- Email: [email protected]
Serum fatty acid composition and insulin resistance are independently associated with liver fat markers in elderly men
diabetes research and clinical practice 87 (2010) 379–384
Contents lists available at ScienceDirect
Diabetes Research and Clinical Practice journ al h omepage: www .elsevier.co m/lo cate/diabres
Serum fatty acid composition and insulin resistance are independently associated with liver fat markers in elderly men§ ¨ rnlo¨v b,c, Bjo¨rn Zethelius b, Ulf Rise´rus a Helena Petersson a,*, Johan A a
Department of Public Health and Caring Sciences/Clinical Nutrition and Metabolism, Uppsala University, Uppsala Science Park, 75185 Uppsala, Sweden b Department of Public Health and Caring Sciences/Geriatrics, Uppsala University, Uppsala Science Park, 75185 Uppsala, Sweden c Department of Health and Social Sciences, Ho¨gskolan Dalarna, 79188 Falun, Sweden
article info
abstract
Article history:
Aim: To investigate the relationships of serum fatty acid (FA) composition and estimated
Received 13 August 2009
desaturase activities with the liver fat marker alanine aminotransferase (ALT).
Received in revised form
Methods: 546 Swedish elderly men of a population-based cohort participated in this cross-
23 October 2009
sectional study. FA composition was assessed in serum cholesterol esters to determine
Accepted 24 November 2009
dietary fat quality (e.g. linoleic) and desaturation products (e.g. dihomo-g-linolenic acid). Desaturase indices, including stearoyl coenzymeA desaturase-1 (SCD-1), were calculated by FA product-to-precursor ratios.
Keywords:
Results: In linear regression analyses adjusting for lifestyle, abdominal obesity and insulin
Fatty acid composition
sensitivity, the dietary biomarker linoleic acid (n-6), but not n-3 FAs, was inversely related to
Liver fat
ALT. Desaturation products including palmitoleic, oleic, g-linolenic and dihomo-g-linolenic
Lipogenesis
acids, and D6-desaturase and SCD-1 indices were directly related to ALT (all p < 0.05). After
Stearoyl coenzymeA desaturase-1
further adjustment for factors previously linked to fatty liver (i.e. serum lipids, adiponectin concentrations), SCD-1 index ( p = 0.004) and insulin resistance ( p < 0.0001) were independent determinants of ALT activity, whereas waist circumference, triglycerides, non-esterified FA and adiponectin were not. Conclusion: A low dietary intake of linoleic acid and elevated SCD-1 index may contribute to higher ALT activity in elderly men, even independently of obesity and insulin resistance. # 2009 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
Non-alcoholic fatty liver is related to obesity, especially abdominal obesity, and is strongly associated with insulin resistance, the metabolic syndrome and type 2 diabetes. A fatty liver also overproduces several cardiovascular risk factors, such as glucose, very low density lipoprotein triglycerides, coagulation factors, and C-reactive protein [1]. The pathogenesis of fatty liver is poorly understood. Dietary §
factors could play a role and recently a low-fat diet was reported to reduce liver fat [2]. Furthermore it was recently shown that self-reported saturated fat intake, was directly associated with liver fat [3], whereas animal studies suggest a protective effect of polyunsaturated fat [4]. Due to major reporting bias of dietary fat intake assessed by food records it is relevant to investigate this relation using objective biomarkers of dietary fat quality, i.e. serum fatty acid (FA) composition [5]. Serum FA composition has previously been
This study was presented at the 44th annual meeting of the European Association for the Study of Diabetes (EASD) 2008. * Corresponding author. Tel.: +46 18 6117980; fax: +46 18 6117976. E-mail address: [email protected] (H. Petersson). 0168-8227/$ – see front matter # 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.diabres.2009.11.019
380
diabetes research and clinical practice 87 (2010) 379–384
related to fatty liver in a limited number of subjects [6], but studies in larger population-based samples are lacking. Also, studies taking abdominal obesity, insulin resistance, adiponectin and other potential confounders into account are not available. In general, the essential FAs linoleic acid (18:2n-6) and a-linolenic acid (18:3n-3) are good markers of dietary intake of these FAs whereas saturated and monounsaturated FAs are somewhat weaker biomarkers [5]. In addition to the dietary intake, specific serum FAs also reflect endogenous fat metabolism and FA desaturation, e.g. palmitoleic, g-linolenic and dihomo-g-linolenic acids. Desaturases induce double-bounds in the FA chains. Stearoyl coenzymeA desaturase (SCD)-1 is the final step in de novo lipogenesis and converts saturated FA to monounsaturated FA, whereas D5- and D6-desaturases participate in the metabolism of polyunsaturated FAs [7]. SCD-1 is of high interest as this lipogenic enzyme has been suggested to play a role in the development of fatty liver [8,9]. In epidemiological studies, desaturase activities can be estimated using FA product-to-precursor ratios. To date, population-based data are lacking regarding the association between markers of endogenous FA metabolism and fatty liver. Based on previous data, we hypothesized that a diet high in saturated fat and low in polyunsaturated fat as well as increased desaturase activity could be involved in the development of non-alcoholic fatty liver. Therefore, we investigated the associations between serum FA composition, reflecting both dietary intake of FAs and endogenous fat metabolism, and alanine aminotransferase (ALT) levels, reflecting liver fat, in a large population-based sample of elderly men. To further explore a potential independent role of dietary FAs and/or desaturases we adjusted for lifestyle factors, as well as various metabolic factors previously linked to fatty liver, i.e. abdominal obesity, insulin resistance (measured directly by euglycemic insulin clamp), plasma adiponectin concentrations, triglycerides and non-esterified FA (NEFA).
2.
Subjects
The subjects were participants of the Uppsala Longitudinal Study of Adult Men (ULSAM, http://www.pubcare.uu.se/ ULSAM/), a population-based prospective study starting in 1970 in Uppsala, Sweden. Twelve-hundred twenty-one (73%) of 1681 men participated in the third investigation cycle that was carried out between 1991 and 1995, when the participants were 71 (0.6) years old. Eleven-hundred eighty-men had measure of ALT. Of these, 583 subjects had data on serum FA composition. Participants with a previous diagnosis of liver disease in the compulsory Swedish hospital discharge register (ICD-8 codes 570-573, ICD-9 codes 570-573 or ICD-10 codes K70-K77) were excluded (n = 4). Subjects consuming 20 g alcohol/day (n = 30) were excluded to avoid alcohol-related fatty liver. Three subjects had ALT levels below detection level (<0.10 mkat/l) and were excluded. Thus, 546 subjects were included in the present study. The study was approved by the Ethics Committee of Uppsala University. All subjects gave written informed consent.
3.
Materials and methods
3.1.
Clinical assessment and biochemistry
The investigation included anthropometric measurements, blood biochemistry, blood pressure, a self-administered medical questionnaire and an interview. All measurements were performed under standardised conditions as previously described [10,11]. Height was measured to the nearest centimeter and body weight to the nearest 0.1 kg. Body mass index (BMI) was calculated as the ratio of weight to height squared (kg/ m2). Waist circumference (WC) was measured midway between the lowest rib and the iliac crest in a supine position. All blood samples were drawn after an overnight fast. Serum triglycerides were analysed by enzymatic technique using IL Test Enzymatic Method for use in a Monarch apparatus (Instrumentation Laboratories, Lexington, USA). Serum NEFA was measured by an enzymatic colorimetric method (Wako Chemicals, Neuss, Germany). Plasma insulin was measured using an enzymatic-immunological assay (Enzymmun, Boehringer Mannheim, Germany) performed in an ES300 automatic analyser (Boehringer Mannheim). FA composition in serum cholesterol esters were analysed as previously described [12]. Serum was extracted with chloroform in the presence of methanol, butylated hydroxytoluene and NaH2PO4. Lipids were separated by thin liquid chromatography and trans methylated. The percentage composition of methylated FA 16:0–22:6 was determined by gas chromatography (a 25 m NB-351 silica capillary column, i.d. 0.32 mm, phase layer 0.20 mm) with helium as a carrier gas. Desaturase activities were estimated by the following FA product-to-precursor ratios; D5-desaturase (20:4n-6/20:3n-6), D6-desaturase (18:3n-6/18:2n-6), and SCD-1 (16:1n-7/16:0). Lipogenic index was calculated as 16:0/18:2n-6 ratio [9]. ALT in serum was used as a surrogate marker of fatty liver and analysed by the enzymatic method ECCLS/IFCC (Hitachi 717, Boehringer Mannheim, Germany) with CV = 1.6% at 1.30 mkat/l. Insulin sensitivity was determined by euglycemic insulin clamp according to DeFronzo et al. [13], slightly modified. The insulin infusion rate was 56 mU/min per body surface area (m2), instead of 40. Plasma glucose concentrations were maintained at 5.1 mmol/l during the clamp by adjusting the rate of infusion of a 20% glucose solution. Glucose disposal; M (mg/kg body weight/min), was calculated as the amount of glucose taken up during the last 60 min of the clamp. Serum adiponectin was analysed using a validated in-house timeresolved immunofluorometric assay (TR-IFMA) based on commercial reagents from R&D Systems (Abingdon, UK). Physical activity (categorised as sedentary, moderate, regular and athletic) was assessed by a self-administered questionnaire [14]. Alcohol intake was assessed by a 7-day dietary record [15].
3.2.
Statistical analyses
The distribution of the variables was examined by Shapiro–Wilk W-test. In order to promote a normal distribution, the following variables were logarithmically transformed before analysis; ALT, 16:1, 18:0, 18:3n-6, 18:3n-3, 20:5n-3, D6-desaturase, SCD-1, fasting insulin, NEFA, triglycerides, and adiponectin. Alcohol intake was divided into quartiles. Linear regression analysis was used to associate FA composition (standardised to one
diabetes research and clinical practice 87 (2010) 379–384
standard deviation) with ALT. The relation between FA composition and ALT was adjusted for abdominal obesity (measured by WC), lifestyle factors (alcohol intake, physical activity) and insulin sensitivity in a multivariable analysis (Model 1, n = 460). To further explore a potential independent link between FAs and liver fat, we added metabolic factors (standardised to one standard deviation) previously related to fatty liver; triglycerides, NEFA and adiponectin (Model 2, n = 460). To avoid colinearity in Model 2, we only included FA variables with a p-value <0.001 in Model 1 (i.e. linoleic acid and SCD-1 index). Since this study included a subgroup of the ULSAM cohort, potential selection bias was investigated by comparing participants with FA measurements (n = 546) with those without FA measurements as regard to ALT (n = 634), WC, BMI, M and triglycerides. A JMP software package was used for statistics (SAS Institute, Cary, NC). p < 0.05 was considered as significant.
381
FA composition with regards to WC, BMI, ALT, M and triglycerides ( p > 0.28 for all).
4.1. Associations between FA composition and ALT— influence of lifestyle factors and insulin sensitivity The association between FA composition and ALT is presented in Table 2 as univariable and multivariable analyses adjusting for WC, alcohol intake, physical activity and insulin sensitivity (multivariable Model 1). The results were similar when adjusting for BMI instead of WC or if M was substituted for fasting insulin. When excluding subjects with diabetes (n = 76), the trend remained for all FAs with the exception of 18:1. Lipogenic index was related to ALT in univariable analysis (b = 0.08, p < 0.0001) and multivariable Model 1 (b = 0.06, p = 0.003).
4.2. Associations between FA composition and ALT— influence of other metabolic factors
4.
Results
Population characteristics are presented in Table 1. The range of ALT activity was between 0.10 and 2.55 mkat/l. The subsample of participants with measurements of FA composition did not differ from the sub-sample without measurements of
Table 1 – Population characteristics. Variable
Mean SD a
Age (years) Waist circumference (cm) BMI (kg/m2) Alcohol intake (g/day)a Insulin sensitivity (M, mg/kg body weight/min) Alanine aminotransferase (mkat/l)a Adiponectin (ng/ml)a Non-esterified fatty acids (mmol/l)a Triglycerides (mmol/l)a Palmitic acid (16:0, %) Palmitoleic acid (16:1, %)a Stearic acid (18:0, %)a Oleic acid (18:1, %) Linoleic acid (18:2n-6, %) g-Linolenic acid (18:3n-6, %)a a-Linolenic acid (18:3n-3, %)a Dihomo-g-linolenic acid (20:3n-6, %) Arachidonic acid (20:4n-6, %) Eicosapentaenoic acid (20:5n-3, %)a Docosahexaenoic acid (22:6n-3, %) D5-Desaturase index D6-Desaturase indexa Stearoyl coA desaturase-1 indexa
71.3 (71.0–71.5) 95.1 9.2 26.3 3.4 3.6 (1.0–7.7) 5.2 2.1 0.38 (0.30–0.48) 9.6 (7.6–12.6) 0.48 (0.38–0.60) 1.3 (1.0–1.7) 11.7 0.9 3.4 (2.9–4.2) 0.92 (0.82–1.05) 20.4 2.2 52.5 4.3 0.60 (0.47–0.80) 0.83 (0.71–0.96) 0.71 0.15 5.8 1.1 1.6 (1.2–2.2) 0.97 0.24 8.5 2.0 0.012 (0.009–0.016) 0.30 (0.25–0.35)
Insulin sensitivity, glucose disposal measured by euglycemic insulin clamp; D5-desaturase = (20:4n-6/20:3n-6); D6-desaturase = (18:3n-6/18:2n-6); stearoyl coA desaturase-1 = (16:1/16:0). a Skewed distributed variables are presented as median (Q1–Q3).
WC, triglycerides and NEFA were positively related whereas insulin sensitivity and adiponectin were inversely related to ALT ( p 0.007, Table 3). After adjusting for abdominal obesity, lifestyle factors and insulin sensitivity in multivariable Model 1, palmitoleic acid (16:1), linoleic acid (18:2n-6) and SCD-1 index were the FA variables most strongly related to ALT ( p < 0.001) and were therefore included in multivariable Model 2. To investigate a possible independent role of FAs in the relationship with ALT, linoleic acid and SCD-1 index were analysed together with metabolic and lifestyle factors in a multivariable model (Model 2) (Table 3). Palmitoleic acid was not included in the analysis due to its inclusion in the SCD-1 index. None of the variables in multivariable Model 2 were interrelated with a correlation coefficient higher than 0.62. Insulin sensitivity ( p < 0.0001) and SCD-1 index ( p = 0.004) were significantly related to ALT, whereas none of the other factors remained associated. When including only one of the two FA variables at a time, the M-value ( p < 0.0001) and the FA variable (linoleic acid, p = 0.0004 and SCD-1 index, p < 0.0001, respectively) were associated with ALT.
5.
Discussion
In this population-based study, we assessed serum cholesterol ester FA composition, as a measure of both dietary FA intake and desaturation activities, to investigate the associations with liver fat markers (ALT). With regard to markers of dietary fat intake, proportions of linoleic acid (n-6), but not n-3 FAs, was inversely related to ALT independently of lifestyle factors, abdominal obesity and insulin resistance. Food sources containing n-6 FAs includes vegetable oils, margarines, nuts and seeds. With regard to endogenous FA metabolism, increased proportions of palmitoleic, oleic, g-linolenic and dihomo-glinolenic acids, and high D6-desaturase, SCD-1 and lipogenic indices were related to ALT independently of obesity, lifestyle factors and insulin resistance. SCD-1 may play a role in the development of fatty liver [8,9] and has been suggested as a therapeutic target in the treatment of obesity-related meta-
382
diabetes research and clinical practice 87 (2010) 379–384
Table 2 – Relationships of serum fatty acid composition and desaturase indices with serum alanine aminotransferase levels. Fatty acid standardised to 1 SD
Multivariable Model 1a
Univariable model p
b Palmitic acid (16:0) Palmitoleic acid (16:1) Stearic acid (18:0) Oleic acid (18:1) Linoleic acid (18:2n-6) g-Linolenic acid (18:3n-6) a-Linolenic acid (18:3n-3) Dihomo-g-linolenic acid (20:3n-6) Arachidonic acid (20:4n-6) Eicosapentaenoic acid (20:5n-3) Docosahexaenoic acid (22:6n-3) D5-desaturase index D6-desaturase index Stearoyl coA desaturase-1 index
0.06 0.11 0.02 0.05 0.09 0.08 0.03 0.11 0.04 0.05 0.02 0.05 0.09 0.10
0.0009 <0.0001 0.25 0.01 <0.0001 <0.0001 0.15 <0.0001 0.01 0.004 0.25 0.01 <0.0001 <0.0001
p
b 0.03 0.09 0.02 0.04 0.07 0.05 0.02 0.06 0.02 0.04 0.01 0.02 0.06 0.09
0.08 <0.0001 0.37 0.03 0.0002 0.02 0.25 0.002 0.25 0.03 0.53 0.34 0.004 <0.0001
b, regression coefficient; D5-desaturase = (20:4n-6/20:3n-6); D6-desaturase = (18:3n-6/18:2n-6); stearoyl coA desaturase-1 = (16:1/16:0). Model 1: n = 460, multivariable model adjusting for waist circumference, alcohol intake, physical activity and insulin sensitivity.
a
bolic disorders [16]. Interestingly, insulin sensitivity did not, in spite of its strong association with liver fat [1], abolish the relation between SCD-1 index and ALT. Instead, SCD-1 index and M, but not WC, triglycerides, NEFA and adiponectin, were significantly related to ALT in multivariable analysis. Our results are in accordance with previous studies relating FA composition in plasma, liver and adipose tissue to liver fat [6,17,18]. Those study samples were however small (n = 20–40) and not population-based, and since all subjects were obese it is difficult to interpret the role of fatty liver independent of obesity. In line with our results in Swedish men, Araya et al. reported lower linoleic acid levels in the total lipids of the liver in South American patients with steatosis, but in contrast found a positive association between linoleic acid measured in adipose tissue and liver triglycerides [17]. There are several potential explanations for the somewhat different results between that and the present study. First, there was a major difference in BMI between the two studies. Whereas the subjects of our cohort were rather lean (mean BMI 26) elderly men, the subjects in the South American study were extremely obese (BMI 46). Therefore they may have a modified fatty acid metabolism [19], such as impaired fat oxidation and down-regulated hepatic D6-desaturase
activity causing accumulation of linoleic acid, possibly as a consequence of long-term obesity and hyperinsulinemia, or due to drugs used for chronic treatment of various obesityrelated complications. Further, the genetic background as well as dietary fat intakes and lifestyle habits in South America (Chile) is very different from those in Sweden, differences that makes comparison of the two studies difficult. FAs were measured in different fat compartments. Since linoleic acid occurs at a high percentage in serum cholesterol esters, it may be easier to detect significant associations in this fat compartment. In accordance with our results, a high SCD-1 index (18:1/18:0) and low proportion of polyunsaturated FAs measured in the liver was recently associated with liver fat, also taking obesity into account [9]. In contrast, Stefan et al. showed a negative correlation between SCD-1 index (18:1/18:0) in serum and liver fat [20]. This association was however only seen in obese subjects and SCD-1 index was estimated by the serum 18:1/18:0-ratio rather than by using the 16:1/16:0-ratio. The 18:1/18:0-ratio was not associated to liver fat in the present study and by others [6]. One explanation for this may be that this index is largely reflected by the diet (e.g. meat fat and olive oil intake) whereas the 16:1/16:0-ratio mainly reflects endogenously
Table 3 – Associations between serum alanine aminotransferase levels and metabolic variables (n = 460). Variables standardised to 1 SD
Univariable regression b
Waist circumference Insulin sensitivity (M) Serum triglycerides Serum non-esterified fatty acids Linoleic acid (18:2n-6) Stearoyl coA desaturase-1 index Serum adiponectin
0.10 0.15 0.08 0.08 0.11 0.12 0.05
p <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.007
Multivariable Model 2 a b 0.02 0.13 0.0001 0.02 0.03 0.07 0.02
p 0.44 <0.0001 0.996 0.34 0.21 0.004 0.35
b, regression coefficient; M, glucose disposal measured by euglycemic insulin clamp; stearoyl coA desaturase-1 = (16:1/16:0). a Model 2 also included alcohol intake and physical activity. None of these variables were significantly related to ALT in the univariable or multivariable models and are therefore not shown.
diabetes research and clinical practice 87 (2010) 379–384
produced FA. Although the SCD-1 index reflects desaturase activity, it could also be affected by diet. Compared to high intake of polyunsaturated FA, a diet high in saturated fat increased the 16:1/16:0-ratio in a randomized controlled study [21]. The ratio may also increase in response to a diet high in sugars and very low in fat (10E%) [22]. Such extreme low-fat diet is however rare in this population (Sjo¨gren P, unpublished data). In line with the results in the present study, linoleic acid and SCD-1 index has also previously been related to CRP, an acute phase protein produced by the liver [23]. The observed association of an increased lipogenic activity and liver fat in the present study is in accordance with the previously reported correlation between lipogenic index (16:0/ 18:2n-6 ratio) and fatty liver [9]. The polyunsaturated FAs, glinolenic and dihomo-g-linolenic acids were positively related to ALT. These FAs are rare in the diet and mainly reflect endogenous FA metabolism. g-Linolenic and dihomo-g-linolenic acids have previously been related to increased diabetes risk [24], whereas high dietary intake of linoleic acid, the most common dietary polyunsaturated n-6 FA, has been consistently associated with decreased diabetes and improved insulin sensitivity [25]. There are several potential mechanisms that could explain our findings. Depending on distinct effects on lipid synthesis and FA oxidation, FAs may affect liver fat content differently. SCD-1 deficiency in mice promotes FA oxidation and decreases lipogenesis [16]. Polyunsaturated FA inhibits SCD-1 and D6desaturase by influencing transcription of sterol regulatory element binding protein (SREBP)-1c and peroxisome proliferator activated receptor (PPAR)-a. By reducing the active form of SREBP-1c, polyunsaturated FA suppresses transcription of genes involved in hepatic FA synthesis [7]. Polyunsaturated FA also serves as ligand for PPAR-a and PPAR-d and thereby stimulates hepatic b-oxidation [26]. Strengths of this study include the large populationbased sample with FA composition, desaturase activity indices and directly measured peripheral insulin sensitivity, as well as the inclusion of various confounding lifestyle and metabolic factors. Another advantage is the use of objective biomarkers of dietary FA intake instead of self-reported intake, which is associated with high reporting bias [5]. There are however limitations of this study. Since this is an observational cross-sectional study no conclusion regarding causality can be drawn. ALT activity was used as a liver fat marker instead of histology or imaging techniques. Although serum ALT correlates with liver fat it only partly explains the variation [1]. This would however weaken our observed associations rather than create false ones. Desaturase activities were estimated and not directly measured but desaturation indices are established in epidemiological studies [5] and are related to desaturase activities in vitro and in animals [27,28]. The FA composition was measured in a subgroup but since this sub-sample did not differ from those without FA composition measurement with regard to WC, BMI, ALT, M, and triglycerides, the generalizability was probably not affected. Only men at the same age participated without data in women, other ethnic groups, or other ages. Alcohol intake was self-reported and thus possibly is underreported.
383
In summary, we observed an association between serum FA composition and ALT activity, independent of obesity and insulin resistance. The results partly support a potential protective role of dietary polyunsaturated fat (linoleic acid) in fatty liver. Endogenously synthesized FAs were even more closely associated with fatty liver. Especially, a significant role of SCD-1 was evident since its link to ALT was independent of other factors known to be closely associated with fatty liver, i.e. abdominal obesity, insulin resistance, dyslipidemia and adiponectin concentrations. Whether SCD-1 activity per se may be causally involved in the development of fatty liver is unknown, and a potential anti-lipogenic effect in the liver of dietary linoleic acid needs to be investigated in future controlled dietary interventions.
Conflict of interest The authors declare that they have no conflict of interest.
Acknowledgements This study was supported by NordForsk (Nordic Centre of Excellence in Food, Nutrition and health [SYSDIET]), Swedish Nutrition Foundation (SNF), and Swedish council for working life and social research (FAS).
references
[1] A. Kotronen, H. Yki-Jarvinen, Fatty liver: a novel component of the metabolic syndrome, Arterioscler. Thromb. Vasc. Biol. 28 (2008) 27–38. [2] J. Westerbacka, K. Lammi, A.M. Hakkinen, A. Rissanen, I. Salminen, A. Aro, et al., Dietary fat content modifies liver fat in overweight nondiabetic subjects, J. Clin. Endocrinol. Metab. 90 (2005) 2804–2809. [3] M. Tiikkainen, R. Bergholm, S. Vehkavaara, A. Rissanen, A.M. Hakkinen, M. Tamminen, et al., Effects of identical weight loss on body composition and features of insulin resistance in obese women with high and low liver fat content, Diabetes 52 (2003) 701–707. [4] M. Sekiya, N. Yahagi, T. Matsuzaka, Y. Najima, M. Nakakuki, R. Nagai, et al., Polyunsaturated fatty acids ameliorate hepatic steatosis in obese mice by SREBP-1 suppression, Hepatology 38 (2003) 1529–1539. [5] L. Hodson, C.M. Skeaff, B.A. Fielding, Fatty acid composition of adipose tissue and blood in humans and its use as a biomarker of dietary intake, Prog. Lipid Res. 47 (2008) 348– 380. [6] I.T. de Almeida, H. Cortez-Pinto, G. Fidalgo, D. Rodrigues, M.E. Camilo, Plasma total and free fatty acids composition in human non-alcoholic steatohepatitis, Clin. Nutr. 21 (2002) 219–223. [7] M.T. Nakamura, T.Y. Nara, Structure, function, and dietary regulation of delta6, delta5, and delta9 desaturases, Annu. Rev. Nutr. 24 (2004) 345–376. [8] K. Chu, M. Miyazaki, W.C. Man, J.M. Ntambi, Stearoylcoenzyme A desaturase 1 deficiency protects against hypertriglyceridemia and increases plasma high-density lipoprotein cholesterol induced by liver X receptor activation, Mol. Cell. Biol. 26 (2006) 6786–6798.
384
diabetes research and clinical practice 87 (2010) 379–384
[9] A. Kotronen, T. Seppanen-Laakso, J. Westerbacka, T. Kiviluoto, J. Arola, A.L. Ruskeepaa, et al., Hepatic stearoylCoA desaturase (SCD)-1 activity and diacylglycerol but not ceramide concentrations are increased in the nonalcoholic human fatty liver, Diabetes 58 (2009) 203–208. [10] H. Hedstrand, A study of middle-aged men with particular reference to risk factors for cardiovascular disease, Ups. J. Med. Sci. Suppl. 19 (1975) 1–61. [11] E.T. Skarfors, K.I. Selinus, H.O. Lithell, Risk factors for developing non-insulin dependent diabetes: a 10 year follow up of men in Uppsala, BMJ 303 (1991) 755–760. [12] M. Boberg, L.B. Croon, I.B. Gustafsson, B. Vessby, Platelet fatty acid composition in relation to fatty acid composition in plasma and to serum lipoprotein lipids in healthy subjects with special reference to the linoleic acid pathway, Clin. Sci. (Lond.) 68 (1985) 581–587. [13] R.A. DeFronzo, J.D. Tobin, R. Andres, Glucose clamp technique: a method for quantifying insulin secretion and resistance, Am. J. Physiol. 237 (1979) E214–223. [14] L. Byberg, B. Zethelius, P.M. McKeigue, H.O. Lithell, Changes in physical activity are associated with changes in metabolic cardiovascular risk factors, Diabetologia 44 (2001) 2134–2139. [15] U. Riserus, E. Ingelsson, Alcohol intake, insulin resistance, and abdominal obesity in elderly men, Obesity (Silver Spring) 15 (2007) 1766–1773. [16] A. Dobrzyn, J.M. Ntambi, Stearoyl-CoA desaturase as a new drug target for obesity treatment, Obes. Rev. 6 (2005) 169–174. [17] J. Araya, R. Rodrigo, L.A. Videla, L. Thielemann, M. Orellana, P. Pettinelli, et al., Increase in long-chain polyunsaturated fatty acid n-6/n-3 ratio in relation to hepatic steatosis in patients with non-alcoholic fatty liver disease, Clin. Sci. (Lond.) 106 (2004) 635–643. [18] J.P. Allard, E. Aghdassi, S. Mohammed, M. Raman, G. Avand, B.M. Arendt, et al., Nutritional assessment and hepatic fatty acid composition in non-alcoholic fatty liver disease (NAFLD): a cross-sectional study, J. Hepatol. 48 (2008) 300– 307. [19] K.R. Westerterp, A. Smeets, M.P. Lejeune, M.P. WoutersAdriaens, M.S. Westerterp-Plantenga, Dietary fat oxidation
[20]
[21]
[22]
[23]
[24]
[25]
[26] [27]
[28]
as a function of body fat, Am. J. Clin. Nutr. 87 (2008) 132– 135. N. Stefan, A. Peter, A. Cegan, H. Staiger, J. Machann, F. Schick, et al., Low hepatic stearoyl-CoA desaturase 1 activity is associated with fatty liver and insulin resistance in obese humans, Diabetologia 51 (2008) 648–656. E. Warensjo, U. Riserus, I.B. Gustafsson, R. Mohsen, T. Cederholm, B. Vessby, Effects of saturated and unsaturated fatty acids on estimated desaturase activities during a controlled dietary intervention, Nutr. Metab. Cardiovasc. Dis. 18 (2008) 683–690. M.F. Chong, L. Hodson, A.S. Bickerton, R. Roberts, M. Neville, F. Karpe, et al., Parallel activation of de novo lipogenesis and stearoyl-CoA desaturase activity after 3 d of high-carbohydrate feeding, Am. J. Clin. Nutr. 87 (2008) 817–823. H. Petersson, S. Basu, T. Cederholm, U. Riserus, Serum fatty acid composition and indices of stearoyl-CoA desaturase activity are associated with systemic inflammation: longitudinal analyses in middle-aged men, Br. J. Nutr. 99 (2008) 1186–1189. L. Wang, A.R. Folsom, Z.J. Zheng, J.S. Pankow, J.H. Eckfeldt, Plasma fatty acid composition and incidence of diabetes in middle-aged adults: the Atherosclerosis Risk in Communities (ARIC) Study, Am. J. Clin. Nutr. 78 (2003) 91– 98. U. Riserus, W.C. Willett, F.B. Hu, Dietary fats and prevention of type 2 diabetes, Prog. Lipid Res. 48 (2009) 44– 51. B. Desvergne, L. Michalik, W. Wahli, Transcriptional regulation of metabolism, Physiol. Rev. 86 (2006) 465–514. A.D. Attie, R.M. Krauss, M.P. Gray-Keller, A. Brownlie, M. Miyazaki, J.J. Kastelein, et al., Relationship between stearoyl-CoA desaturase activity and plasma triglycerides in human and mouse hypertriglyceridemia, J. Lipid Res. 43 (2002) 1899–1907. G. Liu, J.K. Lynch, J. Freeman, B. Liu, Z. Xin, H. Zhao, et al., Discovery of potent, selective, orally bioavailable stearoylCoA desaturase 1 inhibitors, J. Med. Chem. 50 (2007) 3086– 3100.
Contents lists available at ScienceDirect
Diabetes Research and Clinical Practice journ al h omepage: www .elsevier.co m/lo cate/diabres
Serum fatty acid composition and insulin resistance are independently associated with liver fat markers in elderly men§ ¨ rnlo¨v b,c, Bjo¨rn Zethelius b, Ulf Rise´rus a Helena Petersson a,*, Johan A a
Department of Public Health and Caring Sciences/Clinical Nutrition and Metabolism, Uppsala University, Uppsala Science Park, 75185 Uppsala, Sweden b Department of Public Health and Caring Sciences/Geriatrics, Uppsala University, Uppsala Science Park, 75185 Uppsala, Sweden c Department of Health and Social Sciences, Ho¨gskolan Dalarna, 79188 Falun, Sweden
article info
abstract
Article history:
Aim: To investigate the relationships of serum fatty acid (FA) composition and estimated
Received 13 August 2009
desaturase activities with the liver fat marker alanine aminotransferase (ALT).
Received in revised form
Methods: 546 Swedish elderly men of a population-based cohort participated in this cross-
23 October 2009
sectional study. FA composition was assessed in serum cholesterol esters to determine
Accepted 24 November 2009
dietary fat quality (e.g. linoleic) and desaturation products (e.g. dihomo-g-linolenic acid). Desaturase indices, including stearoyl coenzymeA desaturase-1 (SCD-1), were calculated by FA product-to-precursor ratios.
Keywords:
Results: In linear regression analyses adjusting for lifestyle, abdominal obesity and insulin
Fatty acid composition
sensitivity, the dietary biomarker linoleic acid (n-6), but not n-3 FAs, was inversely related to
Liver fat
ALT. Desaturation products including palmitoleic, oleic, g-linolenic and dihomo-g-linolenic
Lipogenesis
acids, and D6-desaturase and SCD-1 indices were directly related to ALT (all p < 0.05). After
Stearoyl coenzymeA desaturase-1
further adjustment for factors previously linked to fatty liver (i.e. serum lipids, adiponectin concentrations), SCD-1 index ( p = 0.004) and insulin resistance ( p < 0.0001) were independent determinants of ALT activity, whereas waist circumference, triglycerides, non-esterified FA and adiponectin were not. Conclusion: A low dietary intake of linoleic acid and elevated SCD-1 index may contribute to higher ALT activity in elderly men, even independently of obesity and insulin resistance. # 2009 Elsevier Ireland Ltd. All rights reserved.
1.
Introduction
Non-alcoholic fatty liver is related to obesity, especially abdominal obesity, and is strongly associated with insulin resistance, the metabolic syndrome and type 2 diabetes. A fatty liver also overproduces several cardiovascular risk factors, such as glucose, very low density lipoprotein triglycerides, coagulation factors, and C-reactive protein [1]. The pathogenesis of fatty liver is poorly understood. Dietary §
factors could play a role and recently a low-fat diet was reported to reduce liver fat [2]. Furthermore it was recently shown that self-reported saturated fat intake, was directly associated with liver fat [3], whereas animal studies suggest a protective effect of polyunsaturated fat [4]. Due to major reporting bias of dietary fat intake assessed by food records it is relevant to investigate this relation using objective biomarkers of dietary fat quality, i.e. serum fatty acid (FA) composition [5]. Serum FA composition has previously been
This study was presented at the 44th annual meeting of the European Association for the Study of Diabetes (EASD) 2008. * Corresponding author. Tel.: +46 18 6117980; fax: +46 18 6117976. E-mail address: [email protected] (H. Petersson). 0168-8227/$ – see front matter # 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.diabres.2009.11.019
380
diabetes research and clinical practice 87 (2010) 379–384
related to fatty liver in a limited number of subjects [6], but studies in larger population-based samples are lacking. Also, studies taking abdominal obesity, insulin resistance, adiponectin and other potential confounders into account are not available. In general, the essential FAs linoleic acid (18:2n-6) and a-linolenic acid (18:3n-3) are good markers of dietary intake of these FAs whereas saturated and monounsaturated FAs are somewhat weaker biomarkers [5]. In addition to the dietary intake, specific serum FAs also reflect endogenous fat metabolism and FA desaturation, e.g. palmitoleic, g-linolenic and dihomo-g-linolenic acids. Desaturases induce double-bounds in the FA chains. Stearoyl coenzymeA desaturase (SCD)-1 is the final step in de novo lipogenesis and converts saturated FA to monounsaturated FA, whereas D5- and D6-desaturases participate in the metabolism of polyunsaturated FAs [7]. SCD-1 is of high interest as this lipogenic enzyme has been suggested to play a role in the development of fatty liver [8,9]. In epidemiological studies, desaturase activities can be estimated using FA product-to-precursor ratios. To date, population-based data are lacking regarding the association between markers of endogenous FA metabolism and fatty liver. Based on previous data, we hypothesized that a diet high in saturated fat and low in polyunsaturated fat as well as increased desaturase activity could be involved in the development of non-alcoholic fatty liver. Therefore, we investigated the associations between serum FA composition, reflecting both dietary intake of FAs and endogenous fat metabolism, and alanine aminotransferase (ALT) levels, reflecting liver fat, in a large population-based sample of elderly men. To further explore a potential independent role of dietary FAs and/or desaturases we adjusted for lifestyle factors, as well as various metabolic factors previously linked to fatty liver, i.e. abdominal obesity, insulin resistance (measured directly by euglycemic insulin clamp), plasma adiponectin concentrations, triglycerides and non-esterified FA (NEFA).
2.
Subjects
The subjects were participants of the Uppsala Longitudinal Study of Adult Men (ULSAM, http://www.pubcare.uu.se/ ULSAM/), a population-based prospective study starting in 1970 in Uppsala, Sweden. Twelve-hundred twenty-one (73%) of 1681 men participated in the third investigation cycle that was carried out between 1991 and 1995, when the participants were 71 (0.6) years old. Eleven-hundred eighty-men had measure of ALT. Of these, 583 subjects had data on serum FA composition. Participants with a previous diagnosis of liver disease in the compulsory Swedish hospital discharge register (ICD-8 codes 570-573, ICD-9 codes 570-573 or ICD-10 codes K70-K77) were excluded (n = 4). Subjects consuming 20 g alcohol/day (n = 30) were excluded to avoid alcohol-related fatty liver. Three subjects had ALT levels below detection level (<0.10 mkat/l) and were excluded. Thus, 546 subjects were included in the present study. The study was approved by the Ethics Committee of Uppsala University. All subjects gave written informed consent.
3.
Materials and methods
3.1.
Clinical assessment and biochemistry
The investigation included anthropometric measurements, blood biochemistry, blood pressure, a self-administered medical questionnaire and an interview. All measurements were performed under standardised conditions as previously described [10,11]. Height was measured to the nearest centimeter and body weight to the nearest 0.1 kg. Body mass index (BMI) was calculated as the ratio of weight to height squared (kg/ m2). Waist circumference (WC) was measured midway between the lowest rib and the iliac crest in a supine position. All blood samples were drawn after an overnight fast. Serum triglycerides were analysed by enzymatic technique using IL Test Enzymatic Method for use in a Monarch apparatus (Instrumentation Laboratories, Lexington, USA). Serum NEFA was measured by an enzymatic colorimetric method (Wako Chemicals, Neuss, Germany). Plasma insulin was measured using an enzymatic-immunological assay (Enzymmun, Boehringer Mannheim, Germany) performed in an ES300 automatic analyser (Boehringer Mannheim). FA composition in serum cholesterol esters were analysed as previously described [12]. Serum was extracted with chloroform in the presence of methanol, butylated hydroxytoluene and NaH2PO4. Lipids were separated by thin liquid chromatography and trans methylated. The percentage composition of methylated FA 16:0–22:6 was determined by gas chromatography (a 25 m NB-351 silica capillary column, i.d. 0.32 mm, phase layer 0.20 mm) with helium as a carrier gas. Desaturase activities were estimated by the following FA product-to-precursor ratios; D5-desaturase (20:4n-6/20:3n-6), D6-desaturase (18:3n-6/18:2n-6), and SCD-1 (16:1n-7/16:0). Lipogenic index was calculated as 16:0/18:2n-6 ratio [9]. ALT in serum was used as a surrogate marker of fatty liver and analysed by the enzymatic method ECCLS/IFCC (Hitachi 717, Boehringer Mannheim, Germany) with CV = 1.6% at 1.30 mkat/l. Insulin sensitivity was determined by euglycemic insulin clamp according to DeFronzo et al. [13], slightly modified. The insulin infusion rate was 56 mU/min per body surface area (m2), instead of 40. Plasma glucose concentrations were maintained at 5.1 mmol/l during the clamp by adjusting the rate of infusion of a 20% glucose solution. Glucose disposal; M (mg/kg body weight/min), was calculated as the amount of glucose taken up during the last 60 min of the clamp. Serum adiponectin was analysed using a validated in-house timeresolved immunofluorometric assay (TR-IFMA) based on commercial reagents from R&D Systems (Abingdon, UK). Physical activity (categorised as sedentary, moderate, regular and athletic) was assessed by a self-administered questionnaire [14]. Alcohol intake was assessed by a 7-day dietary record [15].
3.2.
Statistical analyses
The distribution of the variables was examined by Shapiro–Wilk W-test. In order to promote a normal distribution, the following variables were logarithmically transformed before analysis; ALT, 16:1, 18:0, 18:3n-6, 18:3n-3, 20:5n-3, D6-desaturase, SCD-1, fasting insulin, NEFA, triglycerides, and adiponectin. Alcohol intake was divided into quartiles. Linear regression analysis was used to associate FA composition (standardised to one
diabetes research and clinical practice 87 (2010) 379–384
standard deviation) with ALT. The relation between FA composition and ALT was adjusted for abdominal obesity (measured by WC), lifestyle factors (alcohol intake, physical activity) and insulin sensitivity in a multivariable analysis (Model 1, n = 460). To further explore a potential independent link between FAs and liver fat, we added metabolic factors (standardised to one standard deviation) previously related to fatty liver; triglycerides, NEFA and adiponectin (Model 2, n = 460). To avoid colinearity in Model 2, we only included FA variables with a p-value <0.001 in Model 1 (i.e. linoleic acid and SCD-1 index). Since this study included a subgroup of the ULSAM cohort, potential selection bias was investigated by comparing participants with FA measurements (n = 546) with those without FA measurements as regard to ALT (n = 634), WC, BMI, M and triglycerides. A JMP software package was used for statistics (SAS Institute, Cary, NC). p < 0.05 was considered as significant.
381
FA composition with regards to WC, BMI, ALT, M and triglycerides ( p > 0.28 for all).
4.1. Associations between FA composition and ALT— influence of lifestyle factors and insulin sensitivity The association between FA composition and ALT is presented in Table 2 as univariable and multivariable analyses adjusting for WC, alcohol intake, physical activity and insulin sensitivity (multivariable Model 1). The results were similar when adjusting for BMI instead of WC or if M was substituted for fasting insulin. When excluding subjects with diabetes (n = 76), the trend remained for all FAs with the exception of 18:1. Lipogenic index was related to ALT in univariable analysis (b = 0.08, p < 0.0001) and multivariable Model 1 (b = 0.06, p = 0.003).
4.2. Associations between FA composition and ALT— influence of other metabolic factors
4.
Results
Population characteristics are presented in Table 1. The range of ALT activity was between 0.10 and 2.55 mkat/l. The subsample of participants with measurements of FA composition did not differ from the sub-sample without measurements of
Table 1 – Population characteristics. Variable
Mean SD a
Age (years) Waist circumference (cm) BMI (kg/m2) Alcohol intake (g/day)a Insulin sensitivity (M, mg/kg body weight/min) Alanine aminotransferase (mkat/l)a Adiponectin (ng/ml)a Non-esterified fatty acids (mmol/l)a Triglycerides (mmol/l)a Palmitic acid (16:0, %) Palmitoleic acid (16:1, %)a Stearic acid (18:0, %)a Oleic acid (18:1, %) Linoleic acid (18:2n-6, %) g-Linolenic acid (18:3n-6, %)a a-Linolenic acid (18:3n-3, %)a Dihomo-g-linolenic acid (20:3n-6, %) Arachidonic acid (20:4n-6, %) Eicosapentaenoic acid (20:5n-3, %)a Docosahexaenoic acid (22:6n-3, %) D5-Desaturase index D6-Desaturase indexa Stearoyl coA desaturase-1 indexa
71.3 (71.0–71.5) 95.1 9.2 26.3 3.4 3.6 (1.0–7.7) 5.2 2.1 0.38 (0.30–0.48) 9.6 (7.6–12.6) 0.48 (0.38–0.60) 1.3 (1.0–1.7) 11.7 0.9 3.4 (2.9–4.2) 0.92 (0.82–1.05) 20.4 2.2 52.5 4.3 0.60 (0.47–0.80) 0.83 (0.71–0.96) 0.71 0.15 5.8 1.1 1.6 (1.2–2.2) 0.97 0.24 8.5 2.0 0.012 (0.009–0.016) 0.30 (0.25–0.35)
Insulin sensitivity, glucose disposal measured by euglycemic insulin clamp; D5-desaturase = (20:4n-6/20:3n-6); D6-desaturase = (18:3n-6/18:2n-6); stearoyl coA desaturase-1 = (16:1/16:0). a Skewed distributed variables are presented as median (Q1–Q3).
WC, triglycerides and NEFA were positively related whereas insulin sensitivity and adiponectin were inversely related to ALT ( p 0.007, Table 3). After adjusting for abdominal obesity, lifestyle factors and insulin sensitivity in multivariable Model 1, palmitoleic acid (16:1), linoleic acid (18:2n-6) and SCD-1 index were the FA variables most strongly related to ALT ( p < 0.001) and were therefore included in multivariable Model 2. To investigate a possible independent role of FAs in the relationship with ALT, linoleic acid and SCD-1 index were analysed together with metabolic and lifestyle factors in a multivariable model (Model 2) (Table 3). Palmitoleic acid was not included in the analysis due to its inclusion in the SCD-1 index. None of the variables in multivariable Model 2 were interrelated with a correlation coefficient higher than 0.62. Insulin sensitivity ( p < 0.0001) and SCD-1 index ( p = 0.004) were significantly related to ALT, whereas none of the other factors remained associated. When including only one of the two FA variables at a time, the M-value ( p < 0.0001) and the FA variable (linoleic acid, p = 0.0004 and SCD-1 index, p < 0.0001, respectively) were associated with ALT.
5.
Discussion
In this population-based study, we assessed serum cholesterol ester FA composition, as a measure of both dietary FA intake and desaturation activities, to investigate the associations with liver fat markers (ALT). With regard to markers of dietary fat intake, proportions of linoleic acid (n-6), but not n-3 FAs, was inversely related to ALT independently of lifestyle factors, abdominal obesity and insulin resistance. Food sources containing n-6 FAs includes vegetable oils, margarines, nuts and seeds. With regard to endogenous FA metabolism, increased proportions of palmitoleic, oleic, g-linolenic and dihomo-glinolenic acids, and high D6-desaturase, SCD-1 and lipogenic indices were related to ALT independently of obesity, lifestyle factors and insulin resistance. SCD-1 may play a role in the development of fatty liver [8,9] and has been suggested as a therapeutic target in the treatment of obesity-related meta-
382
diabetes research and clinical practice 87 (2010) 379–384
Table 2 – Relationships of serum fatty acid composition and desaturase indices with serum alanine aminotransferase levels. Fatty acid standardised to 1 SD
Multivariable Model 1a
Univariable model p
b Palmitic acid (16:0) Palmitoleic acid (16:1) Stearic acid (18:0) Oleic acid (18:1) Linoleic acid (18:2n-6) g-Linolenic acid (18:3n-6) a-Linolenic acid (18:3n-3) Dihomo-g-linolenic acid (20:3n-6) Arachidonic acid (20:4n-6) Eicosapentaenoic acid (20:5n-3) Docosahexaenoic acid (22:6n-3) D5-desaturase index D6-desaturase index Stearoyl coA desaturase-1 index
0.06 0.11 0.02 0.05 0.09 0.08 0.03 0.11 0.04 0.05 0.02 0.05 0.09 0.10
0.0009 <0.0001 0.25 0.01 <0.0001 <0.0001 0.15 <0.0001 0.01 0.004 0.25 0.01 <0.0001 <0.0001
p
b 0.03 0.09 0.02 0.04 0.07 0.05 0.02 0.06 0.02 0.04 0.01 0.02 0.06 0.09
0.08 <0.0001 0.37 0.03 0.0002 0.02 0.25 0.002 0.25 0.03 0.53 0.34 0.004 <0.0001
b, regression coefficient; D5-desaturase = (20:4n-6/20:3n-6); D6-desaturase = (18:3n-6/18:2n-6); stearoyl coA desaturase-1 = (16:1/16:0). Model 1: n = 460, multivariable model adjusting for waist circumference, alcohol intake, physical activity and insulin sensitivity.
a
bolic disorders [16]. Interestingly, insulin sensitivity did not, in spite of its strong association with liver fat [1], abolish the relation between SCD-1 index and ALT. Instead, SCD-1 index and M, but not WC, triglycerides, NEFA and adiponectin, were significantly related to ALT in multivariable analysis. Our results are in accordance with previous studies relating FA composition in plasma, liver and adipose tissue to liver fat [6,17,18]. Those study samples were however small (n = 20–40) and not population-based, and since all subjects were obese it is difficult to interpret the role of fatty liver independent of obesity. In line with our results in Swedish men, Araya et al. reported lower linoleic acid levels in the total lipids of the liver in South American patients with steatosis, but in contrast found a positive association between linoleic acid measured in adipose tissue and liver triglycerides [17]. There are several potential explanations for the somewhat different results between that and the present study. First, there was a major difference in BMI between the two studies. Whereas the subjects of our cohort were rather lean (mean BMI 26) elderly men, the subjects in the South American study were extremely obese (BMI 46). Therefore they may have a modified fatty acid metabolism [19], such as impaired fat oxidation and down-regulated hepatic D6-desaturase
activity causing accumulation of linoleic acid, possibly as a consequence of long-term obesity and hyperinsulinemia, or due to drugs used for chronic treatment of various obesityrelated complications. Further, the genetic background as well as dietary fat intakes and lifestyle habits in South America (Chile) is very different from those in Sweden, differences that makes comparison of the two studies difficult. FAs were measured in different fat compartments. Since linoleic acid occurs at a high percentage in serum cholesterol esters, it may be easier to detect significant associations in this fat compartment. In accordance with our results, a high SCD-1 index (18:1/18:0) and low proportion of polyunsaturated FAs measured in the liver was recently associated with liver fat, also taking obesity into account [9]. In contrast, Stefan et al. showed a negative correlation between SCD-1 index (18:1/18:0) in serum and liver fat [20]. This association was however only seen in obese subjects and SCD-1 index was estimated by the serum 18:1/18:0-ratio rather than by using the 16:1/16:0-ratio. The 18:1/18:0-ratio was not associated to liver fat in the present study and by others [6]. One explanation for this may be that this index is largely reflected by the diet (e.g. meat fat and olive oil intake) whereas the 16:1/16:0-ratio mainly reflects endogenously
Table 3 – Associations between serum alanine aminotransferase levels and metabolic variables (n = 460). Variables standardised to 1 SD
Univariable regression b
Waist circumference Insulin sensitivity (M) Serum triglycerides Serum non-esterified fatty acids Linoleic acid (18:2n-6) Stearoyl coA desaturase-1 index Serum adiponectin
0.10 0.15 0.08 0.08 0.11 0.12 0.05
p <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.007
Multivariable Model 2 a b 0.02 0.13 0.0001 0.02 0.03 0.07 0.02
p 0.44 <0.0001 0.996 0.34 0.21 0.004 0.35
b, regression coefficient; M, glucose disposal measured by euglycemic insulin clamp; stearoyl coA desaturase-1 = (16:1/16:0). a Model 2 also included alcohol intake and physical activity. None of these variables were significantly related to ALT in the univariable or multivariable models and are therefore not shown.
diabetes research and clinical practice 87 (2010) 379–384
produced FA. Although the SCD-1 index reflects desaturase activity, it could also be affected by diet. Compared to high intake of polyunsaturated FA, a diet high in saturated fat increased the 16:1/16:0-ratio in a randomized controlled study [21]. The ratio may also increase in response to a diet high in sugars and very low in fat (10E%) [22]. Such extreme low-fat diet is however rare in this population (Sjo¨gren P, unpublished data). In line with the results in the present study, linoleic acid and SCD-1 index has also previously been related to CRP, an acute phase protein produced by the liver [23]. The observed association of an increased lipogenic activity and liver fat in the present study is in accordance with the previously reported correlation between lipogenic index (16:0/ 18:2n-6 ratio) and fatty liver [9]. The polyunsaturated FAs, glinolenic and dihomo-g-linolenic acids were positively related to ALT. These FAs are rare in the diet and mainly reflect endogenous FA metabolism. g-Linolenic and dihomo-g-linolenic acids have previously been related to increased diabetes risk [24], whereas high dietary intake of linoleic acid, the most common dietary polyunsaturated n-6 FA, has been consistently associated with decreased diabetes and improved insulin sensitivity [25]. There are several potential mechanisms that could explain our findings. Depending on distinct effects on lipid synthesis and FA oxidation, FAs may affect liver fat content differently. SCD-1 deficiency in mice promotes FA oxidation and decreases lipogenesis [16]. Polyunsaturated FA inhibits SCD-1 and D6desaturase by influencing transcription of sterol regulatory element binding protein (SREBP)-1c and peroxisome proliferator activated receptor (PPAR)-a. By reducing the active form of SREBP-1c, polyunsaturated FA suppresses transcription of genes involved in hepatic FA synthesis [7]. Polyunsaturated FA also serves as ligand for PPAR-a and PPAR-d and thereby stimulates hepatic b-oxidation [26]. Strengths of this study include the large populationbased sample with FA composition, desaturase activity indices and directly measured peripheral insulin sensitivity, as well as the inclusion of various confounding lifestyle and metabolic factors. Another advantage is the use of objective biomarkers of dietary FA intake instead of self-reported intake, which is associated with high reporting bias [5]. There are however limitations of this study. Since this is an observational cross-sectional study no conclusion regarding causality can be drawn. ALT activity was used as a liver fat marker instead of histology or imaging techniques. Although serum ALT correlates with liver fat it only partly explains the variation [1]. This would however weaken our observed associations rather than create false ones. Desaturase activities were estimated and not directly measured but desaturation indices are established in epidemiological studies [5] and are related to desaturase activities in vitro and in animals [27,28]. The FA composition was measured in a subgroup but since this sub-sample did not differ from those without FA composition measurement with regard to WC, BMI, ALT, M, and triglycerides, the generalizability was probably not affected. Only men at the same age participated without data in women, other ethnic groups, or other ages. Alcohol intake was self-reported and thus possibly is underreported.
383
In summary, we observed an association between serum FA composition and ALT activity, independent of obesity and insulin resistance. The results partly support a potential protective role of dietary polyunsaturated fat (linoleic acid) in fatty liver. Endogenously synthesized FAs were even more closely associated with fatty liver. Especially, a significant role of SCD-1 was evident since its link to ALT was independent of other factors known to be closely associated with fatty liver, i.e. abdominal obesity, insulin resistance, dyslipidemia and adiponectin concentrations. Whether SCD-1 activity per se may be causally involved in the development of fatty liver is unknown, and a potential anti-lipogenic effect in the liver of dietary linoleic acid needs to be investigated in future controlled dietary interventions.
Conflict of interest The authors declare that they have no conflict of interest.
Acknowledgements This study was supported by NordForsk (Nordic Centre of Excellence in Food, Nutrition and health [SYSDIET]), Swedish Nutrition Foundation (SNF), and Swedish council for working life and social research (FAS).
references
[1] A. Kotronen, H. Yki-Jarvinen, Fatty liver: a novel component of the metabolic syndrome, Arterioscler. Thromb. Vasc. Biol. 28 (2008) 27–38. [2] J. Westerbacka, K. Lammi, A.M. Hakkinen, A. Rissanen, I. Salminen, A. Aro, et al., Dietary fat content modifies liver fat in overweight nondiabetic subjects, J. Clin. Endocrinol. Metab. 90 (2005) 2804–2809. [3] M. Tiikkainen, R. Bergholm, S. Vehkavaara, A. Rissanen, A.M. Hakkinen, M. Tamminen, et al., Effects of identical weight loss on body composition and features of insulin resistance in obese women with high and low liver fat content, Diabetes 52 (2003) 701–707. [4] M. Sekiya, N. Yahagi, T. Matsuzaka, Y. Najima, M. Nakakuki, R. Nagai, et al., Polyunsaturated fatty acids ameliorate hepatic steatosis in obese mice by SREBP-1 suppression, Hepatology 38 (2003) 1529–1539. [5] L. Hodson, C.M. Skeaff, B.A. Fielding, Fatty acid composition of adipose tissue and blood in humans and its use as a biomarker of dietary intake, Prog. Lipid Res. 47 (2008) 348– 380. [6] I.T. de Almeida, H. Cortez-Pinto, G. Fidalgo, D. Rodrigues, M.E. Camilo, Plasma total and free fatty acids composition in human non-alcoholic steatohepatitis, Clin. Nutr. 21 (2002) 219–223. [7] M.T. Nakamura, T.Y. Nara, Structure, function, and dietary regulation of delta6, delta5, and delta9 desaturases, Annu. Rev. Nutr. 24 (2004) 345–376. [8] K. Chu, M. Miyazaki, W.C. Man, J.M. Ntambi, Stearoylcoenzyme A desaturase 1 deficiency protects against hypertriglyceridemia and increases plasma high-density lipoprotein cholesterol induced by liver X receptor activation, Mol. Cell. Biol. 26 (2006) 6786–6798.
384
diabetes research and clinical practice 87 (2010) 379–384
[9] A. Kotronen, T. Seppanen-Laakso, J. Westerbacka, T. Kiviluoto, J. Arola, A.L. Ruskeepaa, et al., Hepatic stearoylCoA desaturase (SCD)-1 activity and diacylglycerol but not ceramide concentrations are increased in the nonalcoholic human fatty liver, Diabetes 58 (2009) 203–208. [10] H. Hedstrand, A study of middle-aged men with particular reference to risk factors for cardiovascular disease, Ups. J. Med. Sci. Suppl. 19 (1975) 1–61. [11] E.T. Skarfors, K.I. Selinus, H.O. Lithell, Risk factors for developing non-insulin dependent diabetes: a 10 year follow up of men in Uppsala, BMJ 303 (1991) 755–760. [12] M. Boberg, L.B. Croon, I.B. Gustafsson, B. Vessby, Platelet fatty acid composition in relation to fatty acid composition in plasma and to serum lipoprotein lipids in healthy subjects with special reference to the linoleic acid pathway, Clin. Sci. (Lond.) 68 (1985) 581–587. [13] R.A. DeFronzo, J.D. Tobin, R. Andres, Glucose clamp technique: a method for quantifying insulin secretion and resistance, Am. J. Physiol. 237 (1979) E214–223. [14] L. Byberg, B. Zethelius, P.M. McKeigue, H.O. Lithell, Changes in physical activity are associated with changes in metabolic cardiovascular risk factors, Diabetologia 44 (2001) 2134–2139. [15] U. Riserus, E. Ingelsson, Alcohol intake, insulin resistance, and abdominal obesity in elderly men, Obesity (Silver Spring) 15 (2007) 1766–1773. [16] A. Dobrzyn, J.M. Ntambi, Stearoyl-CoA desaturase as a new drug target for obesity treatment, Obes. Rev. 6 (2005) 169–174. [17] J. Araya, R. Rodrigo, L.A. Videla, L. Thielemann, M. Orellana, P. Pettinelli, et al., Increase in long-chain polyunsaturated fatty acid n-6/n-3 ratio in relation to hepatic steatosis in patients with non-alcoholic fatty liver disease, Clin. Sci. (Lond.) 106 (2004) 635–643. [18] J.P. Allard, E. Aghdassi, S. Mohammed, M. Raman, G. Avand, B.M. Arendt, et al., Nutritional assessment and hepatic fatty acid composition in non-alcoholic fatty liver disease (NAFLD): a cross-sectional study, J. Hepatol. 48 (2008) 300– 307. [19] K.R. Westerterp, A. Smeets, M.P. Lejeune, M.P. WoutersAdriaens, M.S. Westerterp-Plantenga, Dietary fat oxidation
[20]
[21]
[22]
[23]
[24]
[25]
[26] [27]
[28]
as a function of body fat, Am. J. Clin. Nutr. 87 (2008) 132– 135. N. Stefan, A. Peter, A. Cegan, H. Staiger, J. Machann, F. Schick, et al., Low hepatic stearoyl-CoA desaturase 1 activity is associated with fatty liver and insulin resistance in obese humans, Diabetologia 51 (2008) 648–656. E. Warensjo, U. Riserus, I.B. Gustafsson, R. Mohsen, T. Cederholm, B. Vessby, Effects of saturated and unsaturated fatty acids on estimated desaturase activities during a controlled dietary intervention, Nutr. Metab. Cardiovasc. Dis. 18 (2008) 683–690. M.F. Chong, L. Hodson, A.S. Bickerton, R. Roberts, M. Neville, F. Karpe, et al., Parallel activation of de novo lipogenesis and stearoyl-CoA desaturase activity after 3 d of high-carbohydrate feeding, Am. J. Clin. Nutr. 87 (2008) 817–823. H. Petersson, S. Basu, T. Cederholm, U. Riserus, Serum fatty acid composition and indices of stearoyl-CoA desaturase activity are associated with systemic inflammation: longitudinal analyses in middle-aged men, Br. J. Nutr. 99 (2008) 1186–1189. L. Wang, A.R. Folsom, Z.J. Zheng, J.S. Pankow, J.H. Eckfeldt, Plasma fatty acid composition and incidence of diabetes in middle-aged adults: the Atherosclerosis Risk in Communities (ARIC) Study, Am. J. Clin. Nutr. 78 (2003) 91– 98. U. Riserus, W.C. Willett, F.B. Hu, Dietary fats and prevention of type 2 diabetes, Prog. Lipid Res. 48 (2009) 44– 51. B. Desvergne, L. Michalik, W. Wahli, Transcriptional regulation of metabolism, Physiol. Rev. 86 (2006) 465–514. A.D. Attie, R.M. Krauss, M.P. Gray-Keller, A. Brownlie, M. Miyazaki, J.J. Kastelein, et al., Relationship between stearoyl-CoA desaturase activity and plasma triglycerides in human and mouse hypertriglyceridemia, J. Lipid Res. 43 (2002) 1899–1907. G. Liu, J.K. Lynch, J. Freeman, B. Liu, Z. Xin, H. Zhao, et al., Discovery of potent, selective, orally bioavailable stearoylCoA desaturase 1 inhibitors, J. Med. Chem. 50 (2007) 3086– 3100.