Total Postprandial Hepatic Nonesterified and Dietary Fatty Acid Uptake Is Increased and Insufficiently Curbed by Adipose Tissue Fatty Acid Trapping in Prediabetes With Overweight

Excessive uptake of fatty acids (FAs) by lean tissues (e.g., the liver, skeletal muscles, and the heart) plays an important role in the development of insulin resistance (IR), type 2 diabetes (T2D), and its complications. Hepatic FA flux stimulates hepatic glucose production through a number of mechanisms (1), contributing to the development of T2D. Ectopic fat depositions, such as hepatic and myocardial steatosis, are closely associated with cardiometabolic complications (2–6).

The mechanisms of excess lean organ uptake of FA during the progression toward T2D are not fully elucidated. Adipose tissues play a crucial role in the control of systemic FA fluxes (7). First, adipose tissue intracellular triglyceride (TG) lipolysis, which is inhibited by insulin and stimulated by catecholamines, contributes to most of the fasting and postprandial nonesterified FA (NEFA) flux in circulation. In prediabetes and T2D, postprandial NEFA flux is clearly increased (8–10), driven by increased adipose tissue intracellular TG lipolysis (11,12). A second potential source of excessive postprandial NEFA flux is spillover of dietary FAs (DFAs) into the circulating NEFA pool via the lipolysis of chylomicron-TG in the microcirculation of adipose tissues. We recently found that this NEFA spillover does not contribute to the increase in postprandial NEFA flux observed in prediabetes, likely because of an increase in DFA trapping into intra-abdominal adipose tissues (11).

Systemic postprandial NEFA and DFA fluxes have been reported, and organ-specific NEFA uptake during fasting has been studied (13,14), but quantification of total postprandial cardiac and hepatic FA uptake (i.e., NEFA + DFA uptake) has never been performed in humans. Furthermore, the relationship between organ-specific DFA uptake versus NEFA uptake has not been reported. From our recent study showing a reciprocal relationship between postprandial NEFA flux and adipose tissue DFA trapping (11), it is possible that cardiac and hepatic total FA uptake is mitigated by enhanced adipose tissue DFA trapping despite an increase in postprandial NEFA flux in prediabetes. To test this hypothesis, we used dynamic and whole-body positron emission tomography (PET)/computed tomography (CT) metabolic imaging to evaluate cardiac and hepatic NEFA uptake and metabolism and DFA distribution, with measurements of systemic NEFA flux and DFA spillover with stable isotopic tracer methods in subjects without and with prediabetes with overweight (i.e., impaired glucose tolerance [IGT]). In addition to glucose tolerance status and adipose tissue DFA trapping, we also examined other potential determinants of total postprandial cardiac and hepatic FA uptake, including sex, age, obesity, IR, and insulin secretion.

We screened 50 subjects for participation, 43 were included, and 3 subjects dropped out of the study (Supplementary Fig. 1). One participant was found to have fasting glycemia >7.0 mmol after screening and was excluded from the analyses. The analysis included 20 subjects (11 men and 9 women) with IGT—defined as 2 h plasma glucose levels of 7.8–11.1 mmol/L on two separate 75 g oral glucose tolerance tests —and 19 healthy individuals (9 men and 10 women) with normal glucose tolerance (NGT)—defined as fasting glycemia <5.6 mmol/L, 2 h post 75 g glucose tolerance tests <7.8 mmol/L, and HbA_1c <5.8%. Subjects were aged between 47 and 74 years old and had fasting glucose in the nondiabetic range (<7.0 mmol) on screening. All women were postmenopausal. Subjects with active disease, except well-controlled hypertension or mild dyslipidemia, or a contraindication for temporary cessation of antihypertensive drugs or statins were excluded from the study. Subjects demonstrating clinically significant cardiovascular, renal, hepatic, or uncontrolled thyroid disorder, having medical history of other major illnesses, or taking medications affecting lipid or carbohydrate metabolism that could not be rapidly discontinued (e.g., fibrates, thiazolidinediones, or β-blockers), or who participated in any research trial involving radiation exposure within the past 12 months were excluded. Individuals smoking more than one cigarette/day or consuming more than two alcoholic beverages/day were also excluded. Statins (nine subjects) and ezetimibe (one subject) were stopped at least 3 weeks, while metformin (two participants) and antihypertensive drugs (eight under monotherapy [hydrochlorothiazide: one; ACE inhibitor: two; angiotensin II receptor blocker: four], one under dual therapy [hydrochlorothiazide and angiotensin II receptor blocker]) were stopped 7 days before the study days. Results from protocol B0 in 35 of the 39 participants were reported as part of a previous publication (11).

Participants underwent four standardized postprandial protocols (A0, A1, and B0, B1) performed in random order within an average of 57 days (23 to 129 days). The primary outcomes (Clinicaltrials.gov; NCT02808182) of the study were to determine the effect of IGT and sex on plasma NEFA Ra, cardiac and hepatic NEFA uptake and metabolism, adipose tissue NEFA spillover, DFA oxidation, and organ-specific distribution (protocols A0 and B0) (Supplementary Fig. 2). The effect of nicotinic acid administration on these outcomes was also tested (protocols A1 and B1), but is not reported here. Each protocol was performed after 3 days of an isocaloric diet with maintenance of the participants’ regular physical activities and avoidance of strenuous exercise. Upon arrival after a 12 h overnight fast, anthropometric measurements (including height, weight, and waist circumference) were taken. Fat mass and lean mass were quantified using DEXA (Hologic DXA scanner, Hologic, Bedford, MA) and via electrical bioimpedance (Hydra ECF/ICF; Xitron Technologies, San Diego, CA). An intravenous (I.V.) catheter was inserted in the forearm for infusions, while another catheter was placed in the contralateral arm, maintained in a heating pad (∼55°C) for blood sampling. Blood samples were collected in Na₂EDTA-coated tubes containing aprotinin (10 µL/mL; A6279, Sigma-Aldrich, Oakville, Ontario, Canada), which inhibits in vitro proteolysis. At time 0 of each protocol, a standardized 906 kcal liquid meal (400 mL, composed of 33 g of fat, 34 g of protein, and 101 g of carbohydrates; i.e., 33%, 17%, 50% of calories, respectively) was consumed over 20 min (15).

In protocol A0, we used [¹¹C] palmitate PET dynamic acquisition to determine myocardial and hepatic uptake, oxidation and esterification of plasma NEFA, as well as hepatic NEFA release in VLDL-TG. Ninety minutes after ingestion of the liquid meal, [¹¹C]palmitate (∼185 MBq) was administered I.V. over 30 s with dynamic list mode PET scanning for 30 min (18 × 10, 6 × 30, and 6 × 240 s) preceded by CT acquisition (40 mA) centered on the thoracoabdominal segment. The 90 min time point was chosen in order to coincide with the peak of plasma insulin and to avoid PET signal contamination from the meal [¹⁸F] fluoro-thia-heptadecanoic acid (FTHA).

In protocol B0, we measured DFA distribution and spillover using a combination of oral and I.V. stable isotopic tracers and oral [¹⁸F]FTHA (∼70 MBq) PET, as previously described (11,16). Following an I.V. bolus of [1,1,2,3,3-²H]glycerol (1.6 µmol/kg) and NaH[¹³C]O₃ (1.2 µmol/kg; Cambridge Isotopes Laboratories) to prime the bicarbonate pool, a constant I.V. infusion of [7,7,8,8-²H]palmitate (0.01 µmol/kg/min in 100 mL of 25% human albumin; Cambridge Isotopes Laboratories) and [1,1,2,3,3-²H]glycerol (0.05 µmol/kg/min in normal saline; Cambridge Isotopes Laboratories) was administered from time −60 min (in the fasting state) to time +360 min (after meal intake) (10,17,18). To quantify white adipose tissue DFA spillover, subjects also ingested [U-¹³C]palmitate (10.8 µmol/kg; Cambridge Isotopes Laboratories) mixed into the liquid meal (11,19). Blood samples were collected from −60 min to 360 min to determine plasma palmitate, oleate, linoleate levels, and [U-¹³C]palmitate, [7,7,8,8-²H]palmitate and chylomicron TG [U-¹³C]palmitate enrichments. NEFA and glycerol extract was then derivatized with N-methyl-N-tert-butyldimethylsilyltrifluoroacetamide (MTBSTFA) and analyzed on a gas chromatography–tandem mass spectrometer. We use the oral [¹⁸F]FTHA method to determine whole-body DFA distribution with PET (16). Ingested [¹⁸F]FTHA is absorbed in the intestines and reaches the circulation as chylomicron-TG through the thoracic duct, and the subsequent hydrolysis of [¹⁸F]FTHA-containing chylomicron-TG is assumed to be identical to that of naturally occurring chylomicron-TG. Approximately 70 MBq [¹⁸F]FTHA (20) was mixed with olive oil, encapsulated in gel capsules (T.U.B. Enterprises), and administered orally at time 0 min with the liquid meal. At time 360 min, whole-body PET/CT acquisition was performed with the subject supine to measure [¹⁸F]FTHA whole-body distribution (16). Maximal gastrointestinal tract radioactivity exposure for oral [¹⁸F]FTHA administration was estimated at 2.35 mSv at the stomach. Total radioactivity exposure by study participants after all protocols was inferior to 20 mSv. All tracers were tested for sterility and pyrogenicity.

Indirect calorimetry (Vmax 29n, SensorMedics) was performed for 10 min at fasting and every hour after ingestion of the liquid meal. Fasting and postprandial net whole-body carbohydrate and FA oxidation rates were calculated as described previously (21).

For dynamic thoracic scans, regions of interest were drawn in the left ventricular cavity, myocardium, liver, and spleen based on CT images using A Medical Image Data Examiner (AMIDE) 1.0.5 (22). Total tracer activity curves were then obtained using the registered dynamic PET images. Splenic CT radiodensity was subtracted from hepatic radiodensity to yield standardized values for liver radiodensity.

Left ventricular activity was used as the arterial input function for numerical computation of pharmacological constants describing the fluxes of NEFA in multicompartmental models previously described for the heart (23) and liver (13) (Supplementary Fig. 3). Rates of hepatic NEFA uptake, oxidation, esterification, and TG release (in µmol/min/L of tissue) were calculated as previously derived (13):

Constants k are kinetic constants (in min⁻¹) of movements between compartments (Supplementary Fig. 3A). P_NEFA is the plasma concentration of NEFA (in µmol/L).

Similarly, rates of cardiac NEFA uptake, oxidation, and esterification (in µmol/min/L of tissue) were computed using the following formulas (23) (Supplementary Fig. 3B):

Correction for circulating metabolites was performed during the [¹¹C]palmitate acquisition as described in (24) (see Supplementary Methods).

For whole-body scans in protocol B, mean standard uptake values for all tissues of interest were recorded (15,25). Automatic semantic segmentation of the intestines using convolutional neural network (26,27) was used to calculate the fraction of ingested tracer retained in the intestines (DFA_intestine) (see Supplementary Methods). Whole-organ DFA distribution and adipose tissue DFA trapping represent the percentage of the total ingested DFA partitioned between each organ/tissue. These variables were determined, as described previously, for the liver, heart, skeletal muscles, abdominal subcutaneous adipose tissue, peripheral subcutaneous adipose tissue, and visceral adipose tissue (VAT) (11). Total adipose tissue trapping represents the sum of DFA trapping in all adipose tissues.

Plasma appearance rates of palmitate (Ra_palmitate), total NEFA (Ra_NEFA), and glycerol (Ra_glycerol), and plasma NEFA spillover (Ra_{NEFA spillover}) were calculated with the Steele non-steady-state equations (28), as previously described (11). The Ra_NEFA from adipose tissue intracellular lipolysis (Ra_{NEFA lipolysis}) was calculated by subtracting the rate of NEFA spillover from total Ra_NEFA, as previously reported (11). The area under the curve (AUC) was calculated for Ra_NEFA, Ra_{NEFA spillover}, Ra_{NEFA lipolysis}, and Ra_glycerol by integrating over the 6-h postprandial period.

To calculate the total integrated uptake of liver and heart of plasma NEFA (U_NEFA, in mmol) from adipose tissue intracellular lipolysis, we multiplied the whole-organ rate of NEFA uptake (rate of NEFA uptake multiplied by liver and heart volume [V_hepatic and V_cardiac, respectively]) by the fraction of NEFA originating from adipose tissue intracellular lipolysis, and integrated from time (t) 0 to 360 min, using the following equations:

The fractional tissue palmitate uptake rate does not change with change in insulin and NEFA levels (13), whereas change in plasma NEFA and Ra_NEFA were measured and integrated in the equation to extrapolate tissue NEFA uptake rates throughout the 6-h postprandial period. The full rationale for these calculations is presented in the Supplementary Material, last section.

To calculate the total DFA uptake by the heart (cardiac U_DFA, in mmol) during the 6-h postprandial period, we multiplied the total amount of ingested DFA (33 g or 117 mmol of NEFA) in the liquid meal by the fraction of the [¹⁸F]FTHA tracer in the liquid meal that is not retained the intestines (1 − DFA_intestines) and by the cardiac DFA distribution (in %):

Similarly, total DFA uptake by the liver (hepatic U_DFA, in mmol) was calculated, but with the addition of FAs that were secreted by the liver as TG (in mmol), which was in turn computed by integrating the product of the rate of NEFA release in VLDL-TG (hepatic NEFA release in VLDL-TG, in mmol/min/L of tissue) and liver volume (in L) over the 6-h postprandial period:

Uptake of total FAs (U_TFA) was calculated as the sum of organ uptake of both DFA and NEFA:

Finally, the fraction of cardiac and hepatic U_TFA originating from U_DFA was computed by dividing U_DFA by U_TFA.

Glucose, total NEFA, and TG were measured as previously described (29). Plasma C-peptide, insulin, and leptin were measured using Luminex xMAP-based immunoassays (Millipore, Etobicoke, Ontario, Canada). HOMA of IR (HOMA-IR) was also calculated as an alternative index of IR as fasting plasma insulin (units/L) × fasting plasma glucose (mmol/L)/22.5. Insulin secretion rate (ISR) was determined via deconvolution of plasma C-peptide levels with standard two-compartmental kinetic parameters (30). The disposition index (DI), an index of β-cell function, was determined by the product of the postprandial AUC, ISR/AUC glucose, and Matsuda index (25,31).

Data are expressed as mean ± SEM. Two-way ANOVA for repeated measures with NGT versus IGT groups, sex, and interaction as the independent variables was used to analyze the main variables of interest. Parametric correlation was performed after mathematical transformation of variables that failed the Shapiro-Wilk test to normalize their distribution. Multiple linear regression analyses were performed using a stepwise approach to determine the predictors showing significant independent association, with maximum F value and R² of the model (see Supplementary Methods). P values of <0.05 were considered statistically significant. The inclusion of 40 subjects allowed for the detection of a correlation of ρ > 0.48 at β = 0.20 and α = 0.05. All statistical analyses were performed using IBM SPSS Statistics 26.0 for Windows (IBM, Armonk, NY) or GraphPad Prism 8.0 for Windows (GraphPad Software, San Diego, CA).

Informed written consent was obtained, in accordance with the Declaration of Helsinki, and the protocol received approval from the Centre de Recherche du Centre hospitalier universitaire de Sherbrooke Human Ethics of Research Committee. The protocol was registered on the ClinicalTrials.gov platform (NCT02808182).

The data sets generated during and/or analyzed during the current study are available from the corresponding author upon reasonable request. No applicable resources were generated or analyzed during the current study.

Data from 10 NGT women, 9 IGT women, 9 NGT men, and 11 IGT men were analyzed. As shown in Table 1, IGT participants had significantly higher BMI, VAT mass, fasting glucose, postprandial insulin AUC, and fasting TG, and lower Matsuda index (P = 0.03) and tended to have lower liver radiodensity (P = 0.06) than NGT individuals. The IGT and NGT groups displayed no significant difference in HOMA-IR and fasting insulin, but IGT tended to have lower DI (P = 0.07). Women displayed higher HOMA-IR than men (P = 0.04). Independently of sex, IGT participants had higher AUC Ra_{NEFA lipolysis} and AUC Ra_NEFA than NGT individuals. Female participants had marginally higher AUC Ra_glycerol than male participants.

No statistically significant NGT versus IGT group difference was found in cardiac and hepatic rates of U_NEFA, oxidation, and esterification rates, and hepatic rate of NEFA release as TG (Fig. 1). Compared with women, men had increased rates of hepatic NEFA esterification (P = 0.01) (Fig. 1F) and higher rates of postprandial hepatic NEFA release as TG (P = 0.007) (Fig. 1G).

No statistically significant difference was found for DFA distribution between NGT and IGT participants (Fig. 2). Compared with men, women had higher DFA distribution in the liver (P = 0.04) (Fig. 2B) and DFA trapping in the abdominal subcutaneous adipose tissues (P < 0.0001) (Fig. 2E), but lower DFA distribution in skeletal muscles (P = 0.001) (Fig. 2C). Total adipose tissue DFA trapping was also higher (P = 0.04) (Fig. 2G) in women than in men.

Over the 6-h postprandial period, cardiac and hepatic FA U_TFA were primarily driven by U_NEFA. Independently of sex, IGT participants had higher hepatic U_TFA than NGT participants (97 [range 36–215] mmol/6 h vs. 68 [23–132] mmol/6 h, P = 0.03) and hepatic U_NEFA (80 [18–204] vs. 50 [13–113], P = 0.03) and marginally higher cardiac U_TFA (11 [range 4–24] mmol/6 h vs. 8 [3–20] mmol/6 h, P = 0.09); conversely, IGT participants had marginally lower hepatic U_DFA (17 [9–26] vs. 19 [9–34], P = 0.1) (Fig. 3). On average, DFA uptake contributed to 22% (5–47) and 24% (6–52) of the U_TFA in the heart and liver, respectively (Fig. 4). Independently of sex, DFA contributed less to hepatic U_TFA in IGT compared with NGT participants (P = 0.04).

Cardiac and hepatic U_NEFA were associated with obesity (i.e., increased body weight, BMI, waist circumference, fat mass, and VAT mass). Both cardiac and hepatic U_NEFA were positively related to liver fat (reduced hepatic radiodensity), fasting NEFA, AUC Ra_glycerol, and postprandial AUC total energy expenditure. Cardiac U_NEFA was strongly associated with IR (reduced Matsuda index) and peripheral subcutaneous adipose tissue and total adipose tissue DFA trapping. Hepatic U_NEFA showed similar, albeit generally weaker trends (Supplementary Table 1).

Hepatic U_DFA tended to be negatively related to liver fat, and DFA trapping in abdominal subcutaneous, peripheral subcutaneous, and total adipose tissues. There was no significant correlation between U_NEFA and U_DFA in the heart and the liver (ρ = 0.06, P = 0.72 and ρ = 0.09, P = 0.63, respectively).

Cardiac and hepatic U_TFA increased with obesity, fasting NEFA, AUC Ra_glycerol, and AUC total energy expenditure. Cardiac U_TFA was also strongly associated with VAT mass, liver fat, and peripheral subcutaneous and total adipose tissue DFA trapping. However, hepatic U_TFA was only weakly associated with VAT mass and liver fat, but not related to adipose tissue DFA trapping.

In the heart and liver, the percentage of U_TFA originating from U_DFA was negatively associated with weight, BMI, fat mass, waist circumference, VAT mass, fasting NEFA, and DFA trapping in adipose tissues.

To reveal independent predictors of cardiac and hepatic U_NEFA, U_DFA, and U_TFA, stepwise multiple linear regression was performed using factors included in Supplementary Table 1, along with sex, group, age, and AUC Ra_NEFA. Higher waist circumference, AUC Ra_NEFA, fasting NEFA, peripheral subcutaneous adipose tissue DFA trapping, and younger age independently predicted increased cardiac U_NEFA. Increased AUC Ra_NEFA, fasting NEFA, and peripheral subcutaneous DFA trapping were independently associated with increased cardiac U_TFA (Table 2).

In the liver, U_NEFA was predicted by increased BMI, AUC Ra_NEFA, and fasting NEFA. Similarly, fasting NEFA and AUC Ra_NEFA were independently associated with hepatic U_TFA. However, hepatic U_DFA was reduced with increasing muscle DFA distribution, total adipose tissue DFA trapping, and AUC insulin.

In both the liver and heart, higher AUC Ra_NEFA was a negative predictor of the percentage contribution of U_DFA to U_TFA.

The current study reports the first human in vivo assessment of postprandial cardiac and hepatic uptake of NEFA and DFA. Total postprandial hepatic FA uptake was elevated in prediabetes with overweight, independent of sex. Total postprandial cardiac FA uptake also tended to be elevated in prediabetes. In both organs, this increased FA uptake was essentially driven by increased U_NEFA, while U_DFA tended to be lower in prediabetes. Postprandial U_DFA was quantitatively significant and accounted for ∼25%, on average, and up to ∼50% in some individuals, of total cardiac and hepatic FA uptake. While increased adipose tissue DFA trapping was associated with reduced hepatic U_DFA, it was, however, not sufficient to fully curb increased hepatic uptake of FAs in prediabetes with overweight. Furthermore, increased adipose tissue DFA trapping was a significant marker of increased postprandial cardiac FA uptake, but not of decreased cardiac U_DFA, again demonstrating its incapacity to sufficiently blunt cardiac U_DFA to reduce total postprandial FA uptake of this organ.

Two decades ago, increased adipose tissue FA trapping was suggested as a protection mechanism against ectopic fat deposition and the development of IR in lean organs (32,33). Numerous lines of evidence, including experimental preclinical data (34), cross-sectional adipose tissue depot imaging (35), and human genomic studies (36) have supported a role for expansion of adipose tissue storage capacity as an antidiabetic mechanism. We previously demonstrated that increased postprandial NEFA flux from enhanced intracellular adipose tissue lipolysis was associated with adipose tissue DFA trapping in subjects with prediabetes (11). We reported that 7 day high-calorie overfeeding in healthy subjects leads to rapid functional expansion of adipose tissue DFA trapping capacity, with reciprocal reduction of DFA distribution to the heart and muscles during the development of IR, and before significant adipose tissue mass expansion (37). Furthermore, we also showed that increased VAT DFA trapping was closely related with improvement in hepatic insulin sensitivity in patients with T2D within 2 weeks after bariatric surgery (30). These lines of evidence all suggest that adipose tissue DFA trapping is a highly dynamic process that may curb lean organ FA uptake and lipotoxicity in humans. The results of the current study show that adipose tissue DFA trapping increases with increased lean organ FA uptake in prediabetes, but not enough to fully curb this obesity-associated excessive postprandial uptake of FAs.

There was no difference in cardiac or hepatic postprandial FA uptake between women and men. We however found decreased rates of hepatic release of NEFA in VLDL-TG in women compared with men, consistent with previous reports that have shown lower VLDL particles and TG secretion in women (38,39). This may partly explain why hepatic DFA distribution was higher in women participants. Nevertheless, it should be noted that the hepatic rate of VLDL-TG release was estimated using modeling of a radioactive tracer, which has been previously validated only in animal models (13). Removal of the NEFA-VLDL TG hepatic secretion term did not affect our conclusions with regards to hepatic U_DFA and U_TFA.

The current study has limitations. First, all participants were Caucasians, older than 47, almost all were overweight or obese, and all participating women were postmenopausal. Our comparison between men and women therefore may not apply to younger and leaner individuals, and our conclusions on postprandial organ FA uptake in prediabetes may not apply to other ethnic groups.

Furthermore, patients with T2D displayed larger increases in circulating NEFA fluxes than subjects with prediabetes (10) and would therefore be expected to display higher postprandial FA uptake in the heart and liver. More studies are needed to confirm this hypothesis. The absence of statistically significant IGT versus NGT difference (at an α threshold of 0.05) in cardiac U_TFA may be due to the smaller effect size compared with hepatic U_TFA along with small size of the study cohort. Because the field of view of our dynamic scanning procedure after [¹¹C]palmitate injection did not include the portal vein (we had to include the heart and part of the liver in the 18-cm-long field of view), we could not take into account hepatic U_NEFA from the splanchnic circulation in our kinetic modeling. This likely led to an underestimation of the true hepatic U_NEFA rate, but has not affected our quantification of hepatic DFA metabolism or FA uptake in the heart. Likewise, [¹¹C]palmitate PET dynamic acquisition did not include large muscle groups, preventing accurate measure of U_NEFA in skeletal muscles. [¹¹C]palmitate PET also cannot quantify partial (i.e., ketogenesis) versus complete hepatic FA oxidation.

Finally, we have not measured VLDL-TG–mediated FA uptake, which may account for ∼17% of cardiac FA uptake during fasting in humans (40). Our quantitative measure of postprandial cardiac U_TFA is therefore an underestimation, especially in those with higher circulating plasma TG.

In conclusion, subjects with prediabetes and overweight display higher total postprandial hepatic uptake of FAs, caused essentially by an increase in plasma U_NEFA. Hepatic U_DFA tended to be lower and was mitigated by increased adipose tissue DFA trapping. Our data demonstrate that DFA accounts for a significant fraction of hepatic and cardiac postprandial U_TFA, but the increase in adipose tissue DFA trapping with obesity is insufficient to curb the associated lean tissue lipotoxicity in prediabetes with overweight.

Acknowledgments. The authors acknowledge the contributions of Éric Lavallee, Maude Gerard, Caroll-Lynn Thibodeau, and Diane Lessard, all affiliated to the Centre de recherche du Centre hospitalier universitaire de Sherbrooke, for technical support.

Funding. R.Z.Y. receives funding from the Canadian Institutes of Health Research (CIHR Funding Reference Number: 202111FBD-476587-76355). This work was supported by a grant from the Canadian Institutes of Health Research Institute of Nutrition, Metabolism and Diabetes (CIHR Operating Grant MOP53094). A.C.C. holds the Canada Research Chair in Molecular Imaging of Diabetes.

Duality of Interest. A.C.C. received funding from Eli Lilly, HLS Therapeutics, Janssen Inc., Novartis Pharmaceuticals Canada Inc., and Novo Nordisk Canada Inc. as a consultant. None of these commercial relationships are relevant to the current study and all are of less than $10,000 United States. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. R.Z.Y., É.M., C.N., F.F., M.F., L.B., and S.P. acquired data. R.Z.Y., É.M., B.G., É.E.T., and A.C.C. analyzed and interpreted the data. R.Z.Y., É.M., C.N., and A.C.C., redacted the manuscript. All authors revised the final version of the manuscript. A.C.C. designed the study. A.C.C. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.