A heterogeneous response of liver and skeletal muscle fat to the combination of a Paleolithic diet and exercise in obese individuals with type 2 diabetes: a randomised controlled trial

A heterogeneous response of liver and skeletal muscle fat to the combination of a Paleolithic... Aims/hypothesis The aim of the study was to investigate ectopic fat deposition and insulin sensitivity, in a parallel single-blinded randomised controlled trial, comparing Paleolithic diet alone with the combination of Paleolithic diet and exercise in individuals with type 2 diabetes. Methods Thirty-two individuals with type 2 diabetes with BMI 25–40 kg/m and 30–70 years of age followed a Paleolithic diet ad libitum for 12 weeks. In addition, study participants were randomised by computer program to either supervised combined exercise training (PD-EX group) or standard care exercise recommendations (PD group). Staff performing examinations and assessing outcomes were blinded to group assignment. Thirteen participants were analysed in each group: hepatic and peripheral insulin sensitivity were measured using the hyperinsulinaemic–euglycaemic clamp technique combined with [6,6- H ]glucose infusion, and liver fat was assessed by proton magnetic resonance spectroscopy; both analyses were secondary endpoints. Intramyocellular lipid (IMCL) content was measured by magnetic resonance spectroscopy as a secondary analysis. All exam- inations were performed at Umeå University Hospital, Umeå, Sweden. Results Both study groups showed a median body weight loss of 7 kg. Fat mass decreased by 5.7 kg in the PD group and by 6.5 kg in the PD-EX group. Maximum oxygen uptake increased in the PD-EX group only. Liver fat showed a consistent reduction (74% decrease) in the PD group, while the response in the PD-EX group was heterogeneous (p < 0.05 for the difference between groups). IMCL content of the soleus muscle decreased by 40% in the PD group and by 22% in the PD-EX group (p < 0.05 for the difference between groups). Both groups improved their peripheral and adipose tissue insulin sensitivity, but not their hepatic insulin sensitivity. Plasma fetuin-A decreased by 11% in the PD group (p< 0.05) and remained unchanged in the PD-EX group. Liver fat changes during the intervention were correlated with changes in fetuin-A (r = 0.63, p< 0.01). Participants did not report any important adverse events caused by the intervention. Conclusions/interpretation A Paleolithic diet reduced liver fat and IMCL content, while there was a tissue-specific heteroge- neous response to added exercise training. Trial registration ClinicalTrials.gov NCT01513798 Funding Swedish Diabetes Research Foundation, County Council of Västerbotten, Swedish Heart and Lung Foundation, King Gustav V and Queen Victoria’sFoundation . . . . . . Keywords Exercise Hyperinsulinaemic–euglycaemic clamp Insulin sensitivity Intramyocellular fat Liver fat Nutrition . . . Obesity Paleolithic diet Proton magnetic resonance spectroscopy Weight loss Electronic supplementary material The online version of this article (https://doi.org/10.1007/s00125-018-4618-y) contains peer-reviewed but unedited supplementary material, which is available to authorised users. * Julia Otten Department of Food and Nutrition, Umeå University, Umeå, Sweden julia.otten@umu.se Department of Community Medicine and Rehabilitation, Sports Medicine Unit, Umeå University, Umeå, Sweden Department of Public Health and Clinical Medicine, Division of Department of Radiation Sciences, Radiation Physics and Biomedical Medicine, Umeå University, 90185 Umeå, Sweden Engineering, Umeå University, Umeå, Sweden Diabetologia (2018) 61:1548–1559 1549 Abbreviations efficiently improved glucose tolerance in overweight individuals ALT Alanine aminotransferase and in people with type 2 diabetes [10–12]. AST Aspartate aminotransferase In contrast to the well-established relationship between CRP C-reactive protein diet-induced weight loss and decreased liver fat, it remains EGP Endogenous glucose production unclear whether exercise training decreases liver fat indepen- FFM Fat-free mass dently of weight reduction. An earlier intervention study in IMCL Intramyocellular lipid participants with type 2 diabetes reported that 4 months of NAFLD Non-alcoholic fatty liver disease either aerobic or resistance training was associated with a PD group Paleolithic diet group slight decrease in liver fat [13]. However, aerobic exercise PD-EX group Paleolithic diet and exercise group combined with diet intervention does not appear to cause fur- ther liver fat reduction compared with diet intervention alone [14, 15]. To our knowledge, no prior study has investigated liver fat changes associated with diet in combination with both aerobic and resistance training. Interestingly, fetuin-A, a multi- Introduction functional protein secreted from both liver and adipose tissue, has been suggested as a putative link between insulin resistance Fat accumulation outside adipose tissue, i.e. ectopic fat in liver and muscle, is linked to decreased insulin sensitivity and type and liver and adipose tissue function [16, 17]. Several studies show that individuals with obesity, insulin 2diabetes [1, 2]. However, intervention studies are needed to test a putative causal relationship between changes in ectopic resistance and type 2 diabetes have higher intramyocellular lipid deposition and insulin sensitivity. Diet-induced weight lipid (IMCL) content compared with lean and healthy individ- loss in obese individuals is associated with reduction of fat in uals [18, 19]. In obese individuals, weight reduction decreases liver and skeletal muscle [3–5]. This has been linked to im- IMCL content and simultaneously improves insulin sensitivi- proved insulin sensitivity but has not been a universal finding. ty [20, 21]. However, lean endurance-trained athletes exhibit Notably, there are conflicting data regarding the effect of macro- IMCL content measured at rest that is nearly as high as that of people with type 2 diabetes, but with concomitant normal nutrient composition on ectopic fat [3–7]. Two recent studies on obese postmenopausal women found that a Paleolithic diet con- insulin sensitivity, referred to as the athlete’s paradox [18]. Moreover, IMCL content is reduced immediately following sumed ad libitum with a moderately decreased carbohydrate in- take and a high content of mono- and polyunsaturated fatty acids an acute bout of aerobic exercise in young lean individuals [22]. This suggests that IMCL is an important intracellular effectively reduced liver fat [8, 9]. Furthermore, a Paleolithic diet 1550 Diabetologia (2018) 61:1548–1559 source of energy during exercise in people with high insulin was in accordance with the Helsinki Declaration and was ap- sensitivity. This dynamic response does not occur in obese proved by the Regional Ethical Review Board, Umeå, Sweden. individuals, who show unchanged IMCL content after 1 h of cycling [23]. Notably, a 12 week exercise intervention in in- Diet intervention The Paleolithic diet included lean meat, dividuals with type 2 diabetes found an increased IMCL con- eggs, fish, seafood, nuts, fruits and vegetables. Dairy products, tent with concomitantly increased insulin sensitivity [24]. cereals, legumes and added sugar and salt were excluded. Owing to these inconclusive data, it is of major interest to Energy intake was ad libitum. Each study group attended five study how a combination of diet intervention and aerobic and group sessions run by a trained dietitian, and participants resistance training influences ectopic fat deposition and tissue- could contact the dietitian by e-mail or phone between meet- specific insulin sensitivity in individuals with type 2 diabetes. ings. Dietary intake was assessed at baseline and at 12 weeks, We therefore tested the hypothesis that overweight individuals using a 4 day self-reported weighed food record. A trained with type 2 diabetes on a 12 week Paleolithic diet would dietitian converted the food records into estimated energy exhibit a decrease in liver fat and IMCL content, associated and nutrient intakes using the nutritional calculation program with an improvement in hepatic and peripheral insulin sensi- Dietist XP 3.2 (Kost och Näringsdata, Bromma, Sweden). tivity. Moreover, we hypothesised that combined aerobic and resistance exercise training would lead to a further improve- Exercise intervention Prior to randomisation, all participants ment in liver fat and peripheral insulin sensitivity. were advised to perform at least 30 min of moderate exercise daily in accordance with the current guidelines for people with type 2 diabetes. The PD-EX group additionally underwent a training protocol combining aerobic exercise and resistance Methods training in 1 h sessions three times weekly at the Sports Medicine Unit at Umeå University, Umeå, Sweden. Low- Study design Overweight and obese individuals with type 2 intensity aerobic exercise was performed on a cross-trainer, diabetes consumed a Paleolithic diet for 12 weeks. In addition, and moderate- or high-intensity interval training was per- the participants were randomised to receive either supervised formed on a cycle ergometer. Resistance training included exercise training for 3 h per week (PD-EX group) or standard upper and lower body exercises involving multiple muscle care exercise recommendations (PD group). The reduction of groups. All exercise sessions were supervised by personal fat mass was the primary endpoint of this study and has been trainers with a Bachelor of Science degree in sports medicine. published previously [25]. The outcome measurements of this Body composition, liver fat, IMCL, VO and energy article are secondary endpoints (liver fat and peripheral and 2max expenditure Body composition analysis was performed using hepatic insulin sensitivity) and secondary analyses (IMCL). dual-energy x-ray absorptiometry (Lunar Prodigy X-ray Tube Housing Assembly, Brand BX-1L, Model 8743; GE Medical Participants and randomisation We used advertisements in Systems, Madison, WI, USA) at the Clinical Research Centre local newspapers and posters at Umeå University Hospital, Umeå, Sweden, to recruit individuals with type 2 diabetes, at Umeå University Hospital. VO was determined during 2max a standard cardiopulmonary exercise test on a cycle ergometer age 30–70 years and BMI 25–40 kg/m . For inclusion, partici- pantswere requiredtohaveanHbA 48–95 mmol/mol (6.5– at the Department of Clinical Physiology at Umeå University 1c Hospital. Resting energy expenditure was measured by indi- 10.8%) and be treated with diet and/or metformin. Women had to be postmenopausal. Exclusion criteria were smoking, BP >160/ rect calorimetry (Datex-Ohmeda Deltatrac II; Datex-Ohmeda, Madison, WI, USA) and adjusted by subtracting 5% during 100 mmHg, macroalbuminuria, cardiovascular disease, beta blocker use, severe illness and higher levels of training (e.g. 8 h of sleep. Physical activity energy expenditure was estimat- ed using a combined heart rate monitor and accelerometer for moderate endurance training five times a week, resistance train- 7 consecutive days (Actiheart; CamNtech, Cambridge, UK) as ing every other week). A total of 261 individuals were assessed for eligibility, of previously described [28]. Total energy expenditure was cal- culated as the sum of resting energy expenditure and physical whom 32 fulfilled the inclusion criteria and were randomised to the two groups (Fig. 1). Randomisation, using the computer activity energy expenditure with added 10% for diet-induced thermogenesis. program MinimPy version 0.3 [26], was performed by a statis- tician blinded to the study and not involved in data collection or Liver fat and tibialis anterior/soleus muscle fat were analysed by proton magnetic resonance spectroscopy as described in the analysis. Participants were assigned to either the PD group or the PD-EX group, using biased-coin minimisation with an al- electronic supplementary material (ESM) Methods. location ratio of 1:1 [27]. The nurses and technicians who per- formed the examinations were blinded to group affiliation. All Insulin sensitivity and insulin clearance Insulin sensitivity was assessed using the hyperinsulinaemic–euglycaemic clamp participants gave written informed consent. The study protocol Diabetologia (2018) 61:1548–1559 1551 Fig. 1 CONSORT flow diagram Enrolment Assessed for eligibility (n=261) Excluded (n=229) ♦ Did not meet inclusion criteria (n=185) ♦ Declined to participate (n=44) Randomised (n=32) Allocation Allocated to PD (n=16) Allocated to PD-EX (n=16) ♦ Received allocated intervention (n=15) ♦ Received allocated intervention (n=15) ♦ Did not receive allocated intervention (lack of ♦ Did not receive allocated intervention time to participate in diet education) (n=1) (disappointed with allocation) (n=1) Follow-up Discontinued intervention (loss of Discontinued intervention (n=0) motivation) (n=1) Primary analysis Analysed hyperinsulinaemic–euglycaemic clamp Analysed hyperinsulinaemic–euglycaemic (n=13) clamp (n=13) ♦ Did not perform clamp examination owing to ♦ Clamp examination failed (n=1) illness (n=2) Secondary analyses Analysed liver fat (n=12) Analysed liver fat (n=11) ♦ Refused examination (n=1) ♦ Implanted metal (n=2) Analysed tibialis anterior muscle IMCL (n=12) Analysed tibialis anterior muscle IMCL (n=11) ♦ Spectroscopy failed (n=1) ♦ Implanted metal (n=2) Analysed soleus muscle IMCL (n=10) Analysed soleus muscle IMCL (n=11) ♦ Spectroscopy failed (n=3) ♦ Implanted metal (n=2) technique combined with [6,6- H ]glucose infusion as previ- Denmark). Plasma glucose was clamped at 8 mmol/l by infu- ously described [24]. On the day of examination, participants sion of 20% glucose at a variable rate. This plasma glucose came to the Clinical Research Centre at Umeå University level was chosen based on previous hyperinsulinaemic– Hospital in the fasted state having refrained from physical ex- euglycaemic clamp studies [29, 30]. Arterialised blood was ercise for the prior 48 h. A catheter was placed in an antecubital sampled for determination of plasma glucose and [6,6- H ]glu- vein for infusion. For blood sampling, a catheter was placed in cose at t = 0, 150, 160, 170, 180, 330, 340, 350 and 360 min. the contralateral arm, retrograde into a superficial dorsal hand We analysed plasma insulin at t = 0, 180, 240, 300 and vein, with the hand placed in a heated box for arterialisation. 360 min, and NEFA at t = 0, 240, 270, 300, 330 and 360 min. Primed constant infusion of [6,6- H ]glucose (APL, −1 −1 Stockholm, Sweden) at a rate of 0.22 μmol kg min was Gene expression Real-time quantitative PCR was used to de- initiated at t = 0 min and continued until t = 360 min. At t = termine relative gene expression of TNFα and IL6 in subcu- 180 min, we initiated primed constant infusion of short-acting taneous adipose tissue of the abdomen (see ESM Methods for insulin (Actrapid; Novo Nordisk, Bagsværd, Denmark) at a rate further details). −2 −1 of 40 mU m min , which continued until t = 360 min. Between t = 180 min and t = 360 min, blood was sampled ev- Blood sample analysis Blood samples were taken in the ery 5 min for immediate determination of plasma glucose con- fasting state from a peripheral vein, followed by immediate centration (HemoCue 201 RT; Radiometer Medical, Brønshøj, analysis of plasma insulin, HbA , serum triacylglycerols, 1c 1552 Diabetologia (2018) 61:1548–1559 plasma aspartate aminotransferase (AST), plasma alanine performed using Spearman’s rho (r ). A two-sided p value aminotransferase (ALT) and plasma C-reactive protein of <0.05 was considered statistically significant. All statistical (CRP) in the Clinical Chemistry unit at Umeå University analyses were performed using R, version 3.2.2, a language Hospital. For NEFA analysis, plasma was stored at −80°C and environment for statistical computing (R Foundation for and later analysed using the NEFA-HR kit (Wako Statistical Computing, Vienna, Austria). Data are presented as Chemicals, Neuss, Germany). Serum fetuin-A concentrations median (interquartile range). were determined using a human fetuin-A ELISA kit (BioVendor, Brno, Czech Republic). The arterialised venous samples of the hyperinsulinaemic– Results euglycaemic clamp were analysed for [6,6- H ]glucose and unlabelled glucose using GC-MS at the Swedish Participant characteristics and intervention validation The Metabolomics Centre, Umeå, Sweden. The samples (100 μl) main results of the intervention have been previously pub- were extracted with 90% methanol (900 μl), including C -D- lished [25]. In the present substudy, we found no between- glucose as internal standard, then derivatised by addition of group differences in baseline characteristics except a higher freshly prepared acetic anhydride/pyridine (1/1 vol./vol.). fasting glucose in the PD-EX group (Table 1). Median weight Next, the solvent was removed by a stream of N , and there- loss was 7 kg in both groups (Table 1). Fat mass decreased by after the samples were dissolved in ethyl acetate. These sam- 5.7 kg in the PD group and by 6.5 kg in the PD-EX group ples were injected by an Agilent 7693 autosampler (Agilent (Table 1). Food records showed that both study groups simi- Technologies, Atlanta, GA, USA) into an Agilent 7890A gas larly reduced total energy intake during the ad libitum diet chromatograph, and analysed in an Agilent 7010C QQQ mass intervention, mostly by decreasing intake of carbohydrates spectrometer operating in selected ion monitoring mode. The fragment ion m/z 244 was used to detect [6,6- H ]glucose; and saturated fatty acids (Table 2). VO increased by 2 2max m/z 242 was used for unlabelled glucose and m/z 247 for 10% in the PD-EX group and decreased by 3% in the PD 13 2 C -D-glucose. To determine levels of [6,6- H ]glucose and group (p< 0.01 for between-group difference). Total physical 6 2 unlabelled glucose, calibration curves were set up between activity energy expenditure did not change in either group calibrants and internal standard. (Table 2). Calculations Endogenous glucose production (EGP) was de- Insulin sensitivity and insulin clearance Peripheral insulin sen- termined using Steele’s single-pool non-steady-state equation sitivity, measured as the rate of disappearance during the [31]. The glucose distribution volume was estimated as hyperinsulinaemic–euglycaemic clamp, increased by 53% in 0.1625 l/kg body weight. The suppression of EGP (%) was the PD group (p< 0.05) and by 42% in the PD-EX group (p calculated as [EGP basal (t = 150, 160, 170, 180 min) − EGP < 0.01) (Table 3). Hepatic insulin sensitivity, measured as clamp (t = 330, 340, 350, 360 min)] × 100 / EGP basal. To suppression of EGP, remained essentially unchanged during determine the rate of disappearance, the glucose infusion rate the intervention (Table 3). Both study groups showed in- during the last 30 min of clamping was added to the EGP clamp creased NEFA suppression during the hyperinsulinaemic– and corrected for the non-steady-state condition using Steele’s euglycaemic clamp. Adipose tissue insulin sensitivity, calcu- equation. One participant was excluded from calculations of lated as NEFA suppression during the whole clamp examina- EGP and rate of disappearance because the [6,6- H ] tion, increased by 3.4% in both intervention groups (p< 0.05 glucose infusion pump did not work properly during the third in the PD group, p< 0.01 in the PD-EX group) (Table 3). hour of the examination. Suppression of NEFA (%) was calcu- When the insulin sensitivity measures were normalised for lated as [(NEFA at t =0 min) − (NEFA at t = 240, 270, 300, plasma insulin during the clamp, the PD group showed more 330, 360 min)] − 100 / (NEFA at t =0 min) [32]. Insulin clear- pronounced improvement compared with the PD-EX group ance during insulin infusion was calculated by dividing the (Table 3). This was due to increased insulin clearance in the −1 −1 insulin infusion rate [mU (kg FFM) min ], where FFM is PD group (p< 0.05 for the between-group difference; Fig. 2). fat-free mass, by the mean plasma insulin concentration during The intervention groups did not significantly differ in any insulin infusion [32]. measure of insulin sensitivity. Statistical analysis Several variables showed a skewed distri- Fat in soleus and tibialis anterior muscle IMCL content of the bution; thus, we used the Wilcoxon rank-sum test to compare soleus muscle decreased by 40% in the PD group, but showed the treatment effect (change from baseline to 12 weeks) be- no significant change in the PD-EX group (Fig. 3a). Tibialis tween the PD and PD-EX groups. The change over time with- anterior IMCL content did not change significantly in either in each intervention group was determined using the intervention group (data not shown). Tibialis anterior IMCL content at baseline was associated with peripheral insulin Wilcoxon signed-rank test. Correlation analyses were Diabetologia (2018) 61:1548–1559 1553 Table 1 Baseline characteristics, Variable PD group PD-EX group body weight, and fasting blood samples n (male/female) 13 (9/4) 13 (8/5) Age, years 60 (54, 64) 61 (58, 67) Diabetes duration, years 3 (2, 6) 5 (1, 8) Body weight, kg Baseline 90.0 (83.3, 103.2) 97.2 (82.9, 107.4) Change 0–12 weeks −7.1 (−9.8, −5.6)*** −7.0 (−9.7, −5.6)*** BMI, kg/m Baseline 31.4 (29.4, 33.7) 31.4 (29.0, 34.6) Change 0–12 weeks −2.4 (−3.1, −1.8)*** −2.3 (−3.4, −2.2)*** Fat mass, kg Baseline 34.4 (30.1, 37.9) 33.6 (29.2, 39.2) Change 0–12 weeks −5.7 (−8.2, −4.0)*** −6.5 (−8.9, −5.1)*** HbA , mmol/mol 1c Baseline 55 (48, 58) 56 (50, 59) Change 0–12 weeks −11 (−15, −5)** −11 (−18, −7)** HbA ,% 1c Baseline 7.2 (6.5, 7.5) 7.3 (6.7, 7.5) Change 0–12 weeks −1.0 (−1.4, −0.5)** −1.0 (−1.7, −0.6)** Fasting plasma glucose, mmol/l Baseline 8.0 (6.9, 8.5) 8.6 (7.7, 10.5) Change 0–12 weeks −0.9 (−1.8, −0.1)* −2.0 (−3.0, −1.0)** Serum triacylglycerols, mmol/l Baseline 2.4 (1.4, 3.1) 1.7 (1.0, 2.3) Change 0–12 weeks −0.6 (−1.8, −0.2)* −0.4 (−1.0, −0.1)** Plasma NEFA, mmol/l Baseline 0.60 (0.53, 0.78) 0.80 (0.65, 0.90) Change 0–12 weeks 0.03 (−0.02, 0.20) −0.04 (−0.12, 0.17) Plasma AST, μkat/l Baseline 0.59 (0.53, 0.71) 0.55 (0.51, 0.66) Change 0–12 weeks −0.05 (−0.14, 0.14) 0.02 (−0.20, 0.22) Plasma ALT, μkat/l Baseline 0.67 (0.53, 0.91) 0.54 (0.48, 0.78) Change 0–12 weeks −0.12 (−0.34, −0.08)** −0.07 (−0.16, 0.07) Data are reported as median (interquartile range) *p< 0.05, **p< 0.01, ***p< 0.001 for the within-group change over time from baseline to 12 weeks p< 0.05 between the PD and PD-EX groups sensitivity (r = −0.45, p< 0.05) and fasting plasma NEFA (r in liver fat in the PD-EX group was 43% (p< 0.05), with no S S =0.59, p< 0.05). Changes in tibialis and soleus muscle significant difference between the intervention groups (p=0.08). IMCL content during the intervention were not correlated with Excluding the three outliers did not alter the results of the other changes in peripheral insulin sensitivity. statistical analyses (ESM Table 1). The three individuals whose liver fat increased during the intervention had a greater decrease in soleus muscle IMCL content compared with the other partic- Liver fat All participants in the PD group decreased their liver fat, while the intervention response was more heterogeneous ipants in the PD-EX group (ESM Table 1). VO ,plasma 2max in the PD-EX group (Fig. 3b). Median hepatic lipid reduction triacylglycerols, plasma AST, plasma ALTand plasma CRP were was 74% for the PD group and 32% for the PD-EX group (p improved or unchanged in these three individuals (ESM Table 1). < 0.05 for the difference between groups). Liver fat was not associated with hepatic insulin sensitivity Liver fat increased substantially in three individuals in the PD- at baseline. The decrease in liver fat during the intervention EX group. After exclusion of these outliers, the median decrease was correlated with an improvement in hepatic insulin 1554 Diabetologia (2018) 61:1548–1559 Table 2 Energy balance and die- Variable PD group (n = 12) PD-EX group (n =12) tary intake Energy intake, kJ/day Baseline 8330 (6204, 10,778) 6673 (5569, 9443) Change 0–12 weeks −1377 (−3284, −1025)** −2155 (−3330, −649)** Total energy expenditure, kJ/day Baseline 12,619 (11,129, 13,933) 12,485 (9196, 16,778) Change 0–12 weeks −954 (−1485, −285)* −1305 (−2414, 1201) Resting energy expenditure, kJ/day Baseline 6791 (6184, 7243) 7268 (5565, 7958) Change 0–12 weeks −510 (−774, −188)** −381 (−715, −92)** Physical activity energy expenditure, kJ/day Baseline 4276 (3615, 5519) 4201 (3008, 7201) Change 0–12 weeks −117 (−870, 372) −88 (−1686, 1640) Protein, g/day Baseline 80 (69, 95) 77 (63, 106) Change 0–12 weeks 5 (−17, 23) 1 (−16, 14) −1 −1 Protein, g kg day Baseline 0.85 (0.66, 1.14) 0.78 (0.73, 1.02) Change 0–12 weeks 0.10 (−0.09, 0.35) 0.06 (−0.12, 0.25) Carbohydrate, g/day Baseline 204 (148, 280) 169 (152, 197) Change 0–12 weeks −89 (−122, −49)** −92 (−117, −67)** Total fat, g/day Baseline 84 (58, 115) 64 (46, 98) Change 0–12 weeks −12 (−38, 8) −10 (−33, 24) Saturated fatty acids, g/day Baseline 31 (21, 48) 25 (19, 35) Change 0–12 weeks −14 (−33, −5)** −12 (−23, −6)** Monounsaturated fatty acids, g/day Baseline 32 (25, 41) 26 (16, 37) Change 0–12 weeks 4 (−16, 14) 5 (−6, 18) Polyunsaturated fatty acids, g/day Baseline 11 (9, 14) 9 (7, 16) Change 0–12 weeks 1 (−5, 5) 1 (−3, 7) Protein, E% Baseline 17 (14, 19) 18 (17, 20) Change 0–12 weeks 7 (4, 11)** 6 (3, 12)** Carbohydrate, E% Baseline 41 (38, 46) 45 (32, 49) Change 0–12 weeks −10 (−18, −3)* −14 (−21, −7)** Total fat, E% Baseline 40 (36, 41) 32 (31, 44) Change 0–12 weeks 6 (−6, 11) 9 (2, 14)* Saturated fatty acids, E% Baseline 15 (13, 18) 13 (12, 17) Change 0–12 weeks −5(−8, −3)** −4(−7, −2)** Monounsaturated fatty acids, E% Baseline 16 (14, 17) 12 (10, 17) Change 0–12 weeks 5 (−3, 11) 10 (5, 12)** Polyunsaturated fatty acids, E% Baseline 5.0 (4.7, 6.4) 5.6 (3.8, 6.4) Change 0–12 weeks 2.1 (−0.5, 3.6)* 3.2 (0.9, 3.7)** Data are reported as median (interquartile range) *p< 0.05, **p< 0.01 for the within-group change over time from baseline to 12 weeks E%, energy per cent Diabetologia (2018) 61:1548–1559 1555 Table 3 Insulin sensitivity Insulin sensitivity PD group PP-EX group Peripheral insulin sensitivity −1 −1 Rate of disappearance, mg kg min Baseline 3.79 (2.95, 4.23) 3.87 (3.02, 5.26) Change 0–12 weeks 2.05 (0.32, 3.59)* 1.15 (0.67, 2.66)** −1 −1 Rate of disappearance/insulin, μgkg min per mU/l Baseline 34.2 (28.6, 49.2) 46.2 (35.0, 74.5) Change 0–12 weeks 28.9 (11.8, 61.5)** 14.6 (5.0, 30.3)** Hepatic insulin sensitivity −1 −1 EGP, mg kg min Baseline 1.81 (1.56, 1.99) 1.78 (1.51, 2.49) Change 0–12 weeks 0.04 (−0.06, 0.55) 0.11 (−0.18, 0.60) Suppression of EGP, % Baseline 96 (83, 128) 114 (85, 121) Change 0–12 weeks 13 (−10, 44) 11 (−33, 35) Suppression of EGP/insulin, % per mU/l Baseline 0.98 (0.77, 1.36) 1.51 (0.97, 1.74) Change 0–12 weeks 0.22 (−0.03, 0.91)* 0.06 (−0.45, 0.55) Adipose tissue insulin sensitivity Suppression of NEFA, % Baseline 88 (80, 93) 89 (85, 93) Change 0–12 weeks 3.4 (1.3, 5.7)* 3.4 (0.2, 7.8)** Suppression of NEFA/insulin, % per mU/l Baseline 0.83 (0.74, 1.08) 1.12 (0.95, 1.27) Change 0–12 weeks 0.14 (0.03, 0.31)** 0.11 (−0.06, 0.20) Data are reported as median (interquartile range) *p< 0.05, **p< 0.01 for the within-group change over time from baseline to 12 weeks p< 0.05 between the PD and PD-EX groups sensitivity in the PD group (r = −0.62, p< 0.05) but not in the tissue was associated with the change in fasting plasma PD-EX group. Liver fat was associated with adipose tissue NEFA (r =0.45, p< 0.05). insulin sensitivity (NEFA suppression during the clamp) at baseline (r = −0.58, p< 0.01). Discussion Plasma fetuin-A Plasma fetuin-A decreased by 11% in the PD group (p< 0.05) and remained unchanged in the PD-EX This intervention with a Paleolithic diet in overweight indi- group (Fig. 4). Liver fat changes during the intervention were strongly correlated with changes in fetuin-A (r = 0.63, p viduals with type 2 diabetes showed decreased ectopic lipid accumulation in liver and soleus muscle, as well as improved < 0.01; Fig. 5). Changes in adipose tissue insulin sensitivity (suppression of NEFA/insulin) were also associated with peripheral insulin sensitivity. On a group level, the addition of combined resistance and aerobic exercise training to the diet changes in fetuin-A (r = 0.51, p< 0.01). By contrast, we intervention reduced the effect on muscle and liver fat content. found no association between changes in fetuin-A levels and This was due to the considerable heterogeneity in response to changes in hepatic or peripheral sensitivity. The three partici- exercise. Decreased liver fat during the intervention was pants in the PD-EX group who showed increased liver fat also strongly associated with reduction in plasma fetuin-A levels showed increased plasma fetuin-A levels (Fig. 5). in both intervention groups. This was linked to an improve- Inflammation in plasma and adipose tissue Plasma CRP de- ment in adipose tissue insulin sensitivity. creased in the PD-EX group (p< 0.05; Table 4). In adipose Current guidelines for non-alcoholic fatty liver disease tissue, IL6 and TNFα gene expression did not change signif- (NAFLD) recommend lifestyle interventions involving diet icantly during the intervention. During the intervention, the and exercisetodecreaseliver fat [33]. However, earlier studies change in TNFα gene expression in subcutaneous adipose suggest a complex relationship between liver fat and the effect 1556 Diabetologia (2018) 61:1548–1559 ** Baseline 12 weeks Baseline 12 weeks PD PD-EX 0 b Baseline 12 weeks Baseline 12 weeks PD PD-EX *** Fig. 2 Insulin clearance during 12 weeks of intervention in the PD and PD-EX groups. *p< 0.05 for the within-group change over time from baseline to 12 weeks. p< 0.05 for the intervention effect between the PD group and the PD-EX group of lifestyle interventions. Two short-term studies showed that weight reduction by a low-carbohydrate diet effectively re- duced liver fat [4, 5]. We found a reduction in liver fat after a Paleolithic diet with a moderately reduced carbohydrate content. We anticipated that our combined intervention in in- dividuals with type 2 diabetes would show an additional fat- decreasing effect on the liver, as a recent meta-analysis con- Baseline 12 weeks Baseline 12 weeks PD PD-EX cluded that exercise reduces hepatic fat content [34]. Fig. 3 IMCL content of the soleus muscle (a) and liver fat (b)during Unexpectedly, three individuals in the PD-EX group showed 12 weeks of intervention in the PD and PD-EX groups. IMCL is normal- a clear increase in liver fat. After exclusion of these three ised to the creatine concentration of the muscle. **p< 0.01, ***p< 0.001 individuals, we found that liver fat decreased significantly in for the within-group change over time from baseline to 12 weeks. both study groups, with no difference between groups, while p< 0.05 for the intervention effect between the PD group and the PD- EX group all other comparisons were unaltered. There may be several explanations for the unexpected response regarding liver fat in some individuals, including a prolonged exercise-induced in- crease in liver fat in some individuals or increased metabolic flexibility. The three individuals whose liver fat increased dur- ing 12 weeks of exercise thus showed decreased or unchanged inflammation (plasma CRP levels), triacylglycerol levels and liver enzymes, indicating improved metabolic health in this subgroup. A third possibility is that the participants did not refrain from exercise for 48 h before the examination as they were supposed to, leading to a decrease in muscle fat and an increase in liver fat. In healthy people with normal weight and in overweight individuals, an acute bout of aerobic exercise immediately increases liver fat [35, 36]. This exercise-induced increase in Baseline 12 weeks Baseline 12 weeks liver fat seems to be mainly due to the rise in plasma NEFA PD PD-EX during and after exercise [35, 36]. All participants in our study Fig. 4 Plasma fetuin-A levels during 12 weeks of intervention in the PD were told to refrain from exercise for 48 h prior to liver fat and PD-EX groups. *p< 0.05 for the within-group change over time examination. A possible explanation for the conundrum could from baseline to 12 weeks -1 -1 Insulin clearance during insulin infusion (ml [kg FFM] min ) Fetuin-A (μg/ml) Liver fat (%) Soleus muscle IMCL/creatine Diabetologia (2018) 61:1548–1559 1557 substrate during exercise and is replenished during recovery. Endurance athletes who performed 3 h of cycling exercise had a 20% decrease in IMCL content in the legs and a simultaneous 38% increase in IMCL content in the non-exercising arms [22]. Importantly, obesity and type 2 diabetes are associated with a lack of this dynamic response to exercise, which may be related to continuously increased plasma NEFA levels [23]. However, if adipose tissue lipolysis is reduced with a nicotinic acid analogue, r = 0.63 -200 p < 0.01 plasma NEFA levels are reduced and the decrease in IMCL content during one bout of exercise is more pronounced [41]. Another contributing factor to the heterogeneous response in -400 -30 -20 -10 0 10 20 the combined diet and exercise intervention is the intake of car- Liver fat change 0–12 weeks (%) bohydrates. Suppression of adipose tissue lipolysis through in- Fig. 5 Association between change in liver fat and change in plasma sulin administration or carbohydrate ingestion leads to a reduc- fetuin-A during the intervention tion in plasma NEFA levels. Indeed, glucose ingestion during exercise causes plasma NEFA levels to decrease below fasting be that an exercise-induced increase in liver fat may last longer levels during and after exercise [36, 42]. Accordingly, glucose than 2 days in some individuals. supplementation during and after cycling prevents an increase in Lipolysis and inflammation are closely linked [37]. During liver fat during the recovery phase [36]. Furthermore, studies our intervention, decreasing TNFα expression in adipose tis- with isoenergetic diets show a decrease in liver fat only with sue was associated with decreasing plasma NEFA levels. high-carbohydrate diets, not with high-fat diets [6, 7]. Moreover, we found an association between liver fat and sup- Although energy intake was ad libitum in our study, participants pressibility of NEFA production, highlighting the importance reported decreases in carbohydrate and total energy intake. More of plasma NEFA concentration for hepatic lipid content. In detailed studies of macronutrient intake in relation to exercise are NAFLD, most triacylglycerols in the fatty liver originate from therefore of interest regarding the effects on hepatic lipid content. plasma NEFA, and most plasma NEFA originate from adipose Fetuin-A is secreted mainly from liver and adipose tissue tissue [38]. Plasma NEFA uptake in liver cells and esterifica- and is elevated in type 2 diabetes and NAFLD [17, 43, 44]. tion into hepatic triacylglycerols are insulin-independent, de- Circulating fetuin-A levels have been associated with severity pending only on the plasma NEFA concentration [39]. of liver steatosis, independently of insulin resistance, and with Notably, the percentage of plasma NEFA taken up by the liver non-alcoholic steatohepatitis [45, 46]. Our results showed a remains constant both during and after exercise [40]. strong association between fetuin-A levels and changes in Plasma NEFA concentrations are also important for IMCL liver fat content. Changes in circulating fetuin-A levels were content, and we found that fasting plasma NEFA levels were also associated with the ability to suppress NEFA production. closely associated with IMCL content of the tibialis anterior Since fetuin-A can be secreted by both hepatocytes and adi- muscle. In healthy individuals, IMCL is used as an energy pocytes, it remains unclear whether fetuin-A secreted by the liver influences adipose tissue or the other way round. The PD group increased insulin clearance, which might Table 4 Inflammatory markers in plasma and subcutaneous adipose have been due to the improvement in liver fat content. tissue Hepatocytes thus show impaired insulin clearance when load- Marker PD group PD-EX group ed with triacylglycerols in vitro [47], and liver fat is inversely related to insulin clearance in vivo [48]. Plasma CRP, nmol/l Improvement of hepatic insulin sensitivity was less pro- Baseline 11 (6, 27) 13 (6, 23) nounced in our intervention: only if normalised by insulin, Change 0–12 weeks −2(−9, 1) −3(−8, 0)* the PD group increased hepatic insulin sensitivity. This may IL6 in s.c. adipose tissue, relative expression of mRNA relate to the relatively well-preserved hepatic insulin sensitivity Baseline 0.81 (0.42, 1.16) 0.79 (0.61, 1.17) in our study cohort. In most studies, diet-induced weight loss in Change 0–12 weeks 0.15 (−0.45, 0.62) −0.14 (−0.42, 0.01) people with type 2 diabetes causes an increase in hepatic insulin TNFα in s.c. adipose tissue, relative expression of mRNA sensitivity, but some authors report it unchanged [49, 50]. Baseline 2.19 (1.79, 3.03) 2.45 (1.89, 3.81) A limitation of our study is that the insulin dose might have Change 0–12 weeks −0.06 (−0.53, 0.35) −0.70 (−1.02, 0.32) been too high to detect changes in hepatic insulin sensitivity. Moreover, target plasma glucose during the euglycaemic clamp Data are reported as median (interquartile range) was 8 mmol/l, which may not represent euglycaemia, especially *p< 0.05 for the within-group change over time from baseline to 12 weeks after the intervention when fasting glucose was normalised. A Fetuin-A change 0–12 weeks (μg/ml) 1558 Diabetologia (2018) 61:1548–1559 Open Access This article is distributed under the terms of the Creative lower insulin dose and a lower glucose target during the clamp Commons Attribution 4.0 International License (http:// studies might have detected more subtle changes in hepatic in- creativecommons.org/licenses/by/4.0/), which permits unrestricted use, sulin sensitivity. Indeed, we have previously demonstrated an distribution, and reproduction in any medium, provided you give appro- improvement in HOMA-IR after 5 weeks and 6 months follow- priate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. ing a Paleolithic diet in healthy overweight participants [8, 9]. Another limitation is that we had to exclude soleus References muscle measurements from three participants because we could not separate the intramyocellular and 1. Cusi K (2016) Treatment of patients with type 2 diabetes and non- extramyocellular lipid signals. This is a known technical alcoholic fatty liver disease: current approaches and future direc- difficulty related to the fact that soleus muscle fibres are tions. Diabetologia 59:1012–1020 not aligned in parallel to the main magnetic field. This 2. Perseghin G, Scifo P, De Cobelli F et al (1999) Intramyocellular shortage of data limits our ability to draw conclusions triglyceride content is a determinant of in vivo insulin resistance in humans: a 1H-13C nuclear magnetic resonance spectroscopy as- regarding the effects of the intervention on different sessment in offspring of type 2 diabetic parents. Diabetes 48: skeletal muscle types. Finally, analyses of gene variants 1600–1606 that may influence liver fat accumulation, e.g. PNPLA3, 3. Haufe S, Engeli S, Kast P et al (2011) Randomized comparison of are of interest in future intervention studies. reduced fat and reduced carbohydrate hypocaloric diets on intrahepatic fat in overweight and obese human subjects. In conclusion, our results indicate that an exercise interven- Hepatology 53:1504–1514 tion is associated with a heterogeneous response in liver fat 4. Browning JD, Baker JA, Rogers T, Davis J, Satapati S, Burgess SC content in obese individuals with type 2 diabetes, despite im- (2011) Short-term weight loss and hepatic triglyceride reduction: proved metabolic health. Further studies are needed to under- evidence of a metabolic advantage with dietary carbohydrate re- striction. Am J Clin Nutr 93:1048–1052 stand how exercise changes liver fat and hepatic insulin sen- 5. Kirk E, Reeds DN, Finck BN, Mayurranjan SM, Patterson BW, Klein sitivity in relation to energy balance and macronutrient intake S (2009) Dietary fat and carbohydrates differentially alter insulin sen- among individuals with obesity and type 2 diabetes. sitivity during caloric restriction. Gastroenterology 136:1552–1560 6. Westerbacka J, Lammi K, Hakkinen AM et al (2005) Dietary fat content modifies liver fat in overweight nondiabetic subjects. J Clin Acknowledgements The authors thank the study participants; the re- Endocrinol Metab 90:2804–2809 search nurses I. Arnesjö, K. Iselid, L. Uddståhl, C. Ring and L. H. 7. van Herpen NA, Schrauwen-Hinderling VB, Schaart G, Mensink RP, Bergman at the Clinical Research Centre, (Umeå University Hospital, Schrauwen P (2011) Three weeks on a high-fat diet increases Umeå, Sweden) for performing the clamp examinations; A. Tellström intrahepatic lipid accumulation and decreases metabolic flexibility in (Department of Public Health and Clinical Medicine, Umeå University, healthy overweight men. J Clin Endocrinol Metab 96:E691–E695 Umeå, Sweden) for guiding participants during the diet intervention; and 8. Ryberg M, Sandberg S, Mellberg C et al (2013) A Palaeolithic-type K. Lundgren (Swedish Metabolomics Centre, Swedish University of diet causes strong tissue-specific effects on ectopic fat deposition in Agricultural Sciences, Umeå, Sweden) for the GC-MS analysis of obese postmenopausal women. J Intern Med 274:67–76 [6,6- H ]glucose. Parts of this paper were presented at the 52nd Annual 9. Otten J, Mellberg C, Ryberg M et al (2016) Strong and persistent Meeting of the European Association for the Study of Diabetes, Munich, effect on liver fat with a Paleolithic diet during a two-year interven- Germany, 12–16 September 2016. tion. Int J Obes 40:747–753 10. Lindeberg S, Jonsson T, Granfeldt Y et al (2007) A Palaeolithic diet Data availability Data are available on request from the authors. improves glucose tolerance more than a Mediterranean-like diet in individuals with ischaemic heart disease. Diabetologia 50:1795–1807 Funding This study was supported by grants from the Swedish Diabetes 11. Jonsson T, Granfeldt Y, Ahren B et al (2009) Beneficial effects of a Research Foundation (2014-096), the County Council of Västerbotten Paleolithic diet on cardiovascular risk factors in type 2 diabetes: a (VLL-460481), the Swedish Heart and Lung Foundation (20120450), randomized cross-over pilot study. Cardiovasc Diabetol 8:35 and King Gustav V and Queen Victoria’sFoundation. 12. Frassetto LA, Schloetter M, Mietus-Synder M, Morris RC Jr, Sebastian Duality of interest The authors declare that there is no duality of interest A (2009) Metabolic and physiologic improvements from consuming a associated with this manuscript. paleolithic, hunter-gatherer type diet. Eur J Clin Nutr 63:947–955 13. Bacchi E, Negri C, Targher G et al (2013) Both resistance training Contribution statement JO designed the study, recruited the participants, and aerobic training reduce hepatic fat content in type 2 diabetic collected the data, performed the statistical analyses and wrote the manu- subjects with nonalcoholic fatty liver disease (the RAED2 script. AS, MW, MR, MS and JH designed the study, interpreted the data Randomized Trial). Hepatology 58:1287–1295 and revised the manuscript. MW implemented the dietary intervention. AI 14. Tamura Y, Tanaka Y, Sato F et al (2005) Effects of diet and exercise on conducted the exercise intervention, analysed and interpreted the data and muscle and liver intracellular lipid contents and insulin sensitivity in drafted the article. IS analysed fetuin-A, TNFα and IL6 and drafted the type 2 diabetic patients. J Clin Endocrinol Metab 90:3191–3196 article. JH analysed the spectroscopy data. TO designed the study, 15. Bozzetto L, Prinster A, Annuzzi G et al (2012) Liver fat is reduced interpreted the data and wrote the manuscript. All authors gave their final by an isoenergetic MUFA diet in a controlled randomized study in approval of the manuscript to be published. JO is the guarantor of this work type 2 diabetic patients. Diabetes Care 35:1429–1435 and, as such, had full access to all the data in the study and takes respon- 16. Stefan N, Haring HU (2013) Circulating fetuin-A and free fatty acids sibility for the integrity of the data and the accuracy of the data analysis. interact to predict insulin resistance in humans. Nat Med 19:394–395 17. Trepanowski JF, Mey J, Varady KA (2015) Fetuin-A: a novel link between obesity and related complications. Int J Obes 39:734–741 Diabetologia (2018) 61:1548–1559 1559 18. Goodpaster BH, He J, Watkins S, Kelley DE (2001) Skeletal mus- Association for the Study of Liver Diseases, and American College of Gastroenterology. Gastroenterology 142:1592–1609 cle lipid content and insulin resistance: evidence for a paradox in endurance-trained athletes. J Clin Endocrinol Metab 86:5755–5761 34. Golabi P, Locklear CT, Austin P et al (2016) Effectiveness of exer- 19. Falholt K, Jensen I, Lindkaer Jensen S et al (1988) Carbohydrate cise in hepatic fat mobilization in non-alcoholic fatty liver disease: and lipid metabolism of skeletal muscle in type 2 diabetic patients. systematic review. World J Gastroenterol 22:6318–6327 Diabet Med 5:27–31 35. Egger A, Kreis R, Allemann S et al (2013) The effect of aerobic 20. Greco AV, Mingrone G, Giancaterini A et al (2002) Insulin resis- exercise on intrahepatocellular and intramyocellular lipids in tance in morbid obesity: reversal with intramyocellular fat deple- healthy subjects. PLoS One 8:e70865 tion. Diabetes 51:144–151 36. Bilet L, Brouwers B, van Ewijk PA et al (2015) Acute exercise does not 21. Gray RE, Tanner CJ, Pories WJ, MacDonald KG, Houmard JA decrease liver fat in men with overweight or NAFLD. Sci Rep 5:9709 (2003) Effect of weight loss on muscle lipid content in morbidly 37. Chen Z, Yu R, Xiong Y, Du F, Zhu S (2017) A vicious circle obese subjects. Am J Physiol Endocrinol Metab 284:E726–E732 between insulin resistance and inflammation in nonalcoholic fatty 22. Schrauwen-Hinderling VB, van Loon LJ, Koopman R, Nicolay liver disease. Lipids Health Dis 16:203 K, Saris WH, Kooi ME (2003) Intramyocellular lipid content is 38. Donnelly KL, Smith CI, Schwarzenberg SJ, Jessurun J, Boldt MD, increased after exercise in nonexercising human skeletal muscle. Parks EJ (2005) Sources of fatty acids stored in liver and secreted J Appl Physiol (1985) 95:2328–2332 via lipoproteins in patients with nonalcoholic fatty liver disease. J 23. Ipavec-Levasseur S, Croci I, Choquette S et al (2015) Effect of 1-h Clin Invest 115:1343–1351 moderate-intensity aerobic exercise on intramyocellular lipids in 39. Vatner DF, Majumdar SK, Kumashiro N et al (2015) Insulin- obese men before and after a lifestyle intervention. Appl Physiol independent regulation of hepatic triglyceride synthesis by fatty Nutr Metab 40:1262–1268 acids. Proc Natl Acad Sci U S A 112:1143–1148 24. Meex RC, Schrauwen-Hinderling VB, Moonen-Kornips E et al 40. Van Hall G, Bulow J, Sacchetti M, Al Mulla N, Lyngso D, (2010) Restoration of muscle mitochondrial function and metabolic Simonsen L (2002) Regional fat metabolism in human splanchnic flexibility in type 2 diabetes by exercise training is paralleled by and adipose tissues; the effect of exercise. J Physiol 543:1033–1046 increased myocellular fat storage and improved insulin sensitivity. 41. van Loon LJ, Manders RJ, Koopman R et al (2005) Inhibition of Diabetes 59:572–579 adipose tissue lipolysis increases intramuscular lipid use in type 2 25. Otten J, Stomby A, Waling M et al (2017) Benefits of a Paleolithic diabetic patients. Diabetologia 48:2097–2107 diet with and without supervised exercise on fat mass, insulin sen- 42. Schrauwen P, Hesselink MK, Vaartjes I et al (2002) Effect of acute sitivity, and glycemic control: a randomized controlled trial in indi- exercise on uncoupling protein 3 is a fat metabolism-mediated ef- viduals with type 2 diabetes. Diabetes Metab Res Rev https://doi. fect. Am J Physiol Endocrinol Metab 282:E11–E17 org/10.1002/dmrr.2828 43. Ix JH, Wassel CL, Kanaya AM et al (2008) Fetuin-A and incident 26. Saghaei M, Saghaei S (2011) Implementation of an open-source diabetes mellitus in older persons. JAMA 300:182–188 customizable minimization program for allocation of patients to 44. OuHY, YangYC, WuHT, WuJS, LuFH, Chang CJ (2012) parallel groups in clinical trials. J Biomed Sci Eng 4:734–739 Increased fetuin-A concentrations in impaired glucose tolerance 27. Pocock SJ, Simon R (1975) Sequential treatment assignment with with or without nonalcoholic fatty liver disease, but not impaired balancing for prognostic factors in the controlled clinical trial. fasting glucose. J Clin Endocrinol Metab 97:4717–4723 Biometrics 31:103–115 45. Rametta R, Ruscica M, Dongiovanni P et al (2014) Hepatic 28. Brage S, Brage N, Franks PW, Ekelund U, Wareham NJ (2005) steatosis and PNPLA3 I148M variant are associated with serum Reliability and validity of the combined heart rate and movement fetuin-A independently of insulin resistance. Eur J Clin Investig sensor Actiheart. Eur J Clin Nutr 59:561–570 44:627–633 29. Ryysy L, Hakkinen AM, Goto Tet al (2000) Hepatic fat content and 46. Kahraman A, Sowa JP, Schlattjan M et al (2013) Fetuin-A mRNA insulin action on free fatty acids and glucose metabolism rather than expression is elevated in NASH compared with NAFL patients. insulin absorption are associated with insulin requirements during Clin Sci (Lond) 125:391–400 insulin therapy in type 2 diabetic patients. Diabetes 49:749–758 47. Strang BD, Bertics SJ, Grummer RR, Armentano LE (1998) 30. Juurinen L, Kotronen A, Graner M, Yki-Järvinen H (2008) Relationship of triglyceride accumulation to insulin clearance and Rosiglitazone reduces liver fat and insulin requirements and im- hormonal responsiveness in bovine hepatocytes. J Dairy Sci 81: proves hepatic insulin sensitivity and glycemic control in patients 740–747 with type 2 diabetes requiring high insulin doses. J Clin Endocrinol 48. Goto T, Onuma T, Takebe K, Kral JG (1995) The influence of fatty Metab 93:118–124 liver on insulin clearance and insulin resistance in non-diabetic 31. Steele R (1959) Influences of glucose loading and of injected insu- Japanese subjects. Int J Obes Relat Metab Disord 19:841–845 lin on hepatic glucose output. Ann N Y Acad Sci 82:420–430 49. Petersen KF, Dufour S, Befroy D, Lehrke M, Hendler RE, Shulman 32. Kotronen A, Juurinen L, Tiikkainen M, Vehkavaara S, Yki-Järvinen H GI (2005) Reversal of nonalcoholic hepatic steatosis, hepatic insu- (2008) Increased liver fat, impaired insulin clearance, and hepatic and lin resistance, and hyperglycemia by moderate weight reduction in adipose tissue insulin resistance in type 2 diabetes. Gastroenterology patients with type 2 diabetes. Diabetes 54:603–608 135:122–130 50. Nowotny B, Zahiragic L, Bierwagen A et al (2015) Low-energy 33. Chalasani N, Younossi Z, Lavine JE et al (2012) The diagnosis and diets differing in fibre, red meat and coffee intake equally improve management of non-alcoholic fatty liver disease: practice guideline insulin sensitivity in type 2 diabetes: a randomised feasibility trial. by the American Gastroenterological Association, American Diabetologia 58:255–264 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Diabetologia Springer Journals

A heterogeneous response of liver and skeletal muscle fat to the combination of a Paleolithic diet and exercise in obese individuals with type 2 diabetes: a randomised controlled trial

Free
12 pages
Loading next page...
 
/lp/springer_journal/a-heterogeneous-response-of-liver-and-skeletal-muscle-fat-to-the-9kkk0AfyvU
Publisher
Springer Berlin Heidelberg
Copyright
Copyright © 2018 by The Author(s)
Subject
Medicine & Public Health; Internal Medicine; Metabolic Diseases; Human Physiology
ISSN
0012-186X
eISSN
1432-0428
D.O.I.
10.1007/s00125-018-4618-y
Publisher site
See Article on Publisher Site

Abstract

Aims/hypothesis The aim of the study was to investigate ectopic fat deposition and insulin sensitivity, in a parallel single-blinded randomised controlled trial, comparing Paleolithic diet alone with the combination of Paleolithic diet and exercise in individuals with type 2 diabetes. Methods Thirty-two individuals with type 2 diabetes with BMI 25–40 kg/m and 30–70 years of age followed a Paleolithic diet ad libitum for 12 weeks. In addition, study participants were randomised by computer program to either supervised combined exercise training (PD-EX group) or standard care exercise recommendations (PD group). Staff performing examinations and assessing outcomes were blinded to group assignment. Thirteen participants were analysed in each group: hepatic and peripheral insulin sensitivity were measured using the hyperinsulinaemic–euglycaemic clamp technique combined with [6,6- H ]glucose infusion, and liver fat was assessed by proton magnetic resonance spectroscopy; both analyses were secondary endpoints. Intramyocellular lipid (IMCL) content was measured by magnetic resonance spectroscopy as a secondary analysis. All exam- inations were performed at Umeå University Hospital, Umeå, Sweden. Results Both study groups showed a median body weight loss of 7 kg. Fat mass decreased by 5.7 kg in the PD group and by 6.5 kg in the PD-EX group. Maximum oxygen uptake increased in the PD-EX group only. Liver fat showed a consistent reduction (74% decrease) in the PD group, while the response in the PD-EX group was heterogeneous (p < 0.05 for the difference between groups). IMCL content of the soleus muscle decreased by 40% in the PD group and by 22% in the PD-EX group (p < 0.05 for the difference between groups). Both groups improved their peripheral and adipose tissue insulin sensitivity, but not their hepatic insulin sensitivity. Plasma fetuin-A decreased by 11% in the PD group (p< 0.05) and remained unchanged in the PD-EX group. Liver fat changes during the intervention were correlated with changes in fetuin-A (r = 0.63, p< 0.01). Participants did not report any important adverse events caused by the intervention. Conclusions/interpretation A Paleolithic diet reduced liver fat and IMCL content, while there was a tissue-specific heteroge- neous response to added exercise training. Trial registration ClinicalTrials.gov NCT01513798 Funding Swedish Diabetes Research Foundation, County Council of Västerbotten, Swedish Heart and Lung Foundation, King Gustav V and Queen Victoria’sFoundation . . . . . . Keywords Exercise Hyperinsulinaemic–euglycaemic clamp Insulin sensitivity Intramyocellular fat Liver fat Nutrition . . . Obesity Paleolithic diet Proton magnetic resonance spectroscopy Weight loss Electronic supplementary material The online version of this article (https://doi.org/10.1007/s00125-018-4618-y) contains peer-reviewed but unedited supplementary material, which is available to authorised users. * Julia Otten Department of Food and Nutrition, Umeå University, Umeå, Sweden julia.otten@umu.se Department of Community Medicine and Rehabilitation, Sports Medicine Unit, Umeå University, Umeå, Sweden Department of Public Health and Clinical Medicine, Division of Department of Radiation Sciences, Radiation Physics and Biomedical Medicine, Umeå University, 90185 Umeå, Sweden Engineering, Umeå University, Umeå, Sweden Diabetologia (2018) 61:1548–1559 1549 Abbreviations efficiently improved glucose tolerance in overweight individuals ALT Alanine aminotransferase and in people with type 2 diabetes [10–12]. AST Aspartate aminotransferase In contrast to the well-established relationship between CRP C-reactive protein diet-induced weight loss and decreased liver fat, it remains EGP Endogenous glucose production unclear whether exercise training decreases liver fat indepen- FFM Fat-free mass dently of weight reduction. An earlier intervention study in IMCL Intramyocellular lipid participants with type 2 diabetes reported that 4 months of NAFLD Non-alcoholic fatty liver disease either aerobic or resistance training was associated with a PD group Paleolithic diet group slight decrease in liver fat [13]. However, aerobic exercise PD-EX group Paleolithic diet and exercise group combined with diet intervention does not appear to cause fur- ther liver fat reduction compared with diet intervention alone [14, 15]. To our knowledge, no prior study has investigated liver fat changes associated with diet in combination with both aerobic and resistance training. Interestingly, fetuin-A, a multi- Introduction functional protein secreted from both liver and adipose tissue, has been suggested as a putative link between insulin resistance Fat accumulation outside adipose tissue, i.e. ectopic fat in liver and muscle, is linked to decreased insulin sensitivity and type and liver and adipose tissue function [16, 17]. Several studies show that individuals with obesity, insulin 2diabetes [1, 2]. However, intervention studies are needed to test a putative causal relationship between changes in ectopic resistance and type 2 diabetes have higher intramyocellular lipid deposition and insulin sensitivity. Diet-induced weight lipid (IMCL) content compared with lean and healthy individ- loss in obese individuals is associated with reduction of fat in uals [18, 19]. In obese individuals, weight reduction decreases liver and skeletal muscle [3–5]. This has been linked to im- IMCL content and simultaneously improves insulin sensitivi- proved insulin sensitivity but has not been a universal finding. ty [20, 21]. However, lean endurance-trained athletes exhibit Notably, there are conflicting data regarding the effect of macro- IMCL content measured at rest that is nearly as high as that of people with type 2 diabetes, but with concomitant normal nutrient composition on ectopic fat [3–7]. Two recent studies on obese postmenopausal women found that a Paleolithic diet con- insulin sensitivity, referred to as the athlete’s paradox [18]. Moreover, IMCL content is reduced immediately following sumed ad libitum with a moderately decreased carbohydrate in- take and a high content of mono- and polyunsaturated fatty acids an acute bout of aerobic exercise in young lean individuals [22]. This suggests that IMCL is an important intracellular effectively reduced liver fat [8, 9]. Furthermore, a Paleolithic diet 1550 Diabetologia (2018) 61:1548–1559 source of energy during exercise in people with high insulin was in accordance with the Helsinki Declaration and was ap- sensitivity. This dynamic response does not occur in obese proved by the Regional Ethical Review Board, Umeå, Sweden. individuals, who show unchanged IMCL content after 1 h of cycling [23]. Notably, a 12 week exercise intervention in in- Diet intervention The Paleolithic diet included lean meat, dividuals with type 2 diabetes found an increased IMCL con- eggs, fish, seafood, nuts, fruits and vegetables. Dairy products, tent with concomitantly increased insulin sensitivity [24]. cereals, legumes and added sugar and salt were excluded. Owing to these inconclusive data, it is of major interest to Energy intake was ad libitum. Each study group attended five study how a combination of diet intervention and aerobic and group sessions run by a trained dietitian, and participants resistance training influences ectopic fat deposition and tissue- could contact the dietitian by e-mail or phone between meet- specific insulin sensitivity in individuals with type 2 diabetes. ings. Dietary intake was assessed at baseline and at 12 weeks, We therefore tested the hypothesis that overweight individuals using a 4 day self-reported weighed food record. A trained with type 2 diabetes on a 12 week Paleolithic diet would dietitian converted the food records into estimated energy exhibit a decrease in liver fat and IMCL content, associated and nutrient intakes using the nutritional calculation program with an improvement in hepatic and peripheral insulin sensi- Dietist XP 3.2 (Kost och Näringsdata, Bromma, Sweden). tivity. Moreover, we hypothesised that combined aerobic and resistance exercise training would lead to a further improve- Exercise intervention Prior to randomisation, all participants ment in liver fat and peripheral insulin sensitivity. were advised to perform at least 30 min of moderate exercise daily in accordance with the current guidelines for people with type 2 diabetes. The PD-EX group additionally underwent a training protocol combining aerobic exercise and resistance Methods training in 1 h sessions three times weekly at the Sports Medicine Unit at Umeå University, Umeå, Sweden. Low- Study design Overweight and obese individuals with type 2 intensity aerobic exercise was performed on a cross-trainer, diabetes consumed a Paleolithic diet for 12 weeks. In addition, and moderate- or high-intensity interval training was per- the participants were randomised to receive either supervised formed on a cycle ergometer. Resistance training included exercise training for 3 h per week (PD-EX group) or standard upper and lower body exercises involving multiple muscle care exercise recommendations (PD group). The reduction of groups. All exercise sessions were supervised by personal fat mass was the primary endpoint of this study and has been trainers with a Bachelor of Science degree in sports medicine. published previously [25]. The outcome measurements of this Body composition, liver fat, IMCL, VO and energy article are secondary endpoints (liver fat and peripheral and 2max expenditure Body composition analysis was performed using hepatic insulin sensitivity) and secondary analyses (IMCL). dual-energy x-ray absorptiometry (Lunar Prodigy X-ray Tube Housing Assembly, Brand BX-1L, Model 8743; GE Medical Participants and randomisation We used advertisements in Systems, Madison, WI, USA) at the Clinical Research Centre local newspapers and posters at Umeå University Hospital, Umeå, Sweden, to recruit individuals with type 2 diabetes, at Umeå University Hospital. VO was determined during 2max a standard cardiopulmonary exercise test on a cycle ergometer age 30–70 years and BMI 25–40 kg/m . For inclusion, partici- pantswere requiredtohaveanHbA 48–95 mmol/mol (6.5– at the Department of Clinical Physiology at Umeå University 1c Hospital. Resting energy expenditure was measured by indi- 10.8%) and be treated with diet and/or metformin. Women had to be postmenopausal. Exclusion criteria were smoking, BP >160/ rect calorimetry (Datex-Ohmeda Deltatrac II; Datex-Ohmeda, Madison, WI, USA) and adjusted by subtracting 5% during 100 mmHg, macroalbuminuria, cardiovascular disease, beta blocker use, severe illness and higher levels of training (e.g. 8 h of sleep. Physical activity energy expenditure was estimat- ed using a combined heart rate monitor and accelerometer for moderate endurance training five times a week, resistance train- 7 consecutive days (Actiheart; CamNtech, Cambridge, UK) as ing every other week). A total of 261 individuals were assessed for eligibility, of previously described [28]. Total energy expenditure was cal- culated as the sum of resting energy expenditure and physical whom 32 fulfilled the inclusion criteria and were randomised to the two groups (Fig. 1). Randomisation, using the computer activity energy expenditure with added 10% for diet-induced thermogenesis. program MinimPy version 0.3 [26], was performed by a statis- tician blinded to the study and not involved in data collection or Liver fat and tibialis anterior/soleus muscle fat were analysed by proton magnetic resonance spectroscopy as described in the analysis. Participants were assigned to either the PD group or the PD-EX group, using biased-coin minimisation with an al- electronic supplementary material (ESM) Methods. location ratio of 1:1 [27]. The nurses and technicians who per- formed the examinations were blinded to group affiliation. All Insulin sensitivity and insulin clearance Insulin sensitivity was assessed using the hyperinsulinaemic–euglycaemic clamp participants gave written informed consent. The study protocol Diabetologia (2018) 61:1548–1559 1551 Fig. 1 CONSORT flow diagram Enrolment Assessed for eligibility (n=261) Excluded (n=229) ♦ Did not meet inclusion criteria (n=185) ♦ Declined to participate (n=44) Randomised (n=32) Allocation Allocated to PD (n=16) Allocated to PD-EX (n=16) ♦ Received allocated intervention (n=15) ♦ Received allocated intervention (n=15) ♦ Did not receive allocated intervention (lack of ♦ Did not receive allocated intervention time to participate in diet education) (n=1) (disappointed with allocation) (n=1) Follow-up Discontinued intervention (loss of Discontinued intervention (n=0) motivation) (n=1) Primary analysis Analysed hyperinsulinaemic–euglycaemic clamp Analysed hyperinsulinaemic–euglycaemic (n=13) clamp (n=13) ♦ Did not perform clamp examination owing to ♦ Clamp examination failed (n=1) illness (n=2) Secondary analyses Analysed liver fat (n=12) Analysed liver fat (n=11) ♦ Refused examination (n=1) ♦ Implanted metal (n=2) Analysed tibialis anterior muscle IMCL (n=12) Analysed tibialis anterior muscle IMCL (n=11) ♦ Spectroscopy failed (n=1) ♦ Implanted metal (n=2) Analysed soleus muscle IMCL (n=10) Analysed soleus muscle IMCL (n=11) ♦ Spectroscopy failed (n=3) ♦ Implanted metal (n=2) technique combined with [6,6- H ]glucose infusion as previ- Denmark). Plasma glucose was clamped at 8 mmol/l by infu- ously described [24]. On the day of examination, participants sion of 20% glucose at a variable rate. This plasma glucose came to the Clinical Research Centre at Umeå University level was chosen based on previous hyperinsulinaemic– Hospital in the fasted state having refrained from physical ex- euglycaemic clamp studies [29, 30]. Arterialised blood was ercise for the prior 48 h. A catheter was placed in an antecubital sampled for determination of plasma glucose and [6,6- H ]glu- vein for infusion. For blood sampling, a catheter was placed in cose at t = 0, 150, 160, 170, 180, 330, 340, 350 and 360 min. the contralateral arm, retrograde into a superficial dorsal hand We analysed plasma insulin at t = 0, 180, 240, 300 and vein, with the hand placed in a heated box for arterialisation. 360 min, and NEFA at t = 0, 240, 270, 300, 330 and 360 min. Primed constant infusion of [6,6- H ]glucose (APL, −1 −1 Stockholm, Sweden) at a rate of 0.22 μmol kg min was Gene expression Real-time quantitative PCR was used to de- initiated at t = 0 min and continued until t = 360 min. At t = termine relative gene expression of TNFα and IL6 in subcu- 180 min, we initiated primed constant infusion of short-acting taneous adipose tissue of the abdomen (see ESM Methods for insulin (Actrapid; Novo Nordisk, Bagsværd, Denmark) at a rate further details). −2 −1 of 40 mU m min , which continued until t = 360 min. Between t = 180 min and t = 360 min, blood was sampled ev- Blood sample analysis Blood samples were taken in the ery 5 min for immediate determination of plasma glucose con- fasting state from a peripheral vein, followed by immediate centration (HemoCue 201 RT; Radiometer Medical, Brønshøj, analysis of plasma insulin, HbA , serum triacylglycerols, 1c 1552 Diabetologia (2018) 61:1548–1559 plasma aspartate aminotransferase (AST), plasma alanine performed using Spearman’s rho (r ). A two-sided p value aminotransferase (ALT) and plasma C-reactive protein of <0.05 was considered statistically significant. All statistical (CRP) in the Clinical Chemistry unit at Umeå University analyses were performed using R, version 3.2.2, a language Hospital. For NEFA analysis, plasma was stored at −80°C and environment for statistical computing (R Foundation for and later analysed using the NEFA-HR kit (Wako Statistical Computing, Vienna, Austria). Data are presented as Chemicals, Neuss, Germany). Serum fetuin-A concentrations median (interquartile range). were determined using a human fetuin-A ELISA kit (BioVendor, Brno, Czech Republic). The arterialised venous samples of the hyperinsulinaemic– Results euglycaemic clamp were analysed for [6,6- H ]glucose and unlabelled glucose using GC-MS at the Swedish Participant characteristics and intervention validation The Metabolomics Centre, Umeå, Sweden. The samples (100 μl) main results of the intervention have been previously pub- were extracted with 90% methanol (900 μl), including C -D- lished [25]. In the present substudy, we found no between- glucose as internal standard, then derivatised by addition of group differences in baseline characteristics except a higher freshly prepared acetic anhydride/pyridine (1/1 vol./vol.). fasting glucose in the PD-EX group (Table 1). Median weight Next, the solvent was removed by a stream of N , and there- loss was 7 kg in both groups (Table 1). Fat mass decreased by after the samples were dissolved in ethyl acetate. These sam- 5.7 kg in the PD group and by 6.5 kg in the PD-EX group ples were injected by an Agilent 7693 autosampler (Agilent (Table 1). Food records showed that both study groups simi- Technologies, Atlanta, GA, USA) into an Agilent 7890A gas larly reduced total energy intake during the ad libitum diet chromatograph, and analysed in an Agilent 7010C QQQ mass intervention, mostly by decreasing intake of carbohydrates spectrometer operating in selected ion monitoring mode. The fragment ion m/z 244 was used to detect [6,6- H ]glucose; and saturated fatty acids (Table 2). VO increased by 2 2max m/z 242 was used for unlabelled glucose and m/z 247 for 10% in the PD-EX group and decreased by 3% in the PD 13 2 C -D-glucose. To determine levels of [6,6- H ]glucose and group (p< 0.01 for between-group difference). Total physical 6 2 unlabelled glucose, calibration curves were set up between activity energy expenditure did not change in either group calibrants and internal standard. (Table 2). Calculations Endogenous glucose production (EGP) was de- Insulin sensitivity and insulin clearance Peripheral insulin sen- termined using Steele’s single-pool non-steady-state equation sitivity, measured as the rate of disappearance during the [31]. The glucose distribution volume was estimated as hyperinsulinaemic–euglycaemic clamp, increased by 53% in 0.1625 l/kg body weight. The suppression of EGP (%) was the PD group (p< 0.05) and by 42% in the PD-EX group (p calculated as [EGP basal (t = 150, 160, 170, 180 min) − EGP < 0.01) (Table 3). Hepatic insulin sensitivity, measured as clamp (t = 330, 340, 350, 360 min)] × 100 / EGP basal. To suppression of EGP, remained essentially unchanged during determine the rate of disappearance, the glucose infusion rate the intervention (Table 3). Both study groups showed in- during the last 30 min of clamping was added to the EGP clamp creased NEFA suppression during the hyperinsulinaemic– and corrected for the non-steady-state condition using Steele’s euglycaemic clamp. Adipose tissue insulin sensitivity, calcu- equation. One participant was excluded from calculations of lated as NEFA suppression during the whole clamp examina- EGP and rate of disappearance because the [6,6- H ] tion, increased by 3.4% in both intervention groups (p< 0.05 glucose infusion pump did not work properly during the third in the PD group, p< 0.01 in the PD-EX group) (Table 3). hour of the examination. Suppression of NEFA (%) was calcu- When the insulin sensitivity measures were normalised for lated as [(NEFA at t =0 min) − (NEFA at t = 240, 270, 300, plasma insulin during the clamp, the PD group showed more 330, 360 min)] − 100 / (NEFA at t =0 min) [32]. Insulin clear- pronounced improvement compared with the PD-EX group ance during insulin infusion was calculated by dividing the (Table 3). This was due to increased insulin clearance in the −1 −1 insulin infusion rate [mU (kg FFM) min ], where FFM is PD group (p< 0.05 for the between-group difference; Fig. 2). fat-free mass, by the mean plasma insulin concentration during The intervention groups did not significantly differ in any insulin infusion [32]. measure of insulin sensitivity. Statistical analysis Several variables showed a skewed distri- Fat in soleus and tibialis anterior muscle IMCL content of the bution; thus, we used the Wilcoxon rank-sum test to compare soleus muscle decreased by 40% in the PD group, but showed the treatment effect (change from baseline to 12 weeks) be- no significant change in the PD-EX group (Fig. 3a). Tibialis tween the PD and PD-EX groups. The change over time with- anterior IMCL content did not change significantly in either in each intervention group was determined using the intervention group (data not shown). Tibialis anterior IMCL content at baseline was associated with peripheral insulin Wilcoxon signed-rank test. Correlation analyses were Diabetologia (2018) 61:1548–1559 1553 Table 1 Baseline characteristics, Variable PD group PD-EX group body weight, and fasting blood samples n (male/female) 13 (9/4) 13 (8/5) Age, years 60 (54, 64) 61 (58, 67) Diabetes duration, years 3 (2, 6) 5 (1, 8) Body weight, kg Baseline 90.0 (83.3, 103.2) 97.2 (82.9, 107.4) Change 0–12 weeks −7.1 (−9.8, −5.6)*** −7.0 (−9.7, −5.6)*** BMI, kg/m Baseline 31.4 (29.4, 33.7) 31.4 (29.0, 34.6) Change 0–12 weeks −2.4 (−3.1, −1.8)*** −2.3 (−3.4, −2.2)*** Fat mass, kg Baseline 34.4 (30.1, 37.9) 33.6 (29.2, 39.2) Change 0–12 weeks −5.7 (−8.2, −4.0)*** −6.5 (−8.9, −5.1)*** HbA , mmol/mol 1c Baseline 55 (48, 58) 56 (50, 59) Change 0–12 weeks −11 (−15, −5)** −11 (−18, −7)** HbA ,% 1c Baseline 7.2 (6.5, 7.5) 7.3 (6.7, 7.5) Change 0–12 weeks −1.0 (−1.4, −0.5)** −1.0 (−1.7, −0.6)** Fasting plasma glucose, mmol/l Baseline 8.0 (6.9, 8.5) 8.6 (7.7, 10.5) Change 0–12 weeks −0.9 (−1.8, −0.1)* −2.0 (−3.0, −1.0)** Serum triacylglycerols, mmol/l Baseline 2.4 (1.4, 3.1) 1.7 (1.0, 2.3) Change 0–12 weeks −0.6 (−1.8, −0.2)* −0.4 (−1.0, −0.1)** Plasma NEFA, mmol/l Baseline 0.60 (0.53, 0.78) 0.80 (0.65, 0.90) Change 0–12 weeks 0.03 (−0.02, 0.20) −0.04 (−0.12, 0.17) Plasma AST, μkat/l Baseline 0.59 (0.53, 0.71) 0.55 (0.51, 0.66) Change 0–12 weeks −0.05 (−0.14, 0.14) 0.02 (−0.20, 0.22) Plasma ALT, μkat/l Baseline 0.67 (0.53, 0.91) 0.54 (0.48, 0.78) Change 0–12 weeks −0.12 (−0.34, −0.08)** −0.07 (−0.16, 0.07) Data are reported as median (interquartile range) *p< 0.05, **p< 0.01, ***p< 0.001 for the within-group change over time from baseline to 12 weeks p< 0.05 between the PD and PD-EX groups sensitivity (r = −0.45, p< 0.05) and fasting plasma NEFA (r in liver fat in the PD-EX group was 43% (p< 0.05), with no S S =0.59, p< 0.05). Changes in tibialis and soleus muscle significant difference between the intervention groups (p=0.08). IMCL content during the intervention were not correlated with Excluding the three outliers did not alter the results of the other changes in peripheral insulin sensitivity. statistical analyses (ESM Table 1). The three individuals whose liver fat increased during the intervention had a greater decrease in soleus muscle IMCL content compared with the other partic- Liver fat All participants in the PD group decreased their liver fat, while the intervention response was more heterogeneous ipants in the PD-EX group (ESM Table 1). VO ,plasma 2max in the PD-EX group (Fig. 3b). Median hepatic lipid reduction triacylglycerols, plasma AST, plasma ALTand plasma CRP were was 74% for the PD group and 32% for the PD-EX group (p improved or unchanged in these three individuals (ESM Table 1). < 0.05 for the difference between groups). Liver fat was not associated with hepatic insulin sensitivity Liver fat increased substantially in three individuals in the PD- at baseline. The decrease in liver fat during the intervention EX group. After exclusion of these outliers, the median decrease was correlated with an improvement in hepatic insulin 1554 Diabetologia (2018) 61:1548–1559 Table 2 Energy balance and die- Variable PD group (n = 12) PD-EX group (n =12) tary intake Energy intake, kJ/day Baseline 8330 (6204, 10,778) 6673 (5569, 9443) Change 0–12 weeks −1377 (−3284, −1025)** −2155 (−3330, −649)** Total energy expenditure, kJ/day Baseline 12,619 (11,129, 13,933) 12,485 (9196, 16,778) Change 0–12 weeks −954 (−1485, −285)* −1305 (−2414, 1201) Resting energy expenditure, kJ/day Baseline 6791 (6184, 7243) 7268 (5565, 7958) Change 0–12 weeks −510 (−774, −188)** −381 (−715, −92)** Physical activity energy expenditure, kJ/day Baseline 4276 (3615, 5519) 4201 (3008, 7201) Change 0–12 weeks −117 (−870, 372) −88 (−1686, 1640) Protein, g/day Baseline 80 (69, 95) 77 (63, 106) Change 0–12 weeks 5 (−17, 23) 1 (−16, 14) −1 −1 Protein, g kg day Baseline 0.85 (0.66, 1.14) 0.78 (0.73, 1.02) Change 0–12 weeks 0.10 (−0.09, 0.35) 0.06 (−0.12, 0.25) Carbohydrate, g/day Baseline 204 (148, 280) 169 (152, 197) Change 0–12 weeks −89 (−122, −49)** −92 (−117, −67)** Total fat, g/day Baseline 84 (58, 115) 64 (46, 98) Change 0–12 weeks −12 (−38, 8) −10 (−33, 24) Saturated fatty acids, g/day Baseline 31 (21, 48) 25 (19, 35) Change 0–12 weeks −14 (−33, −5)** −12 (−23, −6)** Monounsaturated fatty acids, g/day Baseline 32 (25, 41) 26 (16, 37) Change 0–12 weeks 4 (−16, 14) 5 (−6, 18) Polyunsaturated fatty acids, g/day Baseline 11 (9, 14) 9 (7, 16) Change 0–12 weeks 1 (−5, 5) 1 (−3, 7) Protein, E% Baseline 17 (14, 19) 18 (17, 20) Change 0–12 weeks 7 (4, 11)** 6 (3, 12)** Carbohydrate, E% Baseline 41 (38, 46) 45 (32, 49) Change 0–12 weeks −10 (−18, −3)* −14 (−21, −7)** Total fat, E% Baseline 40 (36, 41) 32 (31, 44) Change 0–12 weeks 6 (−6, 11) 9 (2, 14)* Saturated fatty acids, E% Baseline 15 (13, 18) 13 (12, 17) Change 0–12 weeks −5(−8, −3)** −4(−7, −2)** Monounsaturated fatty acids, E% Baseline 16 (14, 17) 12 (10, 17) Change 0–12 weeks 5 (−3, 11) 10 (5, 12)** Polyunsaturated fatty acids, E% Baseline 5.0 (4.7, 6.4) 5.6 (3.8, 6.4) Change 0–12 weeks 2.1 (−0.5, 3.6)* 3.2 (0.9, 3.7)** Data are reported as median (interquartile range) *p< 0.05, **p< 0.01 for the within-group change over time from baseline to 12 weeks E%, energy per cent Diabetologia (2018) 61:1548–1559 1555 Table 3 Insulin sensitivity Insulin sensitivity PD group PP-EX group Peripheral insulin sensitivity −1 −1 Rate of disappearance, mg kg min Baseline 3.79 (2.95, 4.23) 3.87 (3.02, 5.26) Change 0–12 weeks 2.05 (0.32, 3.59)* 1.15 (0.67, 2.66)** −1 −1 Rate of disappearance/insulin, μgkg min per mU/l Baseline 34.2 (28.6, 49.2) 46.2 (35.0, 74.5) Change 0–12 weeks 28.9 (11.8, 61.5)** 14.6 (5.0, 30.3)** Hepatic insulin sensitivity −1 −1 EGP, mg kg min Baseline 1.81 (1.56, 1.99) 1.78 (1.51, 2.49) Change 0–12 weeks 0.04 (−0.06, 0.55) 0.11 (−0.18, 0.60) Suppression of EGP, % Baseline 96 (83, 128) 114 (85, 121) Change 0–12 weeks 13 (−10, 44) 11 (−33, 35) Suppression of EGP/insulin, % per mU/l Baseline 0.98 (0.77, 1.36) 1.51 (0.97, 1.74) Change 0–12 weeks 0.22 (−0.03, 0.91)* 0.06 (−0.45, 0.55) Adipose tissue insulin sensitivity Suppression of NEFA, % Baseline 88 (80, 93) 89 (85, 93) Change 0–12 weeks 3.4 (1.3, 5.7)* 3.4 (0.2, 7.8)** Suppression of NEFA/insulin, % per mU/l Baseline 0.83 (0.74, 1.08) 1.12 (0.95, 1.27) Change 0–12 weeks 0.14 (0.03, 0.31)** 0.11 (−0.06, 0.20) Data are reported as median (interquartile range) *p< 0.05, **p< 0.01 for the within-group change over time from baseline to 12 weeks p< 0.05 between the PD and PD-EX groups sensitivity in the PD group (r = −0.62, p< 0.05) but not in the tissue was associated with the change in fasting plasma PD-EX group. Liver fat was associated with adipose tissue NEFA (r =0.45, p< 0.05). insulin sensitivity (NEFA suppression during the clamp) at baseline (r = −0.58, p< 0.01). Discussion Plasma fetuin-A Plasma fetuin-A decreased by 11% in the PD group (p< 0.05) and remained unchanged in the PD-EX This intervention with a Paleolithic diet in overweight indi- group (Fig. 4). Liver fat changes during the intervention were strongly correlated with changes in fetuin-A (r = 0.63, p viduals with type 2 diabetes showed decreased ectopic lipid accumulation in liver and soleus muscle, as well as improved < 0.01; Fig. 5). Changes in adipose tissue insulin sensitivity (suppression of NEFA/insulin) were also associated with peripheral insulin sensitivity. On a group level, the addition of combined resistance and aerobic exercise training to the diet changes in fetuin-A (r = 0.51, p< 0.01). By contrast, we intervention reduced the effect on muscle and liver fat content. found no association between changes in fetuin-A levels and This was due to the considerable heterogeneity in response to changes in hepatic or peripheral sensitivity. The three partici- exercise. Decreased liver fat during the intervention was pants in the PD-EX group who showed increased liver fat also strongly associated with reduction in plasma fetuin-A levels showed increased plasma fetuin-A levels (Fig. 5). in both intervention groups. This was linked to an improve- Inflammation in plasma and adipose tissue Plasma CRP de- ment in adipose tissue insulin sensitivity. creased in the PD-EX group (p< 0.05; Table 4). In adipose Current guidelines for non-alcoholic fatty liver disease tissue, IL6 and TNFα gene expression did not change signif- (NAFLD) recommend lifestyle interventions involving diet icantly during the intervention. During the intervention, the and exercisetodecreaseliver fat [33]. However, earlier studies change in TNFα gene expression in subcutaneous adipose suggest a complex relationship between liver fat and the effect 1556 Diabetologia (2018) 61:1548–1559 ** Baseline 12 weeks Baseline 12 weeks PD PD-EX 0 b Baseline 12 weeks Baseline 12 weeks PD PD-EX *** Fig. 2 Insulin clearance during 12 weeks of intervention in the PD and PD-EX groups. *p< 0.05 for the within-group change over time from baseline to 12 weeks. p< 0.05 for the intervention effect between the PD group and the PD-EX group of lifestyle interventions. Two short-term studies showed that weight reduction by a low-carbohydrate diet effectively re- duced liver fat [4, 5]. We found a reduction in liver fat after a Paleolithic diet with a moderately reduced carbohydrate content. We anticipated that our combined intervention in in- dividuals with type 2 diabetes would show an additional fat- decreasing effect on the liver, as a recent meta-analysis con- Baseline 12 weeks Baseline 12 weeks PD PD-EX cluded that exercise reduces hepatic fat content [34]. Fig. 3 IMCL content of the soleus muscle (a) and liver fat (b)during Unexpectedly, three individuals in the PD-EX group showed 12 weeks of intervention in the PD and PD-EX groups. IMCL is normal- a clear increase in liver fat. After exclusion of these three ised to the creatine concentration of the muscle. **p< 0.01, ***p< 0.001 individuals, we found that liver fat decreased significantly in for the within-group change over time from baseline to 12 weeks. both study groups, with no difference between groups, while p< 0.05 for the intervention effect between the PD group and the PD- EX group all other comparisons were unaltered. There may be several explanations for the unexpected response regarding liver fat in some individuals, including a prolonged exercise-induced in- crease in liver fat in some individuals or increased metabolic flexibility. The three individuals whose liver fat increased dur- ing 12 weeks of exercise thus showed decreased or unchanged inflammation (plasma CRP levels), triacylglycerol levels and liver enzymes, indicating improved metabolic health in this subgroup. A third possibility is that the participants did not refrain from exercise for 48 h before the examination as they were supposed to, leading to a decrease in muscle fat and an increase in liver fat. In healthy people with normal weight and in overweight individuals, an acute bout of aerobic exercise immediately increases liver fat [35, 36]. This exercise-induced increase in Baseline 12 weeks Baseline 12 weeks liver fat seems to be mainly due to the rise in plasma NEFA PD PD-EX during and after exercise [35, 36]. All participants in our study Fig. 4 Plasma fetuin-A levels during 12 weeks of intervention in the PD were told to refrain from exercise for 48 h prior to liver fat and PD-EX groups. *p< 0.05 for the within-group change over time examination. A possible explanation for the conundrum could from baseline to 12 weeks -1 -1 Insulin clearance during insulin infusion (ml [kg FFM] min ) Fetuin-A (μg/ml) Liver fat (%) Soleus muscle IMCL/creatine Diabetologia (2018) 61:1548–1559 1557 substrate during exercise and is replenished during recovery. Endurance athletes who performed 3 h of cycling exercise had a 20% decrease in IMCL content in the legs and a simultaneous 38% increase in IMCL content in the non-exercising arms [22]. Importantly, obesity and type 2 diabetes are associated with a lack of this dynamic response to exercise, which may be related to continuously increased plasma NEFA levels [23]. However, if adipose tissue lipolysis is reduced with a nicotinic acid analogue, r = 0.63 -200 p < 0.01 plasma NEFA levels are reduced and the decrease in IMCL content during one bout of exercise is more pronounced [41]. Another contributing factor to the heterogeneous response in -400 -30 -20 -10 0 10 20 the combined diet and exercise intervention is the intake of car- Liver fat change 0–12 weeks (%) bohydrates. Suppression of adipose tissue lipolysis through in- Fig. 5 Association between change in liver fat and change in plasma sulin administration or carbohydrate ingestion leads to a reduc- fetuin-A during the intervention tion in plasma NEFA levels. Indeed, glucose ingestion during exercise causes plasma NEFA levels to decrease below fasting be that an exercise-induced increase in liver fat may last longer levels during and after exercise [36, 42]. Accordingly, glucose than 2 days in some individuals. supplementation during and after cycling prevents an increase in Lipolysis and inflammation are closely linked [37]. During liver fat during the recovery phase [36]. Furthermore, studies our intervention, decreasing TNFα expression in adipose tis- with isoenergetic diets show a decrease in liver fat only with sue was associated with decreasing plasma NEFA levels. high-carbohydrate diets, not with high-fat diets [6, 7]. Moreover, we found an association between liver fat and sup- Although energy intake was ad libitum in our study, participants pressibility of NEFA production, highlighting the importance reported decreases in carbohydrate and total energy intake. More of plasma NEFA concentration for hepatic lipid content. In detailed studies of macronutrient intake in relation to exercise are NAFLD, most triacylglycerols in the fatty liver originate from therefore of interest regarding the effects on hepatic lipid content. plasma NEFA, and most plasma NEFA originate from adipose Fetuin-A is secreted mainly from liver and adipose tissue tissue [38]. Plasma NEFA uptake in liver cells and esterifica- and is elevated in type 2 diabetes and NAFLD [17, 43, 44]. tion into hepatic triacylglycerols are insulin-independent, de- Circulating fetuin-A levels have been associated with severity pending only on the plasma NEFA concentration [39]. of liver steatosis, independently of insulin resistance, and with Notably, the percentage of plasma NEFA taken up by the liver non-alcoholic steatohepatitis [45, 46]. Our results showed a remains constant both during and after exercise [40]. strong association between fetuin-A levels and changes in Plasma NEFA concentrations are also important for IMCL liver fat content. Changes in circulating fetuin-A levels were content, and we found that fasting plasma NEFA levels were also associated with the ability to suppress NEFA production. closely associated with IMCL content of the tibialis anterior Since fetuin-A can be secreted by both hepatocytes and adi- muscle. In healthy individuals, IMCL is used as an energy pocytes, it remains unclear whether fetuin-A secreted by the liver influences adipose tissue or the other way round. The PD group increased insulin clearance, which might Table 4 Inflammatory markers in plasma and subcutaneous adipose have been due to the improvement in liver fat content. tissue Hepatocytes thus show impaired insulin clearance when load- Marker PD group PD-EX group ed with triacylglycerols in vitro [47], and liver fat is inversely related to insulin clearance in vivo [48]. Plasma CRP, nmol/l Improvement of hepatic insulin sensitivity was less pro- Baseline 11 (6, 27) 13 (6, 23) nounced in our intervention: only if normalised by insulin, Change 0–12 weeks −2(−9, 1) −3(−8, 0)* the PD group increased hepatic insulin sensitivity. This may IL6 in s.c. adipose tissue, relative expression of mRNA relate to the relatively well-preserved hepatic insulin sensitivity Baseline 0.81 (0.42, 1.16) 0.79 (0.61, 1.17) in our study cohort. In most studies, diet-induced weight loss in Change 0–12 weeks 0.15 (−0.45, 0.62) −0.14 (−0.42, 0.01) people with type 2 diabetes causes an increase in hepatic insulin TNFα in s.c. adipose tissue, relative expression of mRNA sensitivity, but some authors report it unchanged [49, 50]. Baseline 2.19 (1.79, 3.03) 2.45 (1.89, 3.81) A limitation of our study is that the insulin dose might have Change 0–12 weeks −0.06 (−0.53, 0.35) −0.70 (−1.02, 0.32) been too high to detect changes in hepatic insulin sensitivity. Moreover, target plasma glucose during the euglycaemic clamp Data are reported as median (interquartile range) was 8 mmol/l, which may not represent euglycaemia, especially *p< 0.05 for the within-group change over time from baseline to 12 weeks after the intervention when fasting glucose was normalised. A Fetuin-A change 0–12 weeks (μg/ml) 1558 Diabetologia (2018) 61:1548–1559 Open Access This article is distributed under the terms of the Creative lower insulin dose and a lower glucose target during the clamp Commons Attribution 4.0 International License (http:// studies might have detected more subtle changes in hepatic in- creativecommons.org/licenses/by/4.0/), which permits unrestricted use, sulin sensitivity. Indeed, we have previously demonstrated an distribution, and reproduction in any medium, provided you give appro- improvement in HOMA-IR after 5 weeks and 6 months follow- priate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. ing a Paleolithic diet in healthy overweight participants [8, 9]. Another limitation is that we had to exclude soleus References muscle measurements from three participants because we could not separate the intramyocellular and 1. Cusi K (2016) Treatment of patients with type 2 diabetes and non- extramyocellular lipid signals. This is a known technical alcoholic fatty liver disease: current approaches and future direc- difficulty related to the fact that soleus muscle fibres are tions. Diabetologia 59:1012–1020 not aligned in parallel to the main magnetic field. This 2. Perseghin G, Scifo P, De Cobelli F et al (1999) Intramyocellular shortage of data limits our ability to draw conclusions triglyceride content is a determinant of in vivo insulin resistance in humans: a 1H-13C nuclear magnetic resonance spectroscopy as- regarding the effects of the intervention on different sessment in offspring of type 2 diabetic parents. Diabetes 48: skeletal muscle types. Finally, analyses of gene variants 1600–1606 that may influence liver fat accumulation, e.g. PNPLA3, 3. Haufe S, Engeli S, Kast P et al (2011) Randomized comparison of are of interest in future intervention studies. reduced fat and reduced carbohydrate hypocaloric diets on intrahepatic fat in overweight and obese human subjects. In conclusion, our results indicate that an exercise interven- Hepatology 53:1504–1514 tion is associated with a heterogeneous response in liver fat 4. Browning JD, Baker JA, Rogers T, Davis J, Satapati S, Burgess SC content in obese individuals with type 2 diabetes, despite im- (2011) Short-term weight loss and hepatic triglyceride reduction: proved metabolic health. Further studies are needed to under- evidence of a metabolic advantage with dietary carbohydrate re- striction. Am J Clin Nutr 93:1048–1052 stand how exercise changes liver fat and hepatic insulin sen- 5. Kirk E, Reeds DN, Finck BN, Mayurranjan SM, Patterson BW, Klein sitivity in relation to energy balance and macronutrient intake S (2009) Dietary fat and carbohydrates differentially alter insulin sen- among individuals with obesity and type 2 diabetes. sitivity during caloric restriction. Gastroenterology 136:1552–1560 6. Westerbacka J, Lammi K, Hakkinen AM et al (2005) Dietary fat content modifies liver fat in overweight nondiabetic subjects. J Clin Acknowledgements The authors thank the study participants; the re- Endocrinol Metab 90:2804–2809 search nurses I. Arnesjö, K. Iselid, L. Uddståhl, C. Ring and L. H. 7. van Herpen NA, Schrauwen-Hinderling VB, Schaart G, Mensink RP, Bergman at the Clinical Research Centre, (Umeå University Hospital, Schrauwen P (2011) Three weeks on a high-fat diet increases Umeå, Sweden) for performing the clamp examinations; A. Tellström intrahepatic lipid accumulation and decreases metabolic flexibility in (Department of Public Health and Clinical Medicine, Umeå University, healthy overweight men. J Clin Endocrinol Metab 96:E691–E695 Umeå, Sweden) for guiding participants during the diet intervention; and 8. Ryberg M, Sandberg S, Mellberg C et al (2013) A Palaeolithic-type K. Lundgren (Swedish Metabolomics Centre, Swedish University of diet causes strong tissue-specific effects on ectopic fat deposition in Agricultural Sciences, Umeå, Sweden) for the GC-MS analysis of obese postmenopausal women. J Intern Med 274:67–76 [6,6- H ]glucose. Parts of this paper were presented at the 52nd Annual 9. Otten J, Mellberg C, Ryberg M et al (2016) Strong and persistent Meeting of the European Association for the Study of Diabetes, Munich, effect on liver fat with a Paleolithic diet during a two-year interven- Germany, 12–16 September 2016. tion. Int J Obes 40:747–753 10. Lindeberg S, Jonsson T, Granfeldt Y et al (2007) A Palaeolithic diet Data availability Data are available on request from the authors. improves glucose tolerance more than a Mediterranean-like diet in individuals with ischaemic heart disease. Diabetologia 50:1795–1807 Funding This study was supported by grants from the Swedish Diabetes 11. Jonsson T, Granfeldt Y, Ahren B et al (2009) Beneficial effects of a Research Foundation (2014-096), the County Council of Västerbotten Paleolithic diet on cardiovascular risk factors in type 2 diabetes: a (VLL-460481), the Swedish Heart and Lung Foundation (20120450), randomized cross-over pilot study. Cardiovasc Diabetol 8:35 and King Gustav V and Queen Victoria’sFoundation. 12. Frassetto LA, Schloetter M, Mietus-Synder M, Morris RC Jr, Sebastian Duality of interest The authors declare that there is no duality of interest A (2009) Metabolic and physiologic improvements from consuming a associated with this manuscript. paleolithic, hunter-gatherer type diet. Eur J Clin Nutr 63:947–955 13. Bacchi E, Negri C, Targher G et al (2013) Both resistance training Contribution statement JO designed the study, recruited the participants, and aerobic training reduce hepatic fat content in type 2 diabetic collected the data, performed the statistical analyses and wrote the manu- subjects with nonalcoholic fatty liver disease (the RAED2 script. AS, MW, MR, MS and JH designed the study, interpreted the data Randomized Trial). Hepatology 58:1287–1295 and revised the manuscript. MW implemented the dietary intervention. AI 14. Tamura Y, Tanaka Y, Sato F et al (2005) Effects of diet and exercise on conducted the exercise intervention, analysed and interpreted the data and muscle and liver intracellular lipid contents and insulin sensitivity in drafted the article. IS analysed fetuin-A, TNFα and IL6 and drafted the type 2 diabetic patients. J Clin Endocrinol Metab 90:3191–3196 article. JH analysed the spectroscopy data. TO designed the study, 15. Bozzetto L, Prinster A, Annuzzi G et al (2012) Liver fat is reduced interpreted the data and wrote the manuscript. All authors gave their final by an isoenergetic MUFA diet in a controlled randomized study in approval of the manuscript to be published. JO is the guarantor of this work type 2 diabetic patients. Diabetes Care 35:1429–1435 and, as such, had full access to all the data in the study and takes respon- 16. Stefan N, Haring HU (2013) Circulating fetuin-A and free fatty acids sibility for the integrity of the data and the accuracy of the data analysis. interact to predict insulin resistance in humans. Nat Med 19:394–395 17. Trepanowski JF, Mey J, Varady KA (2015) Fetuin-A: a novel link between obesity and related complications. Int J Obes 39:734–741 Diabetologia (2018) 61:1548–1559 1559 18. Goodpaster BH, He J, Watkins S, Kelley DE (2001) Skeletal mus- Association for the Study of Liver Diseases, and American College of Gastroenterology. Gastroenterology 142:1592–1609 cle lipid content and insulin resistance: evidence for a paradox in endurance-trained athletes. J Clin Endocrinol Metab 86:5755–5761 34. Golabi P, Locklear CT, Austin P et al (2016) Effectiveness of exer- 19. Falholt K, Jensen I, Lindkaer Jensen S et al (1988) Carbohydrate cise in hepatic fat mobilization in non-alcoholic fatty liver disease: and lipid metabolism of skeletal muscle in type 2 diabetic patients. systematic review. World J Gastroenterol 22:6318–6327 Diabet Med 5:27–31 35. Egger A, Kreis R, Allemann S et al (2013) The effect of aerobic 20. Greco AV, Mingrone G, Giancaterini A et al (2002) Insulin resis- exercise on intrahepatocellular and intramyocellular lipids in tance in morbid obesity: reversal with intramyocellular fat deple- healthy subjects. PLoS One 8:e70865 tion. Diabetes 51:144–151 36. Bilet L, Brouwers B, van Ewijk PA et al (2015) Acute exercise does not 21. Gray RE, Tanner CJ, Pories WJ, MacDonald KG, Houmard JA decrease liver fat in men with overweight or NAFLD. Sci Rep 5:9709 (2003) Effect of weight loss on muscle lipid content in morbidly 37. Chen Z, Yu R, Xiong Y, Du F, Zhu S (2017) A vicious circle obese subjects. Am J Physiol Endocrinol Metab 284:E726–E732 between insulin resistance and inflammation in nonalcoholic fatty 22. Schrauwen-Hinderling VB, van Loon LJ, Koopman R, Nicolay liver disease. Lipids Health Dis 16:203 K, Saris WH, Kooi ME (2003) Intramyocellular lipid content is 38. Donnelly KL, Smith CI, Schwarzenberg SJ, Jessurun J, Boldt MD, increased after exercise in nonexercising human skeletal muscle. Parks EJ (2005) Sources of fatty acids stored in liver and secreted J Appl Physiol (1985) 95:2328–2332 via lipoproteins in patients with nonalcoholic fatty liver disease. J 23. Ipavec-Levasseur S, Croci I, Choquette S et al (2015) Effect of 1-h Clin Invest 115:1343–1351 moderate-intensity aerobic exercise on intramyocellular lipids in 39. Vatner DF, Majumdar SK, Kumashiro N et al (2015) Insulin- obese men before and after a lifestyle intervention. Appl Physiol independent regulation of hepatic triglyceride synthesis by fatty Nutr Metab 40:1262–1268 acids. Proc Natl Acad Sci U S A 112:1143–1148 24. Meex RC, Schrauwen-Hinderling VB, Moonen-Kornips E et al 40. Van Hall G, Bulow J, Sacchetti M, Al Mulla N, Lyngso D, (2010) Restoration of muscle mitochondrial function and metabolic Simonsen L (2002) Regional fat metabolism in human splanchnic flexibility in type 2 diabetes by exercise training is paralleled by and adipose tissues; the effect of exercise. J Physiol 543:1033–1046 increased myocellular fat storage and improved insulin sensitivity. 41. van Loon LJ, Manders RJ, Koopman R et al (2005) Inhibition of Diabetes 59:572–579 adipose tissue lipolysis increases intramuscular lipid use in type 2 25. Otten J, Stomby A, Waling M et al (2017) Benefits of a Paleolithic diabetic patients. Diabetologia 48:2097–2107 diet with and without supervised exercise on fat mass, insulin sen- 42. Schrauwen P, Hesselink MK, Vaartjes I et al (2002) Effect of acute sitivity, and glycemic control: a randomized controlled trial in indi- exercise on uncoupling protein 3 is a fat metabolism-mediated ef- viduals with type 2 diabetes. Diabetes Metab Res Rev https://doi. fect. Am J Physiol Endocrinol Metab 282:E11–E17 org/10.1002/dmrr.2828 43. Ix JH, Wassel CL, Kanaya AM et al (2008) Fetuin-A and incident 26. Saghaei M, Saghaei S (2011) Implementation of an open-source diabetes mellitus in older persons. JAMA 300:182–188 customizable minimization program for allocation of patients to 44. OuHY, YangYC, WuHT, WuJS, LuFH, Chang CJ (2012) parallel groups in clinical trials. J Biomed Sci Eng 4:734–739 Increased fetuin-A concentrations in impaired glucose tolerance 27. Pocock SJ, Simon R (1975) Sequential treatment assignment with with or without nonalcoholic fatty liver disease, but not impaired balancing for prognostic factors in the controlled clinical trial. fasting glucose. J Clin Endocrinol Metab 97:4717–4723 Biometrics 31:103–115 45. Rametta R, Ruscica M, Dongiovanni P et al (2014) Hepatic 28. Brage S, Brage N, Franks PW, Ekelund U, Wareham NJ (2005) steatosis and PNPLA3 I148M variant are associated with serum Reliability and validity of the combined heart rate and movement fetuin-A independently of insulin resistance. Eur J Clin Investig sensor Actiheart. Eur J Clin Nutr 59:561–570 44:627–633 29. Ryysy L, Hakkinen AM, Goto Tet al (2000) Hepatic fat content and 46. Kahraman A, Sowa JP, Schlattjan M et al (2013) Fetuin-A mRNA insulin action on free fatty acids and glucose metabolism rather than expression is elevated in NASH compared with NAFL patients. insulin absorption are associated with insulin requirements during Clin Sci (Lond) 125:391–400 insulin therapy in type 2 diabetic patients. Diabetes 49:749–758 47. Strang BD, Bertics SJ, Grummer RR, Armentano LE (1998) 30. Juurinen L, Kotronen A, Graner M, Yki-Järvinen H (2008) Relationship of triglyceride accumulation to insulin clearance and Rosiglitazone reduces liver fat and insulin requirements and im- hormonal responsiveness in bovine hepatocytes. J Dairy Sci 81: proves hepatic insulin sensitivity and glycemic control in patients 740–747 with type 2 diabetes requiring high insulin doses. J Clin Endocrinol 48. Goto T, Onuma T, Takebe K, Kral JG (1995) The influence of fatty Metab 93:118–124 liver on insulin clearance and insulin resistance in non-diabetic 31. Steele R (1959) Influences of glucose loading and of injected insu- Japanese subjects. Int J Obes Relat Metab Disord 19:841–845 lin on hepatic glucose output. Ann N Y Acad Sci 82:420–430 49. Petersen KF, Dufour S, Befroy D, Lehrke M, Hendler RE, Shulman 32. Kotronen A, Juurinen L, Tiikkainen M, Vehkavaara S, Yki-Järvinen H GI (2005) Reversal of nonalcoholic hepatic steatosis, hepatic insu- (2008) Increased liver fat, impaired insulin clearance, and hepatic and lin resistance, and hyperglycemia by moderate weight reduction in adipose tissue insulin resistance in type 2 diabetes. Gastroenterology patients with type 2 diabetes. Diabetes 54:603–608 135:122–130 50. Nowotny B, Zahiragic L, Bierwagen A et al (2015) Low-energy 33. Chalasani N, Younossi Z, Lavine JE et al (2012) The diagnosis and diets differing in fibre, red meat and coffee intake equally improve management of non-alcoholic fatty liver disease: practice guideline insulin sensitivity in type 2 diabetes: a randomised feasibility trial. by the American Gastroenterological Association, American Diabetologia 58:255–264

Journal

DiabetologiaSpringer Journals

Published: Apr 26, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off