Early chronotype with metabolic syndrome favours resting and exercise fat oxidation in relation to insulin‐stimulated non‐oxidative glucose disposal

New Findings What is the central question of this study? Chronotype reflects differences in circadian‐mediated metabolic and hormonal profiles. But, does resting and/or exercise fuel use differ in early versus late chronotype as it relates to insulin sensitivity? What are the main finding and its importance? Early chronotypes with metabolic syndrome utilized more fat during rest and exercise independent of aerobic fitness when compared with late chronotypes. Early chronotypes were also more physically active throughout the day. Greater fat use was related to non‐oxidative glucose disposal. These findings suggest that early chronotypes have differences in fuel selection that associate with type 2 diabetes risk. Abstract Early chronotypes (ECs) are often insulin‐sensitive, in part, due to physical activity behaviour. It is unclear, however, if chronotypes differ in resting and/or exercise fuel oxidation in relation to insulin action. Using the Morningness–Eveningness Questionnaire (MEQ), adults with metabolic syndrome (ATP III criteria) were classified as EC (MEQ = 63.7 ± 0.9, n = 24 (19F), 54.2 ± 1.2 years) or late chronotype (LC; MEQ = 47.2 ± 1.4, n = 27 (23F), 55.3 ± 1.5 years). Carbohydrate (CHO) and fat oxidation (FOX, indirect calorimetry) were determined at rest, 55% and 85% V˙O2max, along with heart rate and rating of perceived exertion. Physical activity patterns (accelerometers), body composition (DXA) and insulin sensitivity (clamp, 40 mU/m2/min, 90 mg/dl) with an indirect calorimetry for non‐oxidative glucose disposal (NOGD) were also determined. While demographics were similar, ECs had higher V˙O2max (P = 0.02), NOGD (P < 0.001) and resting FOX (P = 0.02) than LCs. Both groups increased CHO reliance during exercise at 55% and 85% V˙O2max (test effect, P < 0.01) from rest, although ECs used more fat (group effect, P < 0.01). ECs had lower sedentary behaviour and more physical activity during morning/midday (both, P < 0.05). FOX at 55% V˙O2max correlated with V˙O2max (r = 0.425, P = 0.004) whereas FOX at 85% V˙O2max related to NOGD (r = 0.392, P = 0.022). ECs with metabolic syndrome used more fat in relation to insulin‐stimulated NOGD.

(LC) counterparts (Romero-Cabrera et al., 2022;Ruddick-Collins et al., 2018). The exact biological mechanism(s) by which chronotype confers higher disease risk is unknown, but insulin sensitivity is an important aetiological factor in the progression to type 2 diabetes and CVD (Haffner, 2003). Additionally, the inability to switch between lipid and carbohydrate fuel sources from fasted to fed states, known as metabolic inflexibility, may precede insulin resistance (Faerch & Vaag, 2011;Goodpaster & Sparks, 2017).
Carbohydrate is considered the primary energy source during moderate-to-high-intensity exercise in healthy controls, although training is known to lower reliance on glycogen/glucose flux and raise fat oxidation (Friedlander, Casazza, Horning, Buddinger, et al., 1998;Friedlander, et al., 2007). Somewhat discordant from fasting fat oxidation observations is that people with insulin resistance, prediabetes and/or type 2 diabetes have reduced muscle glycogen utilization and higher fat oxidation during exercise than matched healthy counterparts (Braun et al., 2004;Colberg et al., 1996;Goodpaster et al., 2002;Malin et al., 2013). Several factors have been purported to explain this apparent paradox (e.g., lipid oversupply and/or impaired oxidative capacity) (Braun et al., 2004;Colberg et al., 1996;Goodpaster et al., 2002;Malin et al., 2013), but no study to date has examined the impact of chronotype on exercise fuel selection. This is biologically relevant as circadian rhythm is a critical feature in chronic disease risk, in part, through energy metabolism disturbances (Poggiogalle et al., 2018;Zee et al., 2013). In fact, we recently reported that people classified as EC are more insulin-sensitive in relation to insulin-stimulated carbohydrate utilization, a marker of metabolic flexibility . Moreover, we also demonstrated that ECs with metabolic syndrome were categorized as having lower TCA cycle intermediates as well as higher acylcarnitines, which was interpreted as supporting the hypothesis of mitochondrial disturbances in those with LCs .
It remains unknown, however, if fuel selection patterns differ from rest to moderate-or high-intensity exercise in adults based on chronotype.
Furthermore, the characterization of PA across chronotypes is often performed by questionnaire, with limited data via accelerometry to help depict potential fitness-related differences. Therefore, we tested the hypothesis that ECs would use more fat during rest and transition to greater carbohydrate reliance during moderate-and high-intensity exercise when compared with LCs. We also anticipated that greater resting fat oxidation as well as exercise metabolic flexibility (i.e., shift from rest to exercise fuel use) would associate with higher insulin sensitivity as measured by euglycaemic clamp and PA.

New Findings
• What is the central question of this study?
Chronotype reflects differences in circadianmediated metabolic and hormonal profiles. But, does resting and/or exercise fuel use differ in early versus late chronotype as it relates to insulin sensitivity?
• What are the main finding and its importance?
Early chronotypes with metabolic syndrome utilized more fat during rest and exercise independent of aerobic fitness when compared with late chronotypes. Early chronotypes were also more physically active throughout the day.
Greater fat use was related to non-oxidative glucose disposal. These findings suggest that early chronotypes have differences in fuel selection that associate with type 2 diabetes risk. men: ≥102 cm) was also measured up to three times using a plastic tape measure approximately 2 cm above the umbilicus and averaged. Blood pressure (systolic ≥130 mmHg and/or diastolic ≥85 mmHg) was lastly recorded after 5 min of rest in a seated position with feat positioned flat on the floor. A total of three blood pressure recordings occurred with 1-2 min rest in between and the data were averaged. All work conformed to the standards set by the latest version of the Declaration of Helsinki. Participants gave verbal and written informed consent before participating and the study was approved by our Institutional Review Board (IRB no. 19364 and no. Pro2020002029).

Body composition
Total body weight was assessed on a digital scale measured with participants wearing minimal clothing and without shoes. Waist circumference was measured in duplicate 2 cm above the umbilicus using a standard tape measure. Values within 0.5 cm were averaged.
If a value was not within 0.5 cm, a third measure was recorded and averaged. Height was assessed with a stadiometer. Body mass index (BMI) was then calculated. Body fat, fat-free mass (FFM), and visceral adipose tissue (VAT) were measured using dual-energy X-ray absorptiometry (DXA Horizon DXA System, Marlborough, MA, USA) following an appropriate 4-h fast of food and beverage, including water, and abstaining from engaging in physical exercise for at least 4 h.

Maximal aerobic fitness and non-exercise physical activity
Participants performed an incremental to maximal oxygen consumption test (V O 2 max ) on a treadmill. A self-selected walking speed was maintained for the duration of the test. The incline was raised 2.5% every 2 min until volitional exhaustion.V O 2 max was confirmed by at least three of the following four criteria: (1) a respiratory exchange ratio (RER) >1.1, (2) a plateau inV O 2 (<0.150 ml/min change), (3) heart rate (HR) achieved was within 10 beats of the age-predicated HR max (220-age), and/or (4) rating of perceived exertion (RPE) ≥17 a.u. on the Borg scale. Respiratory gases (i.e.,V O 2 andV CO 2 ) were collected throughout the test using indirect calorimetry (Vmax Encore, CareFusion, Yorba Linda, CA, USA). HR via electrocardiogram and RPE was recorded during the last 30 s of each stage. RPE in particular was assessed by investigators through the use of a 6-20 point scale, in which participants were asked to state/point to the level of exertion.V O 2 max was determined as the highestV O 2 value achieved during the incremental test. Participants were also given a three-axis accelerometer (Actigraph, Pensacola, FL, USA) and instructed to wear the device on the right hip. Participants were instructed to wear the device upon rising in the morning and remove the device before retiring to bed, for 7 days. A valid wear time for this measurement was considered a total of 4 days worn with an average of 10 h/day. Data included the mean wear time per day and the total number of days the device was worn. Freedson VM3 ('11) was used to determine the percentage of wear time spent in sedentary, light (LPA), moderate (MPA), or vigorous PA (VPA) per day (Freedson et al., 1998;Troiano et al., 2008). Based on recent work (Qian et al., 2021), groups were divided into morning (06.00-10.59 H), midday (11.00-14.59 h), and afternoon (15.00-19.00 h) to characterize how chronotype impacted sedentary and activity patterns across the day.

Metabolic control
Participants were instructed to refrain from strenuous non-exercise PA as well as consumption of alcohol, caffeine and medications 24 h prior to the study visits. A low-fat American Heart Associationbased diet consisting of 55% carbohydrates, 15% protein and 30% fat, with <10% from saturated fat, was provided to standardize diet prior to metabolic tests. Energy requirements for breakfast, lunch, dinner and snacks were determined via metabolic rate (RMR) tests from indirect calorimetry following an overnight fast and multiplied by a PA factor of 1.2.

Resting and submaximal exercise fuel metabolism
On a separate day, after an overnight fast, participants reported to the laboratory between 07.00 and 09.00 h. Baseline respiratory gases and ventilation were collected for 5 min while standing quietly on the treadmill. Participants then completed two 15-min bouts of exercise at 55% (moderate intensity) and 85% (high intensity) of theirV O 2 max .
Speed and grades were adjusted per individual to target metabolic loads. An approximate 2-min break was provided after the first 15 min to foster successful completion. Respiratory gases were collected using indirect calorimetry during each 15-min stage. During the last 30 s of each stage, HR via telemetry (Polar Electro, Kempele, Finland) and RPE were recorded.V O 2 andV CO 2 were averaged over the last 2 min of each stage for the determination of fuel oxidation as performed before by our group (Gaitán et al., 2019). Calculations are shown here for convenience, including exercise metabolic flexibility: 4. Metabolic flexibility = difference between exercise and resting fuel oxidation

Statistical analysis
Data were analysed using SPSS Statistics (Version 24, IBM Corp., Armonk, NY, USA). An independent two-tailed Student's t-test was used to analyse baseline or rest group differences. A two-(group × test) or three-way factor (group × test × time) repeated measures analysis of variance (ANOVA) was used when appropriate to determine group mean differences for outcomes. Given aerobic fitness was different between chronotype groups, we co-varied for this on fat oxidation and NOGD in an effort to confirm chronotype effects. Pearson's correlation was used to examine associations. Data are means ± SD, and significance was accepted as P ≤ 0.05.

Participant characteristics
There were no significant statistical differences in age, body composition, or ATP III criteria between chronotypes (Table 1).
However, ECs had lower HDL levels and higherV O 2 max when scaled to FFM than LCs (P = 0.017; Table 1). Further, NOGD was significantly higher in ECs versus LCs (P = 0.001, without co-varying) and remained significant afterV O 2 max and LPA adjustment (P = 0.002 and P = 0.03, respectively). Interestingly, in subgroup analysis (n = 31), ECs were more physically active as reflected by lower total sedentary behaviour in ECs versus LCs (72.3 ± 1.2 vs. 76.5 ± 1.6% P = 0.056). In fact, LCs became more sedentary across the day compared with ECs (group × time interaction; P = 0.041; Table 2). Moreover, ECs performed more activity during the morning (group effect, P = 0.054) and midday (11.00-14.59 h, P = 0.043) when compared with LCs (Table 2). There was no difference in VPA between groups.

Resting and submaximal exercise characteristics
HR, RPE, oxygen consumption and energy expenditure were not statistically different during rest, 55% or 85%V O 2 max tests (Table 3).

Resting and submaximal fuel selection
Resting fat oxidation was higher in ECs compared with LCs independent of FFM (t-test, P = 0.021, Figure 1). While both groups increased CHO oxidation during exercise at 55% and 85%V O 2 max Data are means ± SD and reflect the full data set available. Complete data present. Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; FFM, fat-free mass; FPG, fasting plasma glucose, 1 mM = 18 mg/dl; GIR, glucose infusion rate; HDL, high-density lipoproteins, 1 mM = 38.67 mg/dl; NOGD, non-oxidative glucose disposal; SBP, systolic blood pressure; TG, triglycerides, 1 mM = 88.5 mg/dl.
in an intensity-based manner (intensity effect, P = 0.029, Figure 1), ECs maintained greater fat oxidation across all conditions (group effect, P = 0.029, Figure 1). Fat oxidation at rest and during exercise remained significant after co-varying forV O 2 max and LPA adjustment (group effect, P = 0.017 and P = 0.009, respectively). Nonetheless, there were no statistical differences in metabolic flexibility between groups at either 55% or 85%V O 2 max . Indeed, CHO oxidation increased similarly for ECs and LCs at 55% (11.9 ± 1.5 vs. 11.3 ± 1.

DISCUSSION
The principal finding of this study is that fasting and exercise fat oxidation are higher in early compared with later chronotypes with metabolic syndrome, independent of fitness and LPA. Inter-estingly, these findings were also independent of HR, RPE, oxygen consumption and energy expenditure during exercise. This latter observation highlights that workload does not explain differences between chronotypes and that our results are not mediated by sympathetic nervous system activity on fat mobilization and/or hepatic glucose production (Friedlander, Casazza, Horning, Huie, et al., 1998).
Although these results are consistent with suggestions of low fasting fat oxidation in people with obesity as well as type 2 diabetes (Holloway et al., 2009;Houmard, 2008), it is important to highlight that LCs shifted fuel preference towards CHO oxidation during exercise in an equivalent manner to ECs. This suggests that exercise metabolic flexibility was present regardless of ECs using more fat.
In addition, ECs engaged in less sedentary behaviour throughout the day and performed more PA in the morning and midday than LCs.   Data are means ± SD and reflect the full data set available. Values shown in bold indicate statistical significance. Abbreviations: EE, energy expenditure; FFM: fat-free mass; HR, heart rate; RPE, rating of perceived exertion; VE, ventilation;V O 2 , volume of oxygen.
is, at least partially, mediated by mitochondrial capacity, it would be fair to anticipate better fat oxidation in ECs if they had higherV O 2 max . Interestingly, higher aerobic fitness was directly correlated with exercise fat oxidation and metabolic flexibility at 55%V O 2 max . Fat oxidation was scaled to FFM, suggesting that muscle mass per se is not likely a factor explaining differences between chronotypes. Instead, the fitness association to fat oxidation more likely relates to the quality of muscle within ECs. The present findings are in line with prior work by our group highlighting that LCs have elevations in blood lactate and TCA cycle intermediates when compared with ECs that parallel lower metabolic insulin sensitivity . Somewhat surprisingly, though, we report no direct correlation of higher fat oxidation with metabolic insulin sensitivity herein. Instead, fat oxidation at 85% V O 2 max correlated with NOGD. This suggests that mechanisms related to higher fat oxidation during more intense activity may be related to insulin-stimulated glycogen storage. Therefore, it might be speculated that increases in fat utilization at higher intensities 'rescue' the cell from lipid abnormalities (e.g., ceramides/diacylglycerol/incomplete lipids, etc.) related to insulin resistance (Goodpaster & Coen, 2012).
Interestingly, it is worth noting that lipid abnormalities regarding low HDL were not seen in our cohort of LCs as previously described in evening adults (Romero-Cabrera et al., 2022). In fact, HDL was higher in those with LC in our study when compared with ECs. We did not design the study to specifically discern cholesterol kinetics, but a possibility is that our LCs are a mix of intermediate and evening people as reflected by MEQ score averages when compared with prior work that split people into classifications of morning, intermediate and evening (Romero-Cabrera et al., 2022). In either case, future work using stable isotopes and/or biopsies is required to confirm the location of fat metabolism disturbance to improve precision care by chronotype particularly since fat oxidation remained significant between groups after co-varying for fitness and LPA.
Another reason chronotype may impact fuel use relates to adipose tissue accretion (Goodpaster et al., 2002). Exercise stimulates lipolytic activity in visceral fat tissue to a greater extent than gluteofemoral fat mass (Carey et al., 1996;Kanaley et al., 1993). If individuals with LC had less visceral fat than adults with EC, it may restrain plasma-free fatty acid availability and subsequent rise in whole-body fat utilization.
Although we did not measure lipolysis via tracers in this study, we found no differences in BMI, waist circumference, VAT or body-fat percentage between groups. Moreover, we observed no direct relation of body composition with fat oxidation in this study, despite LCs engaging in more sedentary behaviour and this relating to body weight and BMI -a finding consistent with the literature (Henson et al., 2020;Romero-Cabrera et al., 2022;Shechter & St-Onge, 2014). Thus, adiposity is unlikely to explain our results in fuel metabolism.
The capacity to increase CHO reliance during moderate-and highintensity exercise to a similar extent in people with LC versus EC with metabolic syndrome is noteworthy. This rise in CHO oxidation may be either a direct result of elevated CHO oxidation, or a compensatory rise due to decreased fat oxidation. We did not incorporate stable isotopes in this study to discern glucose flux, nor did we determine substrate levels during exercise to assess these possibilities. However, it is possible that LCs had elevated rates of hepatic glucose production that promoted blood glucose availability for increased oxidation in the fasted state. Prior work we conducted showcased that the gluconeogenic precursors lactate and alanine were elevated in people categorized as LC . Interestingly, however, we report here no differences in fasting blood glucose, suggesting that  (Braun et al., 2004;Colberg et al., 1996;Goodpaster et al., 2002;Malin et al., 2013). Together, the mass action effects of glucose per se are unlikely to affect glucose uptake during exercise in LCs (Colberg et al., 1996). In contrast, muscle glycogen use during exercise is related to resting glycogen concentrations (Sherman et al., 1981). If our participants with LC began exercise with higher glycogen concentrations, an increase in glycogen use would be expected. Although muscle biopsies would be needed to rule out the role of muscle glycogen content on CHO oxidation in this study, both groups were fed a mixed diet (e.g., 55% CHO) on the day prior to testing in an effort to meet energy needs. Thus, differences in initial skeletal muscle glycogen content at the onset of exercise are unlikely to explain the lower CHO utilization seen between ECs and LCs during exercise. In fact, the similar exercise metabolic flexibility seen between groups would suggest comparable glycogen levels. Instead, prior work in chronic disease risk populations has suggested lower glycogen utilization during exercise in people with insulin resistance (Braun et al., 2004;Colberg et al., 1996;Goodpaster et al., 2002;Malin et al., 2013 LCs would induce comparable fat oxidation and/or NOGD remains to be seen. We did not directly assess sympathetic nervous system activity in the current study. Instead, we used HR as a surrogate and observed no difference between groups. Although we interpret this to suggest that neural activity was comparable, future work ought to consider approaches that assess sympathetic responses to exercise across chronotypes to enhance physiological understanding. Lastly, women in the present study were primarily post-menopausal between both groups (EC n = 15 vs. LC n = 21) and remaining premenopausal women (EC n = 3 vs. LC n = 4) were reasonably matched, thereby suggesting modest, if any, influence on the findings.
In conclusion, ECs rely more on fat as an energy source during the fasted and moderate-to-high-intensity exercise state compared to LCs in adults with metabolic syndrome. Interestingly, these findings occur independent of workload, suggesting that chronotype may be characterized by unique alterations in metabolism. Indeed, LC participants were more sedentary and had lower aerobic fitness than ECs. Interestingly, more LPA was positively related to metabolic insulin sensitivity, and higher fitness was related to increased exercise fat oxidation. Elevated exercise fat oxidation during high intensity was also associated with non-oxidative glucose metabolism. Collectively, this work highlights and supports chronotype as a potential risk factor related to type 2 diabetes and CVD risk.