Triglyceride content of lipoprotein subclasses and kidney hemodynamic function and injury in adolescents with type 1 diabetes

Abstract

Aims:

Elevated triglycerides (TG) are associated with development and progression of kidney disease, and TG distributions across lipoprotein subclasses predict kidney dysfunction in adults with type 1 diabetes (T1D). Little is known regarding these relationships in youth.

Methods:

In this single center study conducted from October 2018 – 2019, lipid constituents from lipoprotein subclasses were quantified by targeted nuclear magnetic resonance spectroscopy. Glomerular filtration rate (GFR), renal plasma flow (RPF), afferent arteriolar resistance (R_A), efferent arteriolar resistance (R_E), intraglomerular pressure (P_GLO), urine albumin-to-creatinine ratio (UACR), and chitinase-3-like protein 1 (YKL-40), a marker of kidney tubule injury, were assessed. Cross-sectional relationships were assessed by correlation and multivariable linear regression (adjusted for age, sex, HbA1c) models.

Results:

Fifty youth with T1D (age 16±3 years, 50% female, HbA1c 8.7±1.3%, T1D duration 5.7±2.6 years) were included. Very-low-density lipoprotein (VLDL)-TG concentrations correlated and associated with intraglomerular hemodynamic function markers including GFR, P_GLO, UACR, as did small low-density lipoprotein (LDL)-TG and small high-density lipoprotein (HDL)-TG. YKL-40 correlated with all lipoprotein subclasses.

Conclusion:

TG within lipoprotein subclasses, particularly VLDL, associated with P_GLO, GFR, albuminuria, and YKL-40. Lipid perturbations may serve as novel targets to mitigate early kidney disease.

1.0. Introduction

Diabetes-related kidney disease (DKD) strongly contributes to morbidity and mortality, including higher rates of cardiovascular disease, the leading cause of death in people with diabetes (1). The presence and number of cardiorenal risk factors such as elevated blood pressure, overweight/obesity, inadequate diabetes management, lack of exercise, family history of kidney and/or cardiovascular disease, and dyslipidemia contribute to risk of long-term complications among individuals with DKD (1–3).

Atherogenic dyslipidemia, defined as elevated triglycerides (TG) and decreased high-density lipoprotein cholesterol (HDL-C) concentrations, is strongly associated with the progression of DKD and is exacerbated by insulin resistance present in both type 1 diabetes (T1D) and type 2 diabetes (T2D) (4). Indeed, adults with inadequately controlled T1D demonstrate increased TG-rich lipoprotein particles and TG-rich low-density lipoprotein (LDL) particles, with shifts of LDL subclasses to an excess of small, dense LDL (5). Youth with T1D also demonstrate impaired cholesterol metabolism including increased free fatty acid uptake, impaired fatty acid oxidation, and increased lipid accumulation at the level of the kidneys, which may contribute to the development of kidney disease (6). Additionally, youth with T1D exhibit early intraglomerular hemodynamic dysfunction, characterized by elevated intraglomerular pressure (P_GLO), renal plasma flow (RPF), and glomerular filtration rate (GFR), prior to onset of overt decline in kidney function (7–9).

In the Coronary Artery Calcification in Type 1 Diabetes (CACTI) study, we found that elevated serum TG concentrations predict new-onset albuminuria in adults with T1D (10). TG concentrations within different lipoprotein subclasses also predict chronic kidney disease and mortality in T1D (11, 12), potentially secondary to the effects of insulin resistance on reductions in functional lipoprotein lipase. Additionally, each lipoprotein class is heterogeneous in size and composition and inherent molecular variation may contribute to the development of atherosclerosis and vascular complications including kidney and cardiovascular disease in T1D (13, 14). Yet, the mechanisms underlying associations between lipid aberrations and development of kidney dysfunction are not fully understood. In this study, we evaluated relationships between TG concentrations within lipoprotein subclasses and gold-standard measures of GFR, RPF, P_GLO, and albuminuria in youth with T1D. Additionally, we assessed relationships with chitinase-3-like protein 1 (YKL-40), a marker of kidney tubule injury that is associated with progression of kidney disease (8, 15). We hypothesized that TG concentrations within lipoprotein subclasses associate with YKL-40, parameters of intraglomerular hemodynamic dysfunction, and albuminuria.

2.0. Subjects, Materials, and Methods

2.1. Study Design and Participants

Fifty adolescents aged 12-21 years with T1D of 1-10 years duration and hemoglobin A1c (HbA1c) of <11% from the Copeptin in Adolescent Participants with Type 1 Diabetes and Early Renal Hemodynamic Function study conducted from October 2018 – October 2019 (CASPER, NCT03618420) were included in this cross-sectional analysis. T1D was defined by American Diabetes Association criteria plus the presence of glutamic acid decarboxylase, islet cell, zinc transporter 8 and/or insulin autoantibodies, and the need for exogenous insulin from the onset of diabetes. T1D duration, defined as the time between T1D diagnosis and the study visit, was extracted from the electronic medical record. Exclusion criteria were previously reported (16). Of importance, no participants were taking medication for the treatment of dyslipidemia at the time of study participation. Participants were recruited from the Pediatric Clinic at the Barbara Davis Center for Diabetes on the University of Colorado Anschutz Medical Campus in Aurora, Colorado, USA. All participants were examined at the University of Colorado’s Clinical and Translational Research Center (CTRC) where tests were preceded by 3 days of restricted physical activity and a fixed-macronutrient, sodium, and protein replete, weight-maintenance diet and performed in the morning after a 12-hour fast. Participants remained in a fasting state until the completion of all study procedures.

The CASPER protocol was approved by the Colorado Multiple Institutional Review Board (COMIRB). Participants or parents/guardians, as appropriate, provided written informed assent and/or consent. Puberty stage was assessed by a pediatric endocrinologist using the standards of Tanner and Marshall for breast development in girls (i.e., inspection and palpation) and genital development in boys. Vital signs (height, weight, systolic and diastolic blood pressures) were measured by standard procedures.

2.2. Targeted Nuclear Magnetic Resonance (NMR) Spectroscopy for TG Concentrations Within Lipoprotein Subclasses

Nightingale Health Ltd. (Helsinki, Finland) employs a targeted NMR spectroscopy platform to quantify absolute concentrations of 98 lipid constituents from 14 lipoprotein subclasses using a proprietary Bayesian algorithm. Particle sizes for very-low-density lipoprotein (VLDL), LDL, intermediate-density lipoprotein (IDL), and HDL, were measured. TG content was measured within each lipoprotein subclass and further stratified by particle size, as previously described (17, 18). This targeted NMR spectroscopy platform has been used in epidemiological biomarker profiling studies (11, 19, 20).

2.3. GFR and RPF by Iohexol and p-aminohippurate (PAH) Clearance Techniques, Urine Albumin-to-Creatinine Ratio (UACR), and Blood Pressure

An intravenous (IV) line was placed, and participants were asked to empty their bladders. Spot plasma and urine samples were collected prior to iohexol and PAH infusions. Iohexol was administered through bolus IV injection (5 mL of 300 mg/mL [Omnipaque 300, GE Healthcare]). An equilibration period of 120 minutes was used and blood collections for iohexol plasma clearance were drawn at +120, +150, +180, +210, +240 minutes (21). Because the Brøchner-Mortensen equation underestimates high values of GFR, the Jødal-Brøchner-Mortensen (JBM) equation was used to calculate the GFR (22). PAH (2 g/10 mL, prepared at the University of Minnesota with a dose of [weight in kg/75 x 4.2 mL] [IND #140129]) was given slowly over 5 minutes followed by a continuous infusion of 8 mL of PAH in 42 mL of normal saline at a rate of 24 mL/hour for 2 hours. After an equilibration period, blood was drawn at +90 and +120 minutes and RPF was calculated as PAH clearance divided by the estimated extraction ratio of PAH which varies by the level of GFR (23). We report absolute GFR (mL/min) and RPF (mL/min) in the main analyses because the practice of indexing GFR and RPF for body surface area (BSA) introduces noise into the clearance measurements (24). GFR and RPF were measured in the setting of mild hyperglycemia (goal blood glucose concentration 170-190 mg/dL [9.4-10.6 mmol/L]) achieved by a modified hyperglycemic clamp technique with paired 20% dextrose and insulin intravenous (IV) infusions. This glycemic range was chosen to mimic the typical glycemic milieu of an adolescent with T1D (equivalent to an HbA1c ~7.6-8.2%) and maintain steady-state glycemic and insulin concentrations during kidney function measures (25, 26). UACR was also measured during fasting from spot urine samples before and after renal clearance assessments and averaged.

2.4. Intraglomerular Hemodynamic Parameter Calculations

The following intraglomerular hemodynamic parameters were estimated or calculated, as appropriate, using equations derived by Gomez, et al (27), and assuming K_FG = 0.1012 mL/s/mm Hg for individuals with diabetes: P_GLO (mm Hg), afferent arteriolar resistance [R_A, (dyne*s/cm⁵)], efferent arteriolar resistance [R_E (dyne*s/cm⁵)], filtration pressure across glomerular capillaries [ΔP_F (mm Hg)], glomerular oncotic pressure [π_G (mm Hg)], renal blood flow (RBF), and renal vascular resistance (RVR). Equations utilized to estimate or calculate these parameters are reported in Table 1.

Table 1.

Equations for Calculating Intraglomerular Hemodynamic Parameters

Gomez Equations
RVR = MAP/RBF
ΔPF = GFR/KFG
πG = 5 x (CM − 2)
CM = TP/ FF x ln (1/(1 − FF))
PGLO = ΔPF + PBOW* + πG
RA = [(MAP − PGLO)/RBF] x 1328**
RE = [GFR/(KFG x (RBF-GFR))] x 1328**

^*Assumes P_BOW equals 10 mm Hg.

^**Using Ohm’s law, 1328 is the conversion factor to dyne/s/cm⁵.

Abbreviations: RBF, renal blood flow; RPF, renal plasma flow; RVR, renal vascular resistance; MAP, mean arterial pressure; ΔP_F; filtration pressure across glomerular capillaries; GFR, glomerular filtration rate; K_FG, gross filtration coefficient equal to 0.1012mL/s/mm Hg; π_G, glomerular oncotic pressure; C_M, plasma protein mean concentration; TP, total protein; FF, filtration fraction; ln, natural logarithm; P_GLO, glomerular hydrostatic pressure; P_BOW, hydrostatic pressure in Bowman’s space_; R_A, afferent arteriolar resistance; R_E, efferent arteriolar resistance.

2.5. Laboratory Assessments

All laboratory measurements were performed by standard methods in either the University of Colorado CTRC Core Laboratories, Aurora, CO, USA; the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Laboratory, Phoenix, AZ, USA; or Nightingale Health Ltd., Helsinki, Finland. Insulin concentration was measured via Clinical Laboratory Improvement Amendments (CLIA)-certified chemiluminescent immunoassay (Beckman Coulter, Brea, CA). Iohexol and PAH concentrations were measured in Phoenix by high-performance liquid chromatography (Waters, Milford, MA). Other fasting laboratory evaluations included total cholesterol, low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), TG, glucose, and HbA1c (Diabetes Control and Complications Trial [DCCT]-calibrated). YKL-40 is a marker of kidney tubule injury that holds potential as a biomarker for DKD, as it demonstrates associative relationships with albuminuria and progression of kidney disease (8, 23). Serum YKL-40 was measured using a Meso Scale Discovery (MSD) QuickPlex SQ120 platform through electrochemiluminescence detection via a U-plex sandwich immunoassay with patterned array after a controlled sample thaw (reference range 69-197 ng/mL, Meso Scale Diagnostics, Gaithersburg, MD (24).

2.6. Statistical Analysis

Means and standard deviations were calculated for continuous variables, except those with highly skewed distributions, which were summarized by medians and inter-quartile ranges (IQRs). Categorical variables were described with numbers and percentages. Linear relationships were found between variables, thus Pearson’s correlation coefficient as well as generalized linear regression models were used to examine the relationships among lipoprotein subclasses and markers of intraglomerular hemodynamic function, including RVR, GFR, RPF, P_GLO, albuminuria, R_A, R_E, systolic and diastolic blood pressure, HbA1c, and YKL-40. Multivariable models were adjusted for age, sex, and HbA1c. A priori the decision was made to consider all analyses exploratory and hypothesis generating; thus, corrections for multiple comparisons were not performed. An alpha-value of 0.05 was considered statistically significant. All statistical analyses were performed in SAS version 9.4 (SAS Institute Inc).

2.7. Data and Resource Availability

The dataset generated and analyzed during this study, in addition to GFR, RPF, and modified hyperglycemic clamp protocols, are available from the corresponding author upon reasonable request.

3.0. Results

3.1. Baseline Characteristics

Fifty adolescents with T1D (age 16.0 ± 3.0 years, HbA1c 8.7 ± 1.3%, diabetes duration 5.7 ± 2.6 years) were included. Baseline characteristics of participants are reported in Table 2. Table 3 reports results of a standard lipid panel, including total cholesterol, HDL-C, LDL-C, and TG, as well as an advanced lipid panel including TG content of each lipoprotein subclass stratified by particle size. Standard lipid panel (HDL-C, LDL-C, TG) mean values were within goal ranges for adolescents with T1D. The majority of the total TG concentration was found to be distributed within the VLDL particle sizes, as expected.

Table 2.

Baseline Study Participant Characteristics

Characteristic	Result (n=50)
Age (years)	16.0 ± 3.0
Tanner stage	5 (4-5)
T1D duration (years)	5.7 ± 2.6
Sex (female, %)	50%
Race/ethnicity (%)
Black non-Hispanic	2
Hispanic	6
White non-Hispanic	92
Other	0
BMI (kg/m2)	23.4 ± 5.1
Weight (kg)	67.5 ± 17.6
Blood pressure (mm Hg)
Systolic blood pressure	119 ± 9
Diastolic blood pressure	74 ± 11
Mean arterial pressure	89 ± 9
HbA1c (%)	8.7 ± 1.3
UACR (mg/g)	6 (5-14)
GFR (mL/min)	189 ± 40
GFR (mL/min/1.73 m2)	183 ± 26
RPF (mL/min)	820 ± 125
RPF (mL/min/1.73 m2)	824 ± 120
YKL-40 (ng/mL)	24907 (17808-32343)

Data are presented as mean ± SD, median and IQR, or percent.

Abbreviations: T1D, type 1 diabetes; BMI, body mass index; HbA1c, hemoglobin A1c; UACR, urinary albumin to creatinine ratio; GFR, glomerular filtration rate; RPF, renal plasma flow; YKL-40, chitinase-3-like protein 1.

Table 3.

Baseline Study Participant Lipid Concentrations

Lipid	Concentration
Total cholesterol (mmol/L)	3.84 ± 0.74
LDL-C (mmol/L)	2.30 ± 0.62
HDL-C (mmol/L)	1.32 ± 0.26
TG (mmol/L)	0.88 ± 0.40
Largest VLDL-TG (mmol/L)	0.03 (0.00-0.08)
Very Large VLDL-TG (mmol/L)	0.07 ± 0.05
Large VLDL-TG (mmol/L)	0.14 ± 0.07
Medium VLDL-TG (mmol/L)	0.24 ± 0.08
Small VLDL-TG (mmol/L)	0.12 ± 0.04
Very Small VLDL-TG (mmol/L)	0.05 ± 0.01
IDL-TG (mmol/L)	0.08 ± 0.02
Large LDL-TG (mmol/L)	0.09 ± 0.02
Medium LDL-TG (mmol/L)	0.03 ± 0.01
Small LDL-TG (mmol/L)	0.01 ± 0.00
Very Large HDL-TG (mmol/L)	0.01 ± 0.00
Large HDL-TG (mmol/L)	0.02 ± 0.01
Medium HDL-TG (mmol/L)	0.04 ± 0.01
Small HDL-TG (mmol/L)	0.04 ± 0.01

Data are presented as mean ± SD, median and IQR, or percent.

Abbreviations: LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; TG, triglycerides; VLDL, very low-density lipoprotein; IDL, indeterminate density lipoprotein.

3.2. Relationships among standard lipid analysis and parameters of kidney function

Results of Pearson correlation and multivariate linear regression analyses investigating for relationships among standard lipid analysis components (total cholesterol, TG, LDL-C, and HDL-C) and markers of intraglomerular hemodynamic function are reported in Table 4. Most notably, TG demonstrated significant relationships with multiple markers across both analyses, including GFR, YKL-40, and SBP. Few significant relationships were seen among total cholesterol, LDL-C, or HDL-C.

Table 4.

Pearson correlation (R) and multivariable linear regression analyses (β±SE) among markers of intraglomerular hemodynamic function and standard lipids.

	TG		Total Cholesterol		LDL-C		HDL-C
	R	β±SE	R	β±SE	R	β±SE	R	β±SE
RVR (mm Hg/L/min · 1,000)	R: 0.17 p=0.33	−0.07 ± 0.05 p=0.19	R: 0.07 p=0.69	−0.03 ± 0.07 p=0.64	R: 0.18 p=0.29	0.09 ± 0.07 p=0.25	R: 0.09 p=0.58	0.09 ± 0.18 p=0.63
GFR (mL/min)	R: 0.49 P<0.001	0.48 ± 0.15 p=0.002	R: 0.26 p=0.07	0.33 ± 0.20 p=0.10	R: 0.26 p=0.07	0.34 ± 0.22 p=0.14	R: −0.10 p=0.51	−0.40 ± 0.55 p=0.47
RPF (mL/min)	R: 0.17 p=0.30	0.35 ± 0.60 p=0.57	R: 0.05 p=0.76	0.06 ± 0.73 p=0.93	R: 0.05 p=0.77	0.01 ± 0.83 p=0.99	R: −0.12 p=0.49	−1.78 ± 1.98 p=0.37
PGLO (mm Hg)	R: 0.41 p=0.01	0.07 ± 0.03 p=0.06	R: 0.29 p=0.09	0.07 ± 0.04 p=0.09	R: 0.16 p=0.36	0.03 ± 0.05 p=0.53	R: 0.07 p=0.67	0.05 ± 0.12 p=0.67
RA (dyne · sec · cm−5)	R: 0.11 p=0.52	2.69 ± 2.60 p=0.31	R: −0.11 p=0.53	−1.66 ± 3.19 p=0.61	R: 0.10 p=0.55	2.89 ± 3.63 p=0.43	R: −0.07 p=0.66	−4.30 ± 8.78 p=0.63
RE (dyne · sec · cm−5)	R: 0.31 p=0.07	2.63 ± 1.71 p=0.13	R: 0.18 p=0.27	1.93 ± 2.13 p=0.37	R: 0.14 p=0.40	1.48 ± 2.45 p=0.55	R: 0.12 p=0.49	3.13 ± 5.89 p=0.60
UACR (mg/g)	R: 0.42 p=0.002	0.010 ± 0.003 p=0.002	R: 0.05 p= 0.71	0.0004 ± 0.0043 p=0.93	R: 0.16 p=0.26	0.005 ± 0.005 p=0.32	R: −0.23 p=0.10	−0.03 ± 0.01 p=0.01
YKL-40 (ng/mL)	R: 0.44 p=0.002	0.005 ± 0.002 p=0.003	R: 0.30 p=0.03	0.005 ± 0.002 p=0.06	R: 0.29 p=0.05	0.004 ± 0.003 p=0.08	R: 0.13 p=0.36	0.004 ± 0.01 p=0.53
SBP (mm Hg)	R: 0.44 p=0.002	0.12 ± 0.04 p=0.003	R: 0.24 p=0.09	0.09 ± 0.05 p=0.08	R: 0.35 p=0.01	0.13 ± 0.05 p=0.02	R: −0.11 p=0.46	−0.10 ± 0.14 p=0.51
DBP (mm Hg)	R: 0.34 p=0.02	0.09 ± 0.04 p=0.05	R: 0.09 p=0.52	0.03 ± 0.05 p=0.54	R: 0.22 p=0.13	0.08 ± 0.06 p=0.17	R: −0.01 p=0.96	0.002 ± 0.15 p=0.99

Multivariable linear regression analyses adjusted for age, sex, and HbA1c. Multivariable linear regression results indicate the degree of change in the marker of function for every 1 mmol/L increase in the respective lipid (i.e., for every 1 mmol/L increase in TG, there is 0.48 ± 0.15 increase in GFR). Bold font indicates statistical significance (p<0.05).

Abbreviations: TG, triglycerides; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; RVR, renal vascular resistance; GFR, glomerular filtration rate; RPF, renal plasma flow; P_GLO, intraglomerular pressure; R_A, afferent arteriolar resistance; R_E, efferent arteriolar resistance; UACR, urinary albumin to creatinine ratio; YKL-40, chitinase-3-like protein; SBP, systolic blood pressure; DBP, diastolic blood pressure.

3.3. Correlations among lipoprotein subclasses and parameters of kidney function

Results of correlation studies and multivariate linear regression models are reported by lipoprotein subclass in Tables 5 through 8. Significant positive correlations were seen among most VLDL-TG subclasses with GFR, P_GLO, R_E, UACR, and systolic and diastolic blood pressure (Table 5). Small and medium LDL-TG also demonstrated positive correlations with GFR, P_GLO, R_E, and UACR. Small LDL-TG also correlated with systolic blood pressure. Large LDL-TG only correlated with P_GLO and UACR. (Table 6). For comparison, small HDL-TG also demonstrated positive correlations with GFR, P_GLO, R_E, UACR, and systolic and diastolic blood pressure (Table 7), while IDL-TG was only positively correlated with P_GLO and UACR (Table 8). YKL-40 demonstrated significant positive correlation with the concentrations of all lipoprotein subclasses (Tables 5 through 8).

Table 5.

Pearson correlation (R) and multivariable linear regression analyses (β±SE) among markers of intraglomerular hemodynamic function and TG concentrations across VLDL-TG particles.

	Largest VLDL-TG		Very Large VLDL-TG		Large VLDL-TG		Medium VLDL-TG		Small VLDL-TG		Very Small VLDL-TG
	R	β±SE	R	β±SE	R	β±SE	R	β±SE	R	β±SE	R	β±SE
RVR (mm Hg/L/min · 1,000)	R: 0.08 p=0.62	0.02 ± 0.02 p=0.45	R: 0.11 p=0.52	0.03 ± 0.04 p=0.35	R: 0.10 p=0.55	0.02 ± 0.03 p=0.39	R: 0.10 p=0.56	0.02 ± 0.02 p=0.42	R: 0.19 p=0.26	0.08 ± 0.06 p=0.16	R: 0.22 p=0.19	0.23 ± 0.15 p=0.14
GFR (mL/min)	R: 0.53 p<0.001	222.00 ± 57.80 p<0.001	R: 0.55 p<0.001	384.30 ± 93.70 P<0.001	R: 0.56 p<0.001	303.20 ± 72.70 P<0.001	R: 0.53 p<0.001	232.50 ± 59.60 P<0.001	R: 0.46 P<0.001	457.70 ± 146.90 p=0.003	R: 0.28 p=0.05	711.20 ± 422.00 p=0.10
RPF (mL/min)	R: 0.24 p=0.15	254.20 ± 235.70 p=0.29	R: 0.25 p=0.14	407.60 ± 389.20 p=0.30	R: 0.25 p=0.14	318.20 ± 307.30 p=0.31	R: 0.25 p=0.13	252.80 ± 258.10 p=0.33	R: 0.18 p=0.30	286.40 ± 635.90 p=0.66	R: 0.13 p=0.43	323.70 ± 1733.80 p=0.85
PGLO (mm Hg)	R: 0.44 p=0.01	31.75 ± 13.32 p=0.02	R: 0.53 P<0.001	65.01 ± 20.88 p=0.004	R: 0.55 P<0.001	54.07 ± 16.21 p=0.002	R: 0.58 P<0.001	48.32 ± 13.26 P<0.001	R: 0.52 p=0.001	100.57 ± 34.06 p=0.01	R: 0.35 p=0.03	169.58 ± 100.09 p=0.10
RA (dyne · sec · cm−5)	R: 0.03 p=0.87	499.80 ± 1052.60 p=0.64	R: −0.01 p=0.95	446.70 ± 1740.40 p=0.80	R: −0.04 p=0.83	107.30 ± 1375.00 p=0.94	R: −0.07 p=0.67	−202.00 ± 1152.40 p=0.86	R: 0.01 p=0.93	1064.40 ± 2801.80 p=0.71	R: 0.09 p=0.59	5496.50 ± 7576.30 p=0.47
RE (dyne · sec · cm−5)	R: 0.33 p=0.04	1184.00 ± 677.20 p=0.09	R: 0.40 p=0.02	2355.00 ± 1092.40 p=0.04	R: 0.41 p=0.01	1941.70 ± 856.60 p=0.03	R: 0.44 p=0.01	1693.30 ± 713.40 p=0.02	R: 0.46 p=0.004	4423.70 ± 1714.40 p=0.01	R: 0.35 p=0.03	8557.70 ± 4895.80 p=0.09
UACR (mg/g)	R: 0.44 p=0.002	4.53 ± 1.30 p=0.001	R: 0.49 P<0.001	8.23 ± 2.08 P<0.001	R: 0.48 P<0.001	6.18 ± 1.64 P<0.001	R: 0.50 P<0.001	4.88 ± 1.33 P<0.001	R: 0.44 p=0.001	9.61 ± 3.25 p=0.01	R: 0.38 p=0.01	18.42 ± 9.15 p=0.05
YKL-40 (ng/mL)	R: 0.57 p<0.001	2.99 ± 0.63 p<0.001	R: 0.58 p<0.001	4.91 ± 1.05 p<0.001	R: 0.55 p<0.001	3.61 ± 0.84 p<0.001	R: 0.51 P<0.001	2.56 ± 0.71 P<0.001	R: 0.51 P<0.001	5.88 ± 1.67 p=0.001	R: 0.48 p<0.001	14.20 ± 4.60 p=0.003
SBP (mm Hg)	R: 0.51 P<0.001	53.13 ± 13.29 P<0.001	R: 0.48 P<0.001	81.14 ± 22.46 P<0.001	R: 0.45 P<0.001	59.02 ± 17.82 p=0.002	R: 0.39 p=0.01	39.46 ± 14.90 p=0.01	R: 0.36 p=0.01	86.93 ± 35.31 p=0.02	R: 0.28 p=0.05	175.01 ± 97.64 p=0.08
DBP (mm Hg)	R: 0.36 p=0.01	40.23 ± 17.23 p=0.02	R: 0.36 p=0.01	63.82 ± 28.55 p=0.03	R: 0.34 p=0.02	45.20 ± 22.46 p=0.05	R: 0.29 p=0.04	27.76 ± 18.43 p=0.14	R: 0.25 p=0.08	53.71 ± 43.64 p=0.22	R: 0.19 p=0.19	94.13 ± 118.39 p=0.43

Multivariable linear regression analyses adjusted for age, sex, and HbA1c. Multivariable linear regression results indicate the degree of change in the marker of function for every 1 mmol/L increase in VLDL-TG (i.e., for every 1mmol/L increase in largest VLDL-TG, there is 222.00 ± 57.80 increase in GFR). Bold font indicates statistical significance (p<0.05).

Abbreviations: VLDL, very low-density lipoprotein; TG, triglycerides; RVR, renal vascular resistance; GFR, glomerular filtration rate; RPF, renal plasma flow; P_GLO, intraglomerular pressure; R_A, afferent arteriolar resistance; R_E, efferent arteriolar resistance; UACR, urinary albumin to creatinine ratio; YKL-40, chitinase-3-like protein; SBP, systolic blood pressure; DBP, diastolic blood pressure.

Table 8.

Pearson correlation (R) and multivariable linear regression analyses (β±SE) among markers of intraglomerular hemodynamic function and IDL-TG particles.

	IDL-TG
	R	β±SE
RVR (mm Hg/L/min · 1,000)	R: 0.10 p=0.56	0.07 ± 0.11 p=0.54
GFR (mL/min)	R: 0.23 p=0.10	472.70 ± 312.60 p=0.14
RPF (mL/min)	R: 0.21 p=0.21	877.60 ± 1263.70 p=0.49
PGLO (mm Hg)	R: 0.35 p=0.03	132.69 ± 73.01 p=0.08
RA (dyne · sec · cm−5)	R: −0.01 p=0.95	−188.40 ± 5604.40 p=0.97
RE (dyne · sec · cm−5)	R: 0.28 p=0.09	4601.70 ± 3671.70 p=0.22
UACR (mg/g)	R: 0.38 p=0.01	12.55 ± 6.78 p=0.07
YKL-40 (ng/mL)	R: 0.440 p=0.001	9.33 ± 3.46 p=0.01
SBP (mm Hg)	R: 0.21 p=0.15	94.84 ± 73.06 p=0.20
DBP (mm Hg)	R: 0.14 p=0.35	38.20 ± 87.60 p=0.66

Multivariable linear regression analyses adjusted for age, sex, and HbA1c. Multivariable linear regression results indicate the degree of change in the marker of function for every 1 mmol/L increase in IDL-TG (i.e., for every 1 mmol/L increase in IDL-TG, there is 472.70 ± 312.60 increase in GFR). Bold font indicates statistical significance (p<0.05).

Abbreviations: IDL, intermediate-density lipoprotein; TG, triglycerides; RVR, renal vascular resistance; GFR, glomerular filtration rate; RPF, renal plasma flow; P_GLO, intraglomerular pressure; R_A, afferent arteriolar resistance; R_E, efferent arteriolar resistance; UACR, urinary albumin to creatinine ratio; YKL-40, chitinase-3-like protein; SBP, systolic blood pressure; DBP, diastolic blood pressure.

Table 6.

Pearson correlation (R) and multivariable linear regression analyses (β±SE) among markers of intraglomerular hemodynamic function and TG concentrations across LDL-TG particles.

	Large LDL-TG		Medium LDL-TG		Small LDL-TG
	R	β±SE	R	β±SE	R	β±SE
RVR (mm Hg/L/min · 1,000)	R: 0.02 p=0.88	0.01 ± 0.11 p=0.91	R: 0.05 p=0.76	0.13 ± 0.32 p=0.70	R: 0.08 p=0.63	0.47 ± 0.65 p=0.48
GFR (mL/min)	R: 0.27 p=0.06	548.00 ± 298.10 p=0.07	R: 0.37 p=0.01	2210.60 ± 832.90 p=0.01	R: 0.51 p<0.001	6264.70 ± 1687.70 p<0.001
RPF (mL/min)	R: 0.27 p=0.11	1313.30 ± 1197.10 p=0.28	R: 0.28 p=0.09	3861.80 ± 6441.80 p=0.27	R: 0.27 p=0.10	7522.80 ± 7175.00 p=0.30
PGLO (mm Hg)	R: 0.41 p=0.01	156.45 ± 68.04 p=0.03	R: 0.49 p=0.002	557.84 ± 186.92 p=0.01	R: 0.57 P<0.001	1301.10 ± 374.52 p=0.002
RA (dyne · sec · cm−5)	R: −0.09 p=0.61	−3057.50 ± 5340.50 p=0.57	R: −0.09 p=0.59	−7417.00 ± 15390.40 p=0.63	R: −0.08 p=0.62	−8573.20 ± 32086.70 p=0.79
RE (dyne · sec · cm−5)	R: 0.28 p=0.09	4310.80 ± 3520.40 p=0.23	R: 0.36 p=0.03	17695.00 ± 9881.30 p=0.08	R: 0.44 p=0.01	47935.10 ± 19819.20 p=0.02
UACR (mg/g)	R: 0.43 p=0.002	14.33 ± 6.44 p=0.03	R: 0.49 p<0.001	53.47 ± 17.97 p=0.01	R: 0.53 p<0.001	137.96 ± 37.01 P<0.001
YKL-40 (ng/mL)	R: 0.41 p=0.003	8.05 ± 3.39 p=0.02	R: 0.48 P<0.001	29.09 ± 9.48 p=0.004	R: 0.58 p<0.001	80.82 ± 18.93 P<0.001
SBP (mm Hg)	R: 0.20 p=0.17	85.52 ± 70.64 p=0.23	R: 0.28 p=0.05	368.96 ± 200.63 p=0.07	R: 0.39 p=0.01	1134.34 ± 414.82 p=0.01
DBP (mm Hg)	R: 0.13 p=0.36	32.39 ± 84.54 p=0.70	R: 0.19 p=0.20	181.34 ± 243.92 p=0.46	R: 0.26 p=0.06	681.00 ± 518.60 p=0.20

Multivariable linear regression analyses adjusted for age, sex, and HbA1c. Multivariable linear regression results indicate the degree of change in the marker of function for every 1 mmol/L increase in LDL-TG (i.e., for every 1 mmol/L increase in large LDL-TG, there is a 548.00 ± 298.10 increase in GFR). Bold font indicates statistical significance (p<0.05).

Abbreviations: LDL, low-density lipoprotein; TG, triglycerides; RVR, renal vascular resistance; GFR, glomerular filtration rate; RPF, renal plasma flow; P_GLO, intraglomerular pressure; R_A, afferent arteriolar resistance; R_E, efferent arteriolar resistance; UACR, urinary albumin to creatinine ratio; YKL-40, chitinase-3-like protein; SBP, systolic blood pressure; DBP, diastolic blood pressure.

Table 7.

Pearson correlation (R) and multivariable linear regression analyses (β±SE) among markers of intraglomerular hemodynamic function and TG concentrations across HDL-TG particles.

	Very Large HDL-TG		Large HDL-TG		Medium HDL-TG		Small HDL-TG
	R	β±SE	R	β±SE	R	β±SE	R	β±SE
RVR (mm Hg/L/min · 1,000)	R: −0.03 p=0.85	−0.33 ± 1.23 p=0.79	R: 0.04 p=0.83	−0.02 ± 0.22 p=0.93	R: 0.22 p=0.19	0.20 ± 0.16 p=0.22	R: 0.27 p=0.10	0.35 ± 0.19 p=0.07
GFR (mL/min)	R: 0.20 p=0.16	4529.10 ± 3366.30 p=0.19	R: −0.06 p=0.69	−192.20 ± 603.80 p=0.75	R: 0.12 p=0.42	338.50 ± 452.20 p=0.46	R: 0.34 p=0.01	1157.80 ± 540.10 p=0.04
RPF (mL/min)	R: 0.14 p=0.41	4047.60 ± 13556.00 p=0.77	R: 0.01 p=0.95	−853.70 ± 2464.90 p=0.73	R: 0.08 p=0.66	−108.40 ± 1837.40 p=0.95	R: 0.15 p=0.39	684.20 ± 2171.00 p=0.75
PGLO (mm Hg)	R: 0.34 p=0.04	1474.41 ± 774.94 p=0.07	R: 0.09 p=0.60	45.73 ± 148.51 p=0.76	R: 0.23 p=0.17	107.11 ± 109.04 p=0.33	R: 0.41 p=0.01	246.31 ± 123.51 p=0.05
RA (dyne · sec · cm−5)	R: −0.22 p=0.18	−88543.90 ± 57670.40 p=0.13	R: −0.8 p=0.64	−9087.60 ± 10751.30 p=0.40	R: 0.13 p=0.44	5359.10 ± 8033.20 p=0.51	R: 0.17 p=0.32	11249.90 ± 9362.50 p=0.24
RE (dyne · sec · cm−5)	R: 0.37 p=0.02	75778.20 ± 37794.10 p=0.05	R: 0.17 p=0.31	4936.80 ± 7242.00 p=0.50	R: 0.22 p=0.19	4831.00 ± 5360.20 p=0.37	R: 0.33 p=0.04	9748.10 ± 6187.10 p=0.13
UACR (mg/g)	R: 0.16 p=0.27	17.41 ± 75.32 p=0.82	R: −0.04 p= 0.76	−16.33 ± 13.04 p=0.22	R: 0.23 p=0.11	9.01 ± 9.90 p=0.37	R: 0.46 P<0.001	35.30 ± 11.28 p=0.003
YKL-40 (ng/mL)	R: 0.44 p=0.001	103.78 ± 36.83 p=0.01	R: 0.30 p=0.03	11.87 ± 6.81 p=0.09	R: 0.44 p=0.002	13.78 ± 4.88 p=0.01	R: 0.47 p<0.001	18.79 ± 5.97 p=0.003
SBP (mm Hg)	R: 0.07 p=0.64	203.43 ± 796.80 p=0.80	R: −0.07 p=0.64	−102.24 ± 139.55 p=0.47	R: 0.17 p=0.22	101.15 ± 104.59 p=0.34	R: 0.35 p=0.01	288.25 ± 124.48 p=0.03
DBP (mm Hg)	R: −0.00 p=0.98	−446.84 ± 938.28 p=0.64	R: −0.09 p=0.55	−151.03 ± 164.07 p=0.36	R: 0.13 p=0.37	60.07 ± 124.33 p=0.63	R: 0.31 p=0.03	257.45 ± 150.53 p=0.09

Multivariable linear regression analyses adjusted for age, sex, and HbA1c. Multivariable linear regression results indicate the degree of change in the marker of function for every 1 mmol/L increase in HDL-TG (i.e., for every 1mmol/L increase in very large HDL-TG, there is a 4529.10 ± 3366.30 increase in GFR). Bold font indicates statistical significance (p<0.05).

Abbreviations: HDL, high-density lipoprotein; TG, triglycerides; RVR, renal vascular resistance; GFR, glomerular filtration rate; RPF, renal plasma flow; P_GLO, intraglomerular pressure; R_A, afferent arteriolar resistance; R_E, efferent arteriolar resistance; UACR, urinary albumin to creatinine ratio; YKL-40, chitinase-3-like protein; SBP, systolic blood pressure; DBP, diastolic blood pressure.

3.4. Multivariable linear regression models

Significant associations among most VLDL-TG subclasses and GFR, P_GLO, R_E, UACR, and systolic blood pressures were demonstrated after multivariable adjustment for age, sex, and HbA1c (Table 5). There were no associations among IDL-TG and any intraglomerular hemodynamic parameters (Table 8). There were few associations noted across LDL-TG (Table 6) and HDL-TG subclasses (Table 7), most notably in small LDL-TG and HDL-TG. Parallel to the correlation study results, YKL-40 associated positively with the concentrations of all lipoprotein subclasses with exception of large HDL-TG (Tables 5 through 8).

4.0. Discussion

In this cohort of adolescents and young adults with a relatively short duration of T1D, correlation and multivariable linear regression analyses demonstrated strong positive relationships between TG concentrations within lipoprotein subclasses, most notably most VLDL-TG particle sizes, small LDL-TG, and small HDL-TG, and many markers of intraglomerular hemodynamic function including GFR, P_GLO, R_E, and UACR. Positive relationships were also observed with systolic blood pressure and tubular injury marker YKL-40. DKD is a leading source of morbidity and mortality in the T1D population, yet exact mechanisms underlying development and progression of DKD, including the influence of lipid aberrations, have not been fully elucidated. Our findings serve as a valuable contribution to the current understanding of relationships between TG concentrations within lipoprotein particles and kidney function.

In youth, markers of intraglomerular hemodynamic dysfunction including hyperfiltration, hyperperfusion, and increased intraglomerular pressure are detectable early in the T1D disease course and prior to development of diagnostic DKD (elevated albuminuria and/or reduced eGFR) (7, 8), suggesting these findings may represent early DKD pathophysiology and may serve as risk markers. Furthermore, in a cohort of Pima Indian adults with T2D, an elevated R_A/R_E ratio and elevated P_GLO followed by rapid decline demonstrated association with development of kidney failure (28). Elevated markers of tubular injury, such as YKL-40, are also seen concurrently with intraglomerular hemodynamic dysfunction, suggesting a role in the early pathophysiology of kidney dysfunction (8, 9).

Relationships between TGs and the development of microvascular complications is a prominent topic of study in people with T1D, although existing studies in this area have primarily been in adult populations with longer T1D duration and more advanced disease. One such example is an advanced lipoprotein particle analysis of adults with T1D in the landmark DCCT/Epidemiology of Diabetes Interventions and Complications (EDIC) cohort (14), which included men and women with mean diabetes duration of approximately 17 years. Albumin excretion rate was associated with elevated total TG concentrations, predominantly in the smaller VLDL subclasses, and elevated TG-rich lipoprotein subclasses with a shift toward small LDL and HDL particles in both men and women.

Other studies have suggested a relationship between TG-predominant lipoproteins and progression of kidney disease (29, 30). An additional DCCT cohort analysis by Sibley et al. found that participants with T1D and albuminuria had higher total TG concentrations than similar participants who did not have albuminuria, demonstrating an independent relationship between albuminuria and TG concentrations, as well as total cholesterol, LDL-C, and HDL-C (13). Similarly, the FinnDiane study, a large cohort of 3,544 individuals with T1D, found positive associations between both VLDL-TG and VLDL-C and incident albuminuria, progression of microalbuminuria, and progression of macroalbuminuria, with progressive increases in VLDL particle number and TG content across all stages of kidney disease in T1D (11). Our study results align with and further support existing literature investigating relationships between TG and kidney function, as we found strong relationships among markers of intraglomerular hemodynamic function and TG concentrations within lipoprotein subclasses, most notably across most VLDL-TG sizes, as well as small LDL-TG and small HDL-TG.

Strengths of our study include use of a comprehensive metabolic assessment of lipid constituents from a variety of lipoprotein subclasses, as well as gold-standard methods for direct measurement of GFR and RPF. We controlled for dietary intake and physical activity prior to sampling and achieved strict blood glucose control during testing for improved accuracy. This study is limited by the relatively small sample size that was largely comprised of non-Hispanic White persons. However, it is notable that the increased homogeneity of race/ethnicity in our study does accurately represent the general population of adolescents with T1D in the United States (31). While our study sample size did not allow for stratification of results by sex, as has been seen in other large-scale studies like the DCCT/EDIC, analysis of sex differences in outcomes represents an important future direction for large scale, multi-center studies in youth and young adults with T1D. Furthermore, we did not assess markers of inflammation outside of YKL-40, which is also a limitation of this study. Consideration of other inflammatory markers, such as C-Reactive Protein (CRP), may strengthen our understanding of these associations seen with YKL-40, and thus will be considered for future studies.

This study is also limited by its cross-sectional design; therefore, our results are not appropriate for use in assessing for causality. This design confers risk for residual confounding from other components of lipid metabolism or kidney function not directly assessed in this study. Indeed, it is possible that relationships seen between lipids and markers of kidney dysfunction could reflect impact of kidney dysfunction on lipid metabolism, rather than the impact of lipid aberrations on kidney function. Yet, a strength of this study is the short duration of T1D in our participants, intended to target early alterations in kidney physiology. Study participants demonstrated overall retained kidney function with no evidence of clinical kidney disease (microalbuminuria, reduced eGFR). Thus, we speculate that this study’s findings reflect the influence of lipid metabolism perturbations on the development of early intraglomerular hemodynamic dysfunction.

In conclusion, intraglomerular hemodynamic dysfunction is suggested to be an early marker of kidney dysfunction in T1D. Additionally, previous studies in adult T1D populations have demonstrated relationships between markers of kidney disease and both total TG and TG-rich lipoprotein concentrations. This study demonstrates strong associations between TG concentrations within lipoprotein subclasses, particularly the VLDL-TG subclasses, small LDL-TG, and small HDL-TG, and direct measures of intraglomerular hemodynamic function in youth with short duration of T1D, suggesting TG perturbations may serve a role in the pathophysiology of DKD. Advanced analyses of lipoprotein profiles may represent a novel avenue for future risk assessment for kidney dysfunction in T1D or may hold promise as a novel target for therapeutic intervention. Further studies into utilization of advanced analyses of lipoprotein profiles in T1D are warranted.

5.4. Availability of data and material:

The dataset generated and analyzed during this study is available from the corresponding author upon reasonable request.