.05) 4 – 33 % 17% higher (p<0.005) 7 – 29 % 18% higher (p=0.01) 5 – 33 %
HOMA-IR: Each ½ unit higher 16% higher (p < 0.001) 3 – 29 % 16% higher (p<0.005) 15 – 17 % 16% higher (p=0.03) 1 – 33 %
Fasting ApoC-III: each mg/dL higher 12% higher (p < 0.005) 4 – 20 % not in model not statistically significant
Proteinuria: each 50% higher 16% higher (p<0.005) 6 – 26 % not in model not in model
Magnitude of Effects on apoB48 Excursion (unadjusted) Adjusted(1) Adjusted(2)
Variable Change iAUC of ApoB48 95% CI Change iAUC of ApoB48 95% CI Change iAUC of ApoB48 95% CI
GFR: each 10 ml/min/1.73m2 lower 16% higher (p < 0.05) 1 – 33 % 16% higher (p = 0.01) 4 – 30 % 19% higher (p = 0.02) 3 – 36 %
HOMA-IR: Each ½ unit higher 16% higher (p < 0.005) 3 – 30 % 16% higher (p < 0.01) 5 – 28 % 19% higher (p = 0.03) 2 – 40 %
Fasting ApoC-III: each mg/dL higher 12% higher (p < 0.01) 3 – 20 % not in model not statistically significant
Proteinuria: each 50% higher 15% higher (p < 0.01) 4 – 27 % not in model not in model
Open in a new tab

Association between sub-nephrotic proteinuria and AUCs of TG, apoB-48

Figures 2a and 2b demonstrate the associations between uPCR and both total and incremental postprandial TG clearance. Figures 2c and 2d show the same associations for apoB-48 AUC. All relationships were quantitatively and statistically significant (individual p values shown in figures). Table 2 provides estimates of effect size: for each 50% increase in uPCR in the group, the iAUC of TG was 12% (95% CI: 6–26%, p<0.005) greater and the iAUC of apoB-48 was 15% (95% CI: 4–27%, p<0.01) greater.

Figure 2.

Figure 2

Open in a new tab

Univariate associations for proteinuria expressed as the urine protein/creatinine ratio: 1a) plasma TG AUC, 1b) plasma TG iAUC, 1c) plasma apoB-48 AUC and 1d) plasma apoB-48 iAUC. An increasing degree of proteinuria was associated with greater AUC for each, indicating impaired clearance of TG and chylomicron remnants. Incremental AUC adjusts for the baseline associations that are also present. Not shown: proteinuria and GFR were not independently associated with TG or apoB-48 AUC.

Potential mediators of dyslipidemia and relationship to measures of CKD

ApoC-III: Greater fasting apoC-III was strongly associated with greater iAUC of TG (r=0.65, p<0.005) and greater iAUC of apoB-48 (r=0.60, p<0.01). Higher fasting apoC-III was also related to lower GFR (r=−0.62, p<0.01) and more proteinuria (r=0.74, p<0.001). In multivariable analysis including GFR, proteinuria, and fasting apoC-III, none were independently associated with either iAUC of TG or with iAUC of apoB-48, indicating significant interactions among these variables.

HOMA-IR: The correlations between HOMA-IR with the iAUC of TG (r=0.56, p<0.02) and with the iAUC of apoB-48 (r=0.54, p=0.02) were also robust. Interestingly, HOMA-IR was not related to GFR or proteinuria. Correspondingly, both lower GFR (p<0.005) and higher HOMA-IR (p<0.005) were independently associated with higher iAUC of TG after adjustment for them together in the same model (multivariable R2=0.64, p< 0.001, see also Table 2 for effect size). Likewise, both lower GFR (p=0.01) and higher HOMA-IR (p<0.01) were independently associated with higher iAUC of apoB-48 (multivariable R2=0.55, p< 0.005 see also Table 2 for effect size. Also shown in Table 2, while addition of fasting apoC-III did not improve the model, both GFR and HOMA-IR remained significant independent determinants of lipid excursion after the additional adjustment for fasting apoC-III.

Comparison of fasting measures to AUC measures

We performed several analyses of the total (not incremental) AUC of TG and total AUC of apoB-48. As expected, fasting TG (r=0.87, p<0.001) was the strongest predictor of total AUC of TG; fasting apoC-III (r=0.81, p<0.001) was also a very good predictor of total AUC of TG.

The relationships between the AUC of TG and GFR or uPCR (Figs 1a, 2a) were about the same as were observed between fasting TG and GFR and uPCR measures (r= −0.65, r=0.72 respectively, both p<0.005). In contrast, the associations of fasting apoB-48 with GFR (r= −0.57, p=0.01) and with uPCR (p=ns) were weaker than the associations between the AUC of apoB-48 and GFR or uPCR (Figs 1c, 2c) elicited by the study meal.

Inflammatory markers

Measurement of baseline (fasting) TNF-alpha and IL-6 found normal values (Table 1) and no relationship to any other study variables.

Further analysis of apoC-III

Given the relevance of apoC-III, our protocol allowed us to analyze its distribution, post-prandial variation, and relation to other markers.

The proportion of plasma apoC-III in the HDL fraction (fasting HDL:plasma ratio of apoC-III) did not correlate to other study measures (eGFR, proteinuria, BMI, HOMA-IR, or fastings levels of TG, apoC-III, HDL, or FFA). Post-prandially (Fig 3a), the apoC-III in HDL was stable (trend toward decrease not statistically significant) while the amount of apoC-III in non-HDL demonstrated meal-related changes.

Figure 3.

Figure 3

Open in a new tab

3a) Postprandial changes in apoC-III were observed only in the non-HDL fraction (trend of lower apoC-III in the HDL fraction was not significant by one-way ANOVA). 3b) the iAUC of apoC-III was strongly correlated to insulin resistance.

We compared fasting apoC-III to the excursion (iAUC) of apoC-III following the study meal with respect to key study measures. The relationship between iAUC of TG and apoC-III excursion was similar (r=0.61, p < 0.01) to its relation to fasting apoC-III. However, the magnitude of the relationship between fasting apoC-III and eGFR (r=−0.62, p < 0.01) was significant while the relationship between apoC-III excursion and eGFR was not (r=−0.42, p=ns). The same pattern was noted for proteinuria (r=0.74, p<0.001 for fasting apoC-III vs. r=0.47, p=ns for excursion of apoC-III). Conversely, the relationship between HOMA-IR and apoC-III was stronger for apoC-III excursion (Figure 3b, r=0.69, p = 0.001) as compared to fasting apoC-III (r=0.49, p = 0.04).

The association of higher HOMA-IR and higher iAUC of apoC-III was independent of GFR in multivariable analysis. Lower GFR and higher HOMA-IR were also each independent, statistically significant predictors of higher fasting apoC-III.

Discussion

The hypothesis that CKD impairs CM and TG clearance is supported by the robust associations demonstrated by this study. Focusing on the postprandial state following a standardized meal enabled us to reduce variation due to intrinsic differences in hepatic lipid and lipoprotein synthesis since the bulk of the incremental changes are derived from the meal itself. With a young group of subjects with primary CKD, no co-morbidities, and effects that varied with the degree of renal impairment, it is clear the cause of reduced CM and TG clearance in this group of young people is CKD itself. The results suggest mechanisms that center on insulin resistance, apoC-III, and non-nephrotic proteinuria as mediators of the CKD effect.

Our findings build upon those of past studies and add unique information to the field. To the authors’ knowledge, there are no previous studies of post-prandial lipid metabolism in children with CKD. The CKiD Study examined fasting lipid metabolism, showing an inverse relationship between fasting TG and GFR and a direct relationship between both sub-nephrotic and nephrotic range proteinuria in a cohort of children with primary CKD.(1) Among adults with CKD, two prior studies of 9 and 21 subjects respectively, demonstrated deficits in postprandial CM clearance in adults with dialysis-dependent ESRD but did not address insulin resistance, proteinuria, or less severe CKD.(9, 10) The prior work most similar to ours examined 14 non-dialysis-dependent adult patients with CKD and found impaired postprandial TG clearance was related to insulin resistance, but did not examine the effect of proteinuria nor whether the degree of the lipid clearance deficit was related to the degree of CKD.(11) Like our participants, none of the individuals in that study had abnormal plasma glucose concentrations, supporting the idea that subclinical insulin resistance is relevant in CKD. More recently, Chan, Ooi and colleagues examined the turnover of VLDL, IDL, and apoC-III in a group of adults with non-diabetic CKD. They demonstrated significant reductions in fractional catabolic rate (FCR) for VLDL and IDL, without differences in production rates, as well as 40% lower FCR for apoC-III in CKD compared to healthy control subjects.(12, 13)

Our work is the first to demonstrate that the severity of impairment of postprandial clearance of TG and CM is directly proportional to the severity of CKD. There was a graded response to the degree of CKD whether expressed as level of GFR or by the amount of (sub-nephrotic) proteinuria. Our work is also the first to study postprandial changes in children and teens with CKD and we demonstrate they suffer from a similar metabolic dysfunction as adults. Although many of the relationships we studied were also demonstrable using the fasting-state (baseline) measures, it should be noted that the majority of our subjects had normal fasting lipid concentrations. Abnormal post-prandial responses in CKD patients with normal fasting measures was also noted by Charlesworth, et al.(11) Interestingly, this discrepancy parallels widespread clinical experience: while the CV risks of dyslipidemia are clear, current guidelines suggest that lipid levels are not the best guide to CV risk(14). Thus, our study also supports postprandial testing as a research method in this population since it allowed us to recreate the clinical experience of normal fasting lipids with abnormal lipid metabolism, elucidate a richer view of that dyslipidemia, and show an association between insulin resistance and CKD.

Not only is elevated apoC-III associated with elevated TG (15), but an extended line of research suggests a causal effect. ApoC-III modulates binding of lipoprotein to cell receptors(16, 17) and proteoglycans.(18) Excess apoC-III inhibits LpL-mediated TG lipolysis.(19–21) Transgenic mice overexpressing apoC-III develop hypertriglyceridemia (16, 22) and knockout mice are resistant to postprandial hypertriglyceridemia and experimental antagonism of apo-CIII reduces hypertriglyceridemia.(23, 24) Humans deficient in apoC-III and apoA-I show uninhibited LPL activity with low TG levels and more rapid VLDL catabolism(25), and APOC3 haplo-insufficiency also confers low plasma TG levels and possibly reduced atherosclerosis.(26–28) Anti apoC-III therapy reduces TG levels in humans with the familial chylomicronemia syndrome (LPL deficiency) by up to 86%.(29) Thus, apoC-III is an active mediator of hypertriglyceridemia.

The associations among apoC-III, HOMA-IR, and CKD in our study are intriguing. Each demonstrated a strong individual relationship to post-prandial excursion of TG and apoB-48, with mutual interactions among each other as well. We might conjecture that apoC-III is the most likely immediate mediator of impaired TG metabolism among these factors. This is consistent with the known activity of apoC-III as well as the recent finding of decreased apoC-III FCR in subjects with CKD.(12) Still, if apoC-III is the proximate mediator of impaired triglyceride metabolism, it remains unknown how CKD and insulin resistance independently lead to increased apoC-III.

It is not possible to completely separate fasting from post-prandial manifestations of impaired TG metabolism. Within that context, our analysis suggests that lower GFR and more proteinuria (worse CKD) was more associated with the fasting level of apoC-III than with the post-prandial apoC-III excursion. In contrast, insulin resistance had more balanced effects on fasting and post-prandial apoC-III. We conjecture this follows from the fact that while GFR changes negligibly with a meal, insulin levels and the relevance of its actions change drastically after eating. Regardless, our findings in this regard are very limited. To whatever extent reduced GFR can be studied separately of insulin resistance, further work will be needed to better define their independent effects on apoC-III.

Insulin has been shown to suppress apoC-III gene expression (30, 31) so it should not be surprising to find higher levels of apoC-III with insulin resistance. Still, since insulin resistance may lead to hyperinsulinemia, in theory, the effect on apoC-III synthesis could depend on whether that specific activity is “resistant” or not. For example, hepatic “insulin resistance” is associated with increased hepatic production of fatty acids and TG, an insulin-induced effect. One explanation of that phenomenon is that hepatic lipogenesis follows activation of SREBP-1c, which retains insulin sensitivity despite hepatic resistance to insulin’s effects to regulate glucose metabolism. While it is interesting that an SREBP-1 binding site has been demonstrated on APOC3 (32) its relevance in this setting is not clear. Finally, it is important to note that the relation between plasma levels of apoC-III and insulin in human studies has been inconsistent--the relationship remains incompletely understood.

CKD has the potential to impact multiple sites of apoC-III catabolism. Though the precise distribution of apoC-III degradation sites is not known, most likely it is cleared by holoparticle uptake while bound to HDL, VLDL, or CM remnants. ApoC-III is not known to be directly cleared by glomerular filtration. The impact of subnephrotic proteinuria was significant in this study but its exact role is unclear; the study lacked power to differentiate the effects of proteinuria from decreased GFR as these two markers of CKD interact strongly. Further investigation is required to determine the nature and consequences of post-translational alterations in apoC-III itself due to CKD, such as variation in sialyation that likely influence its function or catabolism. Increased sialyation of apoC-III was noted in a small cohort of several patients with severe CKD(33).

Although our study was particularly focused on the role of apoC-III, a variety of other mechanisms have been shown to result in hyperlipidemia in CKD.(34) Animal studies, while limited in terms of direct applicability to human physiology, have been used to examine the effects of CKD on lipid metabolism in more depth than is possible in humans. Rats with severely reduced GFR via 5/6 nephrectomy have reduced target tissue (adipose, cardiac and skeletal muscle) LPL as well as reduced tissue GPIHBP1 (blycosylphosphatidylinositol-anchored binding protein 1, a molecule that facilitates TG lipolysis from chylomicrons in part by facilitating LPL transport and binding), and rats with the additional insult of FSGS-like nephrotic syndrome also demonstrated reduced hepatic VLDL-receptors.(35–38) All of these alterations would result in reduced triglyceride clearance. Likewise, hepatic lipase and hepatic LDL receptor-related protein (LRP) expression and quantity have also been found to be reduced in rats with reduced GFR, representing additional defects that could reduce triglyceride clearance.(34, 39, 40). Our results must be considered in the context of these and other mechanisms studied in animals with CKD.(34)

This study has strengths and limitations. The primary limitation its small size, which did not allow as much analysis of multiple variables as desired. Another limitation is that the study did not include interventions such as labeled (radioactive or stable isotope) metabolic “tracers” that could potentially illuminate turnover and catabolism more clearly. Likewise, a direct measure of GFR, such as plasma disappearance of intravenously injected iohexol would have been more precise than an estimated GFR. In planning our study we considered barriers to child study participation, and both of the latter interventions would have added to the burden on subjects with a combination of more complexity and additional risk (actual and perceived).

Treating eGFR as a continuous variable (“exposure”) is a strength of this study. Prior studies had already shown differences between healthy subjects and those with CKD, so our use of a study cohort with a range of renal dysfunction (from healthy to severe) allowed us to determine not only whether CKD is associated with poor TG clearance but also whether it is related to the degree of renal dysfunction. Just as a dose-response relation provides additional positive evidence of an association, our demonstration of a direct relationship between the level of renal dysfunction and the degree of impaired TG clearance provides additional evidence of this relationship. Other strengths include the assessment of proteinuria and insulin resistance as potentially important predictors of TG metabolism.

The novel findings of this study inform further investigation and also have clinically relevant implications. In particular, the association between sub-nephrotic proteinuria and TG metabolism is not only highly plausible (in that the effect of nephrotic proteinuria is profound), but it is also an association that can be clinically tested and treated since proteinuria is modifiable for example by angiotensin converting enzyme inhibitors. The role of subclinical insulin resistance too can be clinically tested and addressed for example by exercise or metformin. Finally, by demonstrating these effects for the first time in children and teens with primary CKD and no other conditions, it is clear that the metabolic defects result from CKD itself rather than secondary phenomena.

In conclusion, postprandial TG and CM metabolism is diminished in direct proportion to the degree of CKD. Linked to increased fasting and post-prandial levels of apoC-III respectively, CKD and subclinical insulin resistance appear to contribute independently to the observed defects in TG and CM clearance.

Materials and Methods

Study Participants

Approval for human research was obtained from The Icahn School of Medicine at Mount Sinai. To determine if lipid abnormalities vary with GFR, we created a study group with a range renal dysfunction that varied from none to severe. Thus, 16 individuals with CKD due to congenital renal defects or primary glomerular disease and two healthy individuals were recruited for a total group of 18 subjects. Exclusions included nephrotic syndrome, serum albumin < 3.5, uncontrolled hypertension, acute illness, severe anemia (Hb < 8 g/dL), smoking, corticosteroids, lipid-lowering agents, transplant, dialysis, or other significant non-renal conditions.

Participants arrived in the morning after a 10–14 hour fast. Following a fasting blood draw, each drank a single high-fat “milk shake” in an amount normalized to their body surface area up to a maximum of 2 m2 (Table 3). For several initial subjects the milk-shake included 50,000 IU/m2 retinyl palmitate to mark CM particles; this was later abandoned as standardized apoB-48 measurement became available. Postprandial blood samples were taken at 2,4,6, and 8 hours, during which time the subject was permitted fluids but no caloric intake of any kind.

Table 3. Nutritional information for the “Milkshake” diet.

Nutritional information for the study milkshake. The amount provided was uniform and based on the body surface area of each participant, to a maximum of 2m2.

Calories 1095
Fat (g) 95
Saturated Fat (g) 47
Carbohydrate (g) 44
Protein (g) 24
Cholesterol (mg) 273
Open in a new tab

Measurements

Serum creatinine, potassium, albumin, urine protein and urine creatinine were all measured by standard auto analyzer (Roche Diagnostics) in the Mount Sinai Hospital clinical laboratory. The remaining measures were performed in the Biomarkers Core Laboratory of the Irving Institute for Clinical and Translational Research at Columbia University Medical Center. Lipoproteins were isolated from fasting and postprandial samples by sequential ultracentrifugation of plasma, beginning with CM/VLDL at d < 1.006 g/mL, in a 50.3ti rotor using a Beckman TL100 ultracentrifuge (Beckman Coulter, USA). Cholesterol, TG, and glucose were measured using enzymatic techniques, while direct HDL-C, apoB-100, apoC-III, apoA-I, Lp(a), and cystatin-C were measured by turbidimetric techniques, all on a standard auto analyzer (Roche Diagnostics). FFA were measured colorimetrically (Wako Diagnostics, Richmond, VA). apoB-48 (Biovendor, Candler, NC), insulin (Millipore, Billerica, MA), IL-6 and TNF-alpha (R&D Systems, Minneapolis, MN) were measured by ELISA.(41)

Calculations

GFR was calculated using an equation which takes into account height, gender, creatinine, BUN, and cystatin-C.(42) uPCR was calculated in mg/mg. TG, apoB-48, and uPCR were log transformed for normality. The response to the high fat challenge was calculated as both total area under the curves (AUC) and incremental AUC (iAUC: only that area above fasting value, Supplemental Fig 1). To allow comparison of children of differing ages and gender, BMI and waist circumference Z scores were calculated from normative national growth data.(43, 44) Insulin sensitivity was estimated by the standard calculation of the HOMA-IR = Insulin (mU/L) × glucose (mg/dl) ÷ 405; values normally increase during puberty.(45–47)

Statistics

Statistical analyses were performed with Sigmaplot (V.11, Systat Software, Chicago IL). As this was the first study to assess whether TG clearance varies with GFR, sample size estimates required several assumptions. We extrapolated the effect sizes in dialysis patients to milder CKD by assuming a linear association across a range of GFR from zero to 100 and allowed for large standard deviations. This estimation suggested a sample size of approximately 24 participants would be required to detect effects of this magnitude with a power of 0.80 and alpha of 0.05. Interim analysis demonstrated a smaller than expected variability in measures and allowed fewer participants.

Univariate associations were assessed by linear regression, reporting correlations as r (Pearson product-moment correlation coefficient) where positive values indicate a direct relationship and negative values indicate an inverse relationship. The squared value, r2, quantifies the proportion of variation explained by the variable in question The magnitude of associations (regression slope) was reported in the units of measure or as percent change for log-transformed variables. 95% confidence intervals for effect size were calculated as the product of standard error of r by the t-distribution value for the 2.5th and 97.5th percentiles (degrees of freedom = N of 18 – 1 – number of variables in the model). Limited by cohort size, multivariable analysis was used to define only the most robust relationships.

Supplementary Material

1
2
3
4
5

Acknowledgments

The authors thank the nursing staff of the clinical research unit at the Mount Sinai Hospital, the staff of the core laboratory of the Core Laboratory at Columbia University Medical Center, and Colleen Ngai for her assistance with this project.

Support:

Dr. Saland was supported by the NIH/NIDDK via K23DK071871.

Clinical research at the Mount Sinai Hospital was supported by NIH/NCRR via UL1RR029887.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Disclosure: The authors have no financial conflicts of interest related to this work.

This work was presented in preliminary form at the 2012 annual meeting of the Pediatric Academic Societies in Boston, MA

References

  • 1.Saland JM, Pierce CB, Mitsnefes MM, Flynn JT, Goebel J, Kupferman JC, et al. Dyslipidemia in children with chronic kidney disease. Kidney Int. 2010 Aug 25; doi: 10.1038/ki.2010.311. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Baigent C, Landray MJ, Reith C, Emberson J, Wheeler DC, Tomson C, et al. The effects of lowering LDL cholesterol with simvastatin plus ezetimibe in patients with chronic kidney disease (Study of Heart and Renal Protection): a randomised placebo-controlled trial. Lancet. 2011 Jun 25;377(9784):2181–2192. doi: 10.1016/S0140-6736(11)60739-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Levey AS. Controlling the epidemic of cardiovascular disease in chronic renal disease: where do we start? Am J Kidney Dis. 1998 Nov;32(5 Suppl 3):S5–S13. doi: 10.1053/ajkd.1998.v32.pm9820463. [DOI] [PubMed] [Google Scholar]
  • 4.Saland JM, Ginsberg HN. Lipoprotein metabolism in chronic renal insufficiency. Pediatr Nephrol. 2007 Aug;22(8):1095–1112. doi: 10.1007/s00467-007-0467-5. [DOI] [PubMed] [Google Scholar]
  • 5.Vaziri ND. Causes of dysregulation of lipid metabolism in chronic renal failure. Semin Dial. 2009 Nov-Dec;22(6):644–651. doi: 10.1111/j.1525-139X.2009.00661.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Nestel PJ, Fidge NH, Tan MH. Increased lipoprotein-remnant formation in chronic renal failure. N Engl J. 1982 Aug 05;307(6):329–333. doi: 10.1056/NEJM198208053070601. 1982. [DOI] [PubMed] [Google Scholar]
  • 7.Wilson DE, Chan IF, Cheung AK, Dutz W, Buchi KN. Retinyl ester retention in chronic renal failure. Further evidence for a defect in chylomicron remnant metabolism. Atherosclerosis. 1985 Nov;57(2–3):189–197. doi: 10.1016/0021-9150(85)90032-2. [DOI] [PubMed] [Google Scholar]
  • 8.Brunzell JD, Hazzard WR, Porte D, Jr, Bierman EL. Evidence for a common, saturable, triglyceride removal mechanism for chylomicrons and very low density lipoproteins in man. J Clin Invest. 1973 Jul;52(7):1578–1585. doi: 10.1172/JCI107334. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Miyamoto T, Rashid Qureshi A, Yamamoto T, Nakashima A, Lindholm B, Stenvinkel P, et al. Postprandial metabolic response to a fat- and carbohydrate-rich meal in patients with chronic kidney disease. Nephrol Dial Transplant. 2011 Jul;26(7):2231–2237. doi: 10.1093/ndt/gfq697. [DOI] [PubMed] [Google Scholar]
  • 10.Weintraub M, Burstein A, Rassin T, Liron M, Ringel Y, Cabili S, et al. Severe defect in clearing postprandial chylomicron remnants in dialysis patients. Kidney Int. 1992 Nov;42(5):1247–1252. doi: 10.1038/ki.1992.411. [DOI] [PubMed] [Google Scholar]
  • 11.Charlesworth JA, Kriketos AD, Jones JE, Erlich JH, Campbell LV, Peake PW. Insulin resistance and postprandial triglyceride levels in primary renal disease. Metabolism. 2005 Jun;54(6):821–828. doi: 10.1016/j.metabol.2005.01.028. [DOI] [PubMed] [Google Scholar]
  • 12.Ooi EM, Chan DT, Watts GF, Chan DC, Ng TW, Dogra GK, et al. Plasma apolipoprotein C-III metabolism in patients with chronic kidney disease. J Lipid Res. 2011 Apr;52(4):794–800. doi: 10.1194/jlr.M011163. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Chan DT, Dogra GK, Irish AB, Ooi EM, Barrett PH, Chan DC, et al. Chronic kidney disease delays VLDL-apoB-100 particle catabolism: potential role of apolipoprotein C-III. J Lipid Res. 2009 Dec;50(12):2524–2531. doi: 10.1194/jlr.P900003-JLR200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Sarnak MJ, Bloom R, Muntner P, Rahman M, Saland JM, Wilson PW, et al. KDOQI US commentary on the 2013 KDIGO Clinical Practice Guideline for Lipid Management in CKD. Am J Kidney Dis. 2015 Mar;65(3):354–366. doi: 10.1053/j.ajkd.2014.10.005. [DOI] [PubMed] [Google Scholar]
  • 15.Curry MD, McConathy WJ, Fesmire JD, Alaupovic P. Quantitative determination of human apolipoprotein C-III by electroimmunoassay. Biochim Biophys Acta. 1980 Mar 21;617(3):503–513. doi: 10.1016/0005-2760(80)90016-8. [DOI] [PubMed] [Google Scholar]
  • 16.Aalto-Setala K, Fisher EA, Chen X, Chajek-Shaul T, Hayek T, Zechner R, et al. Mechanism of hypertriglyceridemia in human apolipoprotein (apo) CIII transgenic mice. Diminished very low density lipoprotein fractional catabolic rate associated with increased apo CIII and reduced apo E on the particles. J Clin Invest. 1992 Nov;90(5):1889–1900. doi: 10.1172/JCI116066. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Shelburne FA, Quarfordt SH. A new apoprotein of human plasma very low density lipoproteins. J Biol Chem. 1974 Mar 10;249:1428–1433. (5 UR - PM:4361735) [PubMed] [Google Scholar]
  • 18.van Barlingen HH, de Jong H, Erkelens DW, de Bruin TW. Lipoprotein lipase-enhanced binding of human triglyceride-rich lipoproteins to heparan sulfate: modulation by apolipoprotein E and apolipoprotein C. J Lipid Res. 1996 Apr;37(4):754–763. [PubMed] [Google Scholar]
  • 19.Krauss RM, Herbert PN, Levy RI, Fredrickson DS. Further observations on the activation and inhibition of lipoprotein lipase by apolipoproteins. Circ Res. 1973 Oct;33:403–411. doi: 10.1161/01.res.33.4.403. (4 UR - PM:4355035) [DOI] [PubMed] [Google Scholar]
  • 20.McConathy WJ, Gesquiere JC, Bass H, Tartar A, Fruchart JC, Wang CS. Inhibition of lipoprotein lipase activity by synthetic peptides of apolipoprotein C-III. J Lipid Res. 1992 Jul;33(7):995–1003. [PubMed] [Google Scholar]
  • 21.Brown WV, Baginsky ML. Inhibition of lipoprotein lipase by an apoprotein of human very low density lipoprotein. Biochem Biophys Res Commun. 1972 Jan 31;46:375–382. doi: 10.1016/s0006-291x(72)80149-9. (2 UR - PM:5057882) [DOI] [PubMed] [Google Scholar]
  • 22.Ito Y, Azrolan N, O'Connell A, Walsh A, Breslow JL. Hypertriglyceridemia as a result of human apo CIII gene expression in transgenic mice. Science. 1990 Aug 17;249(4970):790–793. doi: 10.1126/science.2167514. [DOI] [PubMed] [Google Scholar]
  • 23.Maeda N, Li H, Lee D, Oliver P, Quarfordt SH, Osada J. Targeted disruption of the apolipoprotein C-III gene in mice results in hypotriglyceridemia and protection from postprandial hypertriglyceridemia. J Biol Chem. 1994 Sep 23;269(38):23610–23616. [PubMed] [Google Scholar]
  • 24.Graham MJ, Lee RG, Bell TA, 3rd, Fu W, Mullick AE, Alexander VJ, et al. Antisense oligonucleotide inhibition of apolipoprotein C-III reduces plasma triglycerides in rodents, nonhuman primates, and humans. Circ. Res. 2013 May 24;112(11):1479–1490. doi: 10.1161/CIRCRESAHA.111.300367. [DOI] [PubMed] [Google Scholar]
  • 25.Ginsberg HN, Le NA, Goldberg IJ, Gibson JC, Rubinstein A, Wang-Iverson P, et al. Apolipoprotein B metabolism in subjects with deficiency of apolipoproteins CIII and AI. Evidence that apolipoprotein CIII inhibits catabolism of triglyceride-rich lipoproteins by lipoprotein lipase in vivo. J Clin Invest. 1986 Nov;78(5):1287–1295. doi: 10.1172/JCI112713. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Pollin TI, Damcott CM, Shen H, Ott SH, Shelton J, Horenstein RB, et al. A null mutation in human APOC3 confers a favorable plasma lipid profile and apparent cardioprotection. Science. 2008 Dec 12;322(5908):1702–1705. doi: 10.1126/science.1161524. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Jorgensen AB, Frikke-Schmidt R, Nordestgaard BG, Tybjaerg-Hansen A. Loss-of-function mutations in APOC3 and risk of ischemic vascular disease. N Engl J. Med. 2014 Jul 3;371(1):32–41. doi: 10.1056/NEJMoa1308027. [DOI] [PubMed] [Google Scholar]
  • 28.Crosby J, Peloso GM, Auer PL, Crosslin DR, Stitziel NO, Lange LA, et al. Loss-of-function mutations in APOC3, triglycerides, and coronary disease. N Engl J. Med. 2014 Jul 3;371(1):22–31. doi: 10.1056/NEJMoa1307095. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Gaudet D, Brisson D, Tremblay K, Alexander VJ, Singleton W, Hughes SG, et al. Targeting APOC3 in the familial chylomicronemia syndrome. N Engl J Med. 2014 Dec 4;371(23):2200–2206. doi: 10.1056/NEJMoa1400284. [DOI] [PubMed] [Google Scholar]
  • 30.Altomonte J, Cong L, Harbaran S, Richter A, Xu J, Meseck M, et al. Foxo1 mediates insulin action on apoC-III and triglyceride metabolism. J Clin Invest. 2004 Nov;114(10):1493–1503. doi: 10.1172/JCI19992. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Chen M, Breslow JL, Li W, Leff T. Transcriptional regulation of the apoC-III gene by insulin in diabetic mice: correlation with changes in plasma triglyceride levels. J Lipid Res. 1994 Nov;35(11):1918–1924. [PubMed] [Google Scholar]
  • 32.Motallebipour M, Enroth S, Punga T, Ameur A, Koch C, Dunham I, et al. Novel genes in cell cycle control and lipid metabolism with dynamically regulated binding sites for sterol regulatory element-binding protein 1 and RNA polymerase II in HepG2 cells detected by chromatin immunoprecipitation with microarray detection. FEBS J. 2009 Apr;276(7):1878–1890. doi: 10.1111/j.1742-4658.2009.06914.x. [DOI] [PubMed] [Google Scholar]
  • 33.Holdsworth G, Stocks J, Dodson P, Galton DJ. An abnormal triglyceride-rich lipoprotein containing excess sialylated apolipoprotein C-III. J Clin Invest. 1982 Apr;69(4):932–939. doi: 10.1172/JCI110532. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Vaziri ND. Dyslipidemia of chronic renal failure: the nature, mechanisms, and potential consequences. Am J Physiol Renal Physiol. 2006 Feb;290(2):F262–F272. doi: 10.1152/ajprenal.00099.2005. [DOI] [PubMed] [Google Scholar]
  • 35.Vaziri ND, Liang K. Down-regulation of VLDL receptor expression in chronic experimental renal failure. Kidney Int. 1997 Mar;51(3):913–919. doi: 10.1038/ki.1997.129. [DOI] [PubMed] [Google Scholar]
  • 36.Vaziri ND, Liang K. Down-regulation of tissue lipoprotein lipase expression in experimental chronic renal failure. Kidney Int. 1996 Dec;50(6):1928–1935. doi: 10.1038/ki.1996.515. [DOI] [PubMed] [Google Scholar]
  • 37.Sato T, Liang K, Vaziri ND. Down-regulation of lipoprotein lipase and VLDL receptor in rats with focal glomerulosclerosis. Kidney Int. 2002 Jan;61(1):157–162. doi: 10.1046/j.1523-1755.2002.00104.x. [DOI] [PubMed] [Google Scholar]
  • 38.Vaziri ND, Yuan J, Ni Z, Nicholas SB, Norris KC. Lipoprotein lipase deficiency in chronic kidney disease is accompanied by down-regulation of endothelial GPIHBP1 expression. Clin Exp Nephrol. 2012 Apr;16(2):238–243. doi: 10.1007/s10157-011-0549-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Klin M, Smogorzewski M, Ni Z, Zhang G, Massry SG. Abnormalities in hepatic lipase in chronic renal failure: role of excess parathyroid hormone. J Clin Invest. 1996 May 15;97(10):2167–2173. doi: 10.1172/JCI118657. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Kim C, Vaziri ND. Down-regulation of hepatic LDL receptor-related protein (LRP) in chronic renal failure. Kidney Int. 2005 Mar;67(3):1028–1032. doi: 10.1111/j.1523-1755.2005.00166.x. [DOI] [PubMed] [Google Scholar]
  • 41.Kinoshita M, Kojima M, Matsushima T, Teramoto T. Determination of apolipoprotein B-48 in serum by a sandwich ELISA. Clin Chim Acta. 2005 Jan;351(1–2):115–120. doi: 10.1016/j.cccn.2004.08.008. [DOI] [PubMed] [Google Scholar]
  • 42.Schwartz GJ, Munoz A, Schneider MF, Mak RH, Kaskel F, Warady BA, et al. New equations to estimate GFR in children with CKD. J Am Soc Nephrol. 2009 Mar;20(3):629–637. doi: 10.1681/ASN.2008030287. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Ritz E, Orth SR. N Engl J. Med. Vol. 341. Heidelberg, Germany: Department of Internal Medicine, Ruperto Carola University; 1999. Oct 07, Nephropathy in patients with type 2 diabetes mellitus; pp. 1127–1133. (15 AD - UR - PM:10511612) [DOI] [PubMed] [Google Scholar]
  • 44.McDowell M, Fryar C, Ogden C, Flegal K. Anthropometric reference data for children and adults: United States, 2003–2006. In: Statistics NCfH, editor. National health statistics reports. Hyattsville, MD: 2008. [PubMed] [Google Scholar]
  • 45.Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985 Jul;28(7):412–419. doi: 10.1007/BF00280883. [DOI] [PubMed] [Google Scholar]
  • 46.Kurtoglu S, Hatipoglu N, Mazicioglu M, Kendirici M, Keskin M, Kondolot M. Insulin resistance in obese children and adolescents: HOMA-IR cut-off levels in the prepubertal and pubertal periods. J Clin Res Pediatr Endocrinol. 2010;2(3):100–106. doi: 10.4274/jcrpe.v2i3.100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.d'Annunzio G, Vanelli M, Pistorio A, Minuto N, Bergamino L, Lafusco D, et al. Insulin resistance and secretion indexes in healthy Italian children and adolescents: a multicentre study. Acta Biomed. 2009 Apr;80(1):21–28. [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS785171-supplement-1.docx (13.8KB, docx)
2
NIHMS785171-supplement-2.tif (174.8KB, tif)
3
NIHMS785171-supplement-3.tif (206.2KB, tif)
4
NIHMS785171-supplement-4.tif (180KB, tif)
5
NIHMS785171-supplement-5.tif (150.6KB, tif)

RESOURCES