Combined diet and exercise intervention reverses the metabolic syndrome in middle-aged males: results from the oslo diet and exercise study

Scand J Med Sci Sports 2007: 17: 687-695 Printed in Singapore. All rights reserved Journal compilation & 2007 Blackwell Munksgaard Combined diet and exercise intervention reverses the metabolicsyndrome in middle-aged males: results from the Oslo Diet andExercise Study S. A. Anderssen1, S. Carroll2, P. Urdal3, I. Holme1,4 1Department of Sports Medicine, Norwegian School of Sports Sciences, Oslo, Norway, 2Department of Sport, Health & ExerciseScience, Faculty of Applied Science & Technology, The University of Hull, Hull, UK, 3Department of Clinical Chemistry, UllevaalUniversity Hospital, Oslo, Norway, 4Centre of Preventive Medicine, Department of Preventive Cardiology, Ullevaal UniversityHospital, Oslo, NorwayCorresponding author: Sigmund A. Anderssen, Department of Sports Medicine, The Norwegian School of Sport Sciences,Sognsveien 220, Mail box 4014, Ulleva˚l Stadion; 0806 Oslo, Norway. Tel: 47-23262301; Fax: 47-22234220, E-mail:Sigmund.Anderssen@nih.no We examined the single and combined effects of a 1-year syndrome after 1-year intervention. In the diet-only group, diet and exercise intervention on the International Diabetes 22 participants (64.7%) (P 5 0.023 vs control) and in the Federation (IDF) metabolic syndrome among middle-aged exercise-only group 26 participants (76.5%) (P 5 0.23 vs males. The study was a randomized, controlled, 2 Â 2 control) had the metabolic syndrome following the interven- factorial intervention study. Participants included 137 men tion. Utilizing the factorial design, both dietary and exercise with metabolic syndrome according to the IDF criteria aged intervention had significant effects (Po0.005) on the reso- 40-49 years randomly allocated to four intervention groups: lution of the metabolic syndrome. Both exercise and dietary diet alone (n 5 34), exercise alone (n 5 34), the combination intervention reduced metabolic syndrome prevalence com- of the diet and exercise intervention (n 5 43) or control pared with control after 1 year of intervention. However, the (n 5 26). The main outcome measure was metabolic syn- combined diet and exercise intervention was significantly drome as defined by IDF criteria (2005). In the combined more effective than diet or exercise alone in the treatment of (Po0.0001 as compared with control) had the metabolic The metabolic syndrome has been a variably defined diabetes and reinforcing lifestyle modification within clustering of several cardiovascular disease (CVD) a multiple risk factor therapeutic strategy (Grundy, risk factors (Alberti & Zimmet, 1998; Balkau et al., 2006b). Indeed, Grundy (2006a) restates that the 2002); National Cholesterol Education Program primary reason why the NCEP ATP III introduced Adult Treatment Panel III (NCEP ATP III, 2001).
the metabolic syndrome into its clinical guidelines Insulin resistance, impaired glucose regulation, hy- was to emphasize the importance of lifestyle therapy pertension, obesity, and dyslipidemia are core meta- in clinical practice (Grundy, 2006a). Healthy lifestyle bolic syndrome components, independent of the promotion was also proposed as the ''primary inter- parent organization definition (Fulop et al., 2006).
vention'' and ''best first-line treatment'' within the Clustering of metabolic risk factors in individuals has IDF guidelines (Alberti et al., 2005; Grundy et al., been associated with increased risk of development 2005; Alberti et al., 2006). Moderate calorie restric- of diabetes (Laaksonen et al., 2002a; Sattar et al., tion and changes in physical activity and dietary 2003), CVD, and premature death (Isomaa et al., composition have been emphasized as primary man- 2001; Lakka et al., 2002; Sattar et al., 2003). A consensus worldwide clinically useful definition of The results from the Diabetes Prevention Program the metabolic syndrome has recently been recom- Research Group (DPPR) and Finnish Diabetes Pre- mended by the International Diabetes Federation vention studies have shown the marked clinical benefits of intensive lifestyle intervention on both Metabolic syndrome diagnosis has been proposed the resolution of the metabolic syndrome (Orchard as being clinically useful to identify patients who are et al., 2005) and risk of diabetes and among indivi- at a higher lifetime risk of both CVD and type 2 duals with glucose intolerance (Tuomilehto et al., 2001; Knowler et al., 2002). In these studies, lifestyle intervention involved concurrent changes in dietary According to the worldwide definition by IDF (Alberti et al., and physical activity habits, with one study particu- 2005), men have the metabolic syndrome if the following larly emphasizing weight loss (Hamman et al., 2006).
However, the studies did not randomly assign differ-  Central obesity: waist circumference 494 cm for Eur- ent components of the lifestyle interventions. Other studies have examined the specific effects of a Med- Plus any two of the following four components: iterranean-style diet alone (Azadbakht et al., 2005) orwithin a combined lifestyle modification approach  Raised triglyceride level: ! 1.7 mmol/L or treatment for (Esposito et al., 2004) on the metabolic syndrome.
Additional investigations have focused solely on the  Reduced HDL cholesterol: o1.03 mmol/L or treatment for effects of structured exercise training (Katzmarzyk et al., 2003; Stewart et al., 2005). To our knowledge,  Raised blood pressure: systolic BP ! 130 or diastolic BP ! 85 mmHg or antihypertensive medication; or however, there is a lack of controlled interven-tion trials that have been published concerning the  Raised fasting plasma glucose ( ! 5.6 mmol/L) or pre- efficacy of the single and combined effects of dietand exercise for treating the metabolic syndrome.
The Oslo Diet and Exercise Study (ODES) wasoriginally designed to study the single and combined effects of diet and exercise upon a broad spectrum of Dietary counseling was given together with the spouse at the CVD risk among persons with clustering of CVD start, and then to the participants alone after 3 and 9 months.
The advice was individualized and adapted according to each risk factors (Anderssen et al., 1995). The aim of this person's dietary history and risk profile (estimated from total paper was to examine the single and combined effects cholesterol, HDL, triglycerides, blood pressure, and body of a 1-year diet and exercise intervention within weight). The intervention focused primarily on energy restric- ODES participants meeting the IDF metabolic syn- tion in those who were overweight. Fish and fish products, and reduced intake of saturated fat and cholesterol, were recom-mended to all participants, but especially to those whoseelevated total cholesterol was the more important componentof the risk profile. In order to assess dietary compliance, eachparticipant responded to a 180-item food frequency question- naire (Solvoll et al., 1993). Smoking habits were recorded by a questionnaire as well as estimated through serum thiocyanateconcentration.
Data were retrieved from ODES, a randomized 2 Â 2 factorial The exercise program entailed supervised endurance-based trial of 1-year duration for each participant (The ODES exercise, such as aerobics, circuit training, and fast walking/ investigators, 1993). The participants were recruited from a jogging, three times per week. The duration of each workout continuous ongoing screening examination for all 40-year-old was 60 min. The intensity of the training was 60-80% of the persons in Oslo in the period 1990-1991. The ODES inclusion participant's individual peak heart rate as measured by a tread- criteria were as follows: physically inactive (exercising at most mill test at baseline. The exercise group and combined diet and once per week), body mass index (BMI) 424 (kg/m2), exercise group intermingled during supervised training sessions.
diastolic blood pressure 86-99 mmHg, total serum cholesterol The attendance of each workout was recorded, as was additional 5.2-7.7 mmol/L, HDL o1.20 mmol/L and fasting serum tri- physical activity performed by some participants. A Polar glycerides 41.4 mmol/L. Those with CVD, diabetes or other Sportstester heart rate recorder (Polar Electro OY, Kempele, diseases, or those using drugs that could interfere with the test Finland) was used to measure training intensity. Participants in results were excluded. A detailed description of the recruit- the control group were told not to change their lifestyle during ment of the 209 participants (21 females) within the ODES has the trial, but as all the other participants, they were advised been reported previously (The ODES investigators, 1993). The against smoking. At randomization, the control group partici- data for the present study are based on a retrospective pants were told that after the 1-year trial period, they would be examination of 188 men who were randomly allocated into offered dietary advice and supervised physical training.
one of four intervention groups: diet alone (n 5 45), exercisealone (n 5 48), combinations of diet and exercise intervention(n 5 58) or control (n 5 37). A pre-randomized sequence of the four possible intervention groups was prepared using simplerandomization without blocking. There was a 13% probabil- Each participant was examined for CVD risk factors, dietary ity of obtaining this distribution or a more extreme one, given habits, and maximal oxygen uptake before and after the 1-year equal allocation probabilities to the four groups. Of the 188 of intervention. Blood samples were collected after a 10-min men who met the original ODES study inclusion criteria, 137 recumbence between 08.00 and 10.00 hours after an overnight satisfied the IDF criteria for the metabolic syndrome. These fast and abstinence from smoking. Furthermore, the partici- men had been randomly allocated to the intervention groups pants were told to abstain from vigorous exercise for 4 days as follows: diet alone (n 5 34), exercise alone (n 5 34), combi- before blood sampling. The components were analyzed batch nations of diet and exercise intervention (n 5 43) or control wise from frozen samples at the end of the trial. Insulin was (n 5 26). A written declaration of informed consent was signed measured by a radioimmunoassay (Linco Research, St before the randomization took place. The study was approved Charles, Missouri, USA) before and 60 min after a standard oral glucose load of 75 g. Triglycerides, total cholesterol, and HDL were analyzed using enzymatic methods (Anderssen Table 1. Baseline values (SD) in all participants classified as having the metabolic syndrome according to the IDF criteria (n 5 137) Blood pressure was measured in the supine position after a 10-min rest in a quiet room at a room temperature of 22 1C.
Three recordings were made at 1-min intervals with automaticequipment (Vita-Stat blood pressure monitor; VitaStat Med- ical Services Inc., Bellevue, Washington, USA), and the mean of the last two measurements was used for statistical analysis.
Participants wore only underclothes when body weight, height, and waist circumference measurements were per- formed. Body weight was measured by using a Lindel balance scale (Samhald, Klippan Sweden). Waist circumference was measured while standing at the umbilical level. Two record- ings were made, and the mean of the two was used in the analysis. BMI was calculated as weight in kilograms dividedby the square of height in meters (kg/m2).
IDF, International Diabetes Federation; BMI, body mass index.
Cardio-respiratory fitness was measured during a maximal exercise test on a treadmill according to a modified Balkeprotocol (Balke, 1954) after a 15-min warmup. The test wasended when the participants were close to exhaustion accord- The mean total energy intake (based on the food ing to the Borg scale (Borg, 1970) (above 18) and/or when the frequency questionnaire) was decreased in the diet increase in oxygen uptake with increasing workload was group and in the combined diet1exercise group leveling off with a respiratory exchange ratio above 1.05.
Expired air was analyzed using an MMC Horizon System compared with the control and the exercise groups (SensorMedics, Yorba Linda, California, USA).
(Table 2). Also, the composition of the diet changedin those who received dietary counseling, with asignificant decrease in total fat intake in conjunction with increased polyunsaturated/saturated fatty acidsratio compared with control and exercise-only The response to intervention was measured as the differencebetween the starting and the final value for all variables (year groups. No significant changes in any of the mea- 1-year 0). The results were analyzed according to the inten- sured dietary variables took place in the control or tion-to-treat principle. Testing of, and confidence intervals the exercise-only group. The combined group did not (CIs), for between-intervention group contrasts for continu- adhere better to the dietary advice than the diet-alone ous variables were performed by Student's t-method. Differ- group, nor did this group adhere better to the ences in the prevalence of metabolic syndrome at 1-year weretested using w2 tests. Odds ratios for the relationship between exercise intervention than the exercise-alone group.
1-year changes in lifestyle-related variables (cardio-respiratory There were no statistically significant differences in fitness and selected dietary variables) and resolution of the smoking habits between either of the groups assessed metabolic syndrome were calculated using logistic regression.
by questionnaire or as estimated from serum thio- Interaction between diet and exercise was tested by including terms for diet, exercise and the combined cross-product termbetween the two in the model. All analyses were conducted Figure 1 shows the percentage of individuals not using SPSS statistical software (version 13.0, SPSS Inc., having the metabolic syndrome after the 1-year intervention period. There was a significant differ-ence between intervention groups with respect to theprevalence of metabolic syndrome at 1-year follow- up (Po0.0001). The diet1exercise group had aprevalence of 32.6% as compared with 88.5% in Table 1 displays the baseline data of the 137 parti- the control group (Po0.0001). Also, the diet-only cipants classified as having the metabolic syndrome.
group with a prevalence of 64.7% was reduced vs There was no difference between the intervention control (P 5 0.037). In contrast, the exercise-only groups at baseline in any of the variables.
group with a prevalence of 76.5% did not differ Table 2 gives the adherence to the lifestyle pro- significantly from control (P 5 0.23). The combined gram. The average exercise adherence was 61.3% in diet and exercise interventions were significantly the exercise and 64.7% in the combined group. This more effective in the treatment of metabolic syn- corresponds to an average of 110 and 116 min of drome compared with diet (Po0.006) and exercise exercise per week throughout the year, respectively.
alone (Po0.001). As no interaction (P 5 0.59) be- Cardio-respiratory fitness increased significantly in tween interventions was found for any variable, the the exercise groups compared with the control and factorial design was used to compare the effect of diet diet-only group. The net change in percent vs control and exercise separately. In the comparison of diet vs was 20.4% in the combined group and 15.0% in the no diet and exercise vs no exercise, the effect on exercise-only group. The diet-only and control group metabolic syndrome regression was statistically did not change their physical activity habits.
highly significant (Po0.005; Fig. 2).
Table 2. Baseline values of and changes in physical activity, cardiorespiratory fitness, body weight, dietary specific variables, and serum thiocyanate after1 year [year 1-year 0, mean (SE)] p/s, polyunsaturated/saturated.
aPo0.05 compared with control.
bPo0.05 compared with diet.
cPo0.05 compared with the exercise group.
following adjustment for changes in body weight[odds ratio 1.18 95% CI 5 1.04-1.35; P 5 0.011].
Changes in different dietary factors were also sig-nificantly associated with the reversal of the meta- bolic syndrome (Table 3), but were attenuated when adjusted for changes in body weight.
In this study, we examined the single and combinedeffects of a 1-year diet and physical activity interven- Fig. 1. Percentage reduction of participants not having the tion on the treatment of the metabolic syndrome.
metabolic syndrome after a 1-year intervention (Po0001 Our findings extend existing data by demonstrating diet1exercise as compared with control; P 5 0.023 diet onlyas compared with control; P 5 0.23 exercise only as com- in a randomized-controlled trial that lifestyle changes pared with controls.) The combined diet and exercise group in diet and physical activity habits have a substantial compared with diet only (Po0.006) and exercise only effect on the reversal of metabolic syndrome in men as defined by IDF. The two intervention modalitiesseemed to have direct additive effects on the resolu- Figure 3 gives the 1-year net changes in percent of tion of the metabolic syndrome. Furthermore, subjects in each intervention group below the cut changes in body weight and cardio-respiratory fitness point with respect to each variable in the IDF appeared to be important factors mediating the definition. All participants were above the waist change in metabolic syndrome components.
circumference criterion of 94 cm at baseline. For In observational studies, it has been shown that each variable, with the exception of systolic blood physical activity (Laaksonen et al., 2002b) or physi- pressure, the combined intervention group exhibited cal fitness (LaMonte et al., 2005) is inversely asso- the largest change. Also, in the combined group, all ciated with the incidence of metabolic syndrome variables included in the metabolic syndrome, except using either WHO or ATP III definitions. There fasting glucose, changed significantly compared with are, however, few studies in the literature that have reported on the single or combined effect of diet and Logistic regression analyses (Table 3) examining physical activity on the prevention of metabolic the relationship between changes in cardio-respira- syndrome. Katzmarzyk et al. (2003) found that tory fitness and reversal of the metabolic syndrome 30.5% of the participants were no longer classified gave a significant odds ratio of 1.32. Thus, the as having the metabolic syndrome (using the ATP III likelihood of reversal of metabolic syndrome com- criteria) following 20 weeks of aerobic exercise inter- pared with having the metabolic syndrome is 32% vention. However, this study is limited by the fact for a 1 U improvement of cardio-respiratory fitness that there was no control group and therefore the (mL/kg/min). This relationship remained significant authors could not control for the regression toward Fig. 2. Percentage reduction of participants not having the metabolic syndrome after a 1-year intervention using the 2 Â 2factorial design. The left panel displays comparison physical activity vs no physical activity (Po0.005). The right panel displayscomparison diet vs no diet (Po0.005).
Fig. 3. One year net changes in the prevalence of separate IDF metabolic syndrome components among participants bytreatment group. WC, waist circumference; TG, triglycerides; SBP, systolic blood pressure; DBP, diastolic blood pressure;Gluc, glucose.
Table 3. OR and 95% CI for the relationship between 1-year changes in cardio-respiratory fitness and selected dietary variables and reversal of themetabolic syndrome OR (95% CI) adjusted forchanges in body weight The OR corresponds to a 1 U change in cardio-respiratory fitness or dietary variables, except the p/s fatty acids ratio, where the OR corresponds to a 0.1 Uchange.
Adjusted and unadjusted for changes in body weight.
OR, odds ratios; CI, confidence interval; p/s, polyunsaturated/saturated.
the mean. Stewart et al. (2005) evaluated the effects and eight controls (15.1%) no longer exhibited the of 6 months of exercise training (aerobic and resis- tance exercise) in older men and women (55-75 Both short- (Roberts et al., 2006) and longer-term years) with untreated mild hypertension (Stewart (Muzio et al., 2005) uncontrolled studies have shown et al., 2005). At baseline, 42.3% of participants had reversal of the metabolic syndrome among obese metabolic syndrome by ATP III criteria. Following 6 adults with combined diet and physical activity months of exercise training, nine exercisers (17.7%) interventions, but differ regarding the independent effect of weight reduction. Moreover, Esposito III metabolic syndrome compared with placebo. In et al. (2004) randomized 180 patients (99 men and this study, 18% of the placebo group and 38% of the 81 women) with the metabolic syndrome (as defined lifestyle intervention group no longer had the syn- by ATP III) to a Mediterranean-style dietary inter- drome after 3 years (Orchard et al., 2005). The effect vention group, or to a control group. Following 2 of lifestyle on the reduction of the overall prevalence years of intervention, body weight decreased signifi- of the metabolic syndrome was reported to be most cantly more in patients in the intervention group strongly related to a reduction in the proportion of ( À 4.0 kg) than in those in the control group participants exceeding waist circumference and ( À 1.2 kg). The level of physical activity increased blood pressure thresholds and not, by comparison, comparably in both groups. Following completion of the criteria for lipid abnormalities of triglycerides the study, 44% of participants in the intervention group continued to exhibit the metabolic syndrome Furthermore, randomized-controlled intervention compared with 87% in the control. The dietary studies have shown that a shift to a healthy diet (with changes were also associated with improvement of about 30% energy derived from fat, and rich in fruit endothelial function and a significant reduction of and vegetables and wholegrain cereal products), in markers of systemic vascular inflammation. The fact combination with regular physical activity, can re- that participants in both groups received guidance on duce the risk of diabetes by almost 60% in indivi- increasing their level of physical activity makes it duals with impaired glucose tolerance (Pan et al., difficult to separate the effects of different lifestyle 1997; Tuomilehto et al., 2001; Knowler et al., 2002).
changes. The prevalence of metabolic syndrome has Interestingly, a subsequent analysis of the DPPRG also been shown to decrease significantly following study has identified a clustering of behavioral (diet- the DASH dietary intervention compared with ary and activity) changes within the intervention weight reduction and control diets (Azadbakht group, with participants adherent to one aspect of et al., 2005). In that study, the prevalence of the the lifestyle changes more likely to adhere to other metabolic syndrome after 6 months was 81% with a aspects (The Diabetes Prevention Program Research low-calorie diet, 65% with the DASH diet and Group, 2004). Our study adds to the knowledge in this field showing that such combined lifestyle It has previously been shown that increases in changes are also more effective in treating persons physical activity or changes in dietary habits improve single risk factors included in the metabolic syn- In order to evaluate which lifestyle variables con- drome and potential underlying etiological factors, tributed statistically to the reversal of the metabolic such as abdominal adiposity and insulin resistance syndrome, we calculated the odds ratios of metabolic (Ross et al., 2000; Janssen et al., 2002). The dietary syndrome according to the degree of changes in and exercise interventions in the present study ap- lifestyle variables considered as important. Changes peared to have independent and additive effects in in cardio-respiratory fitness were shown to be a resolving several metabolic syndrome components.
highly significant mediator even after adjustment Other randomized studies conducted among seden- for changes in body weight. Further adjustment of tary, overweight/obese non-diabetic adults have the analyses for changes in body weight may be an shown independent and occasionally additive effects over-correction as the fitness measure is normally of dietary and exercise changes on components of the expressed relative to body weight. Our finding that metabolic syndrome. Dietary modification (including cardio-respiratory fitness changes were strongly con- energy restriction) and vigorous exercise most often nected to metabolic syndrome is in line with observa- have additive effects on fasting and insulin response tional studies (LaMonte et al., 2005). Weight loss was to OGTT (Anderssen et al., 1996; Dengel et al., 1998; the dominant predictor of reduced diabetes incidence Watkins et al., 2003) and blood pressure (Cox et al., in the intervention arm of the DPPRG study, ad- 1996). The addition of exercise to a hypo-caloric, justed for changes in diet and physical activity reduced-fat diet improves high-density lipoprotein- (Hamman et al., 2006). In the present study, essential cholesterol and triglycerides in men and women with dietary variables did not have any independent im- relatively normal baseline lipoprotein-lipid levels pact when adjusted for changes in body weight. In (Stefanick, 1999), with less consistent effects in obese, contrast to these findings, post-hoc analysis of a hypertensives with metabolic syndrome characteris- randomized study showed that the relative reduction tics (Watkins et al., 2003) and other dyslipidemic in the metabolic syndrome associated with the states (Wood et al., 1991; Stefanick et al., 1998).
DASH diet was not due to weight loss alone (Azad- Within a secondary analysis of the Diabetes Pre- bakht et al., 2005). Dietary fat, saturated fat, and vention Program Research Group (DPPRG) study, fiber intake were also shown to be significant pre- lifestyle intervention showed a significant effect on dictors of progression to type 2 diabetes in the the cumulative incidence of the resolution of the ATP Finnish Diabetes Prevention Study participants, even after adjustment for intervention assignment, circumference. Whether dietary modification to- weight and weight change, physical activity and gether with regular exercise will reduce CVD out- baseline intake of the nutrient being investigated comes among persons with the metabolic syndrome has not been adequately tested in controlled clinical We have shown that 1-year changes in diet and trials. Nonetheless, epidemiological data are suppor- structured supervised exercise have a profound effect tive (Katzmarzyk et al., 2004), and the favorable on the prevalence of the metabolic syndrome. Nota- effects on the metabolic syndrome provide strong bly, in the combined diet and exercise intervention support and justification for their recommendation group, the net reduction compared with control was an impressive 56.0%. The beneficial effect of lifestyleintervention on all metabolic syndrome components(except for fasting glucose level) is particularly en- couraging and provides important evidence of the It is estimated that up to 30% of the adult population value of a combined lifestyle approach to persons in the Westernized society has several features of the with multiple metabolic risk factors.
metabolic syndrome, and increased long-term risk of There are proponents (Grundy, 2006c) and skep- CVD and diabetes (Isomaa et al., 2001; Lakka et al., tics (Kahn et al., 2005) concerning the metabolic 2002; Sattar et al., 2003; Wannamethee et al., 2005; syndrome and its diagnostic utility. Furthermore, Wilson et al., 2005; Sundstrom et al., 2006). Preva- debate continues concerning the underlying patho- lence increases with increasing age and in overweight genesis and heterogeneity in the clinical expression of and obese persons. Dietary modification and in- the syndrome. Nevertheless, it is our opinion that creased exercise in combination has a profound effect diagnosing the IDF metabolic syndrome is a clini- on the reversal of metabolic syndrome. Moderate cally simple, but meaningful method of identifying improvements in cardio-respiratory fitness and persons at a higher lifetime risk for both diabetes and weight loss seemed to be key intermediatory factors CVD. A high lifetime risk consistently indicates the in the success of the intervention. Clinicians and their need for lifestyle modification and may minimize the associates should be more anxious to advise their need for poly-pharmacy to treat the metabolic syn- patients to increase their physical activity and im- prove their dietary habits in the treatment of meta- Additional studies are needed to verify our lifestyle intervention findings in older and younger subjectsand females, given the age-specific and sex differences Key words: physical activity, nutrition, cardio-respira- in the prevalence of the metabolic syndrome and separate components (Orchard et al., 2005). Central(abdominal) obesity has been placed in pivotal posi-tion within the IDF definition. It should be notedthat the alternative WHO and ATP III metabolic syndrome criteria have different thresholds for ab- This study was supported by a grant from the Research dominal obesity and HDL level for each sex. The Council of Norway, The Norwegian Council of Cardiovascu- IDF (2005) used the same approach by maintaining lar Diseases and the Department of Sports Medicine, Norwe- ATP III thresholds for all measures except waist diagnosis and classification of diabetes A, Azizi T, Azizi F. Beneficial effects of randomized intervention trial; effect on ''atherothrombogenic syndrome''. Oslo Vera¨nderung infolge Arbeitsermu¨dung.
Arbeitsphysiologie 1954: 15: 311-323.
the Finnish Diabetes Prevention Study.
Janssen I, Fortier A, Hudson R, Ross R.
Effects of an energy-restrictive diet with Kahn R, Buse J, Ferrannini E, Stern M.
oral glucose tolerance in older men.
Am J Hypertens 1998: 11: 1405-1412.
Cardiorespiratory fitness attenuates the S, Sindhu RK, Vaziri ND, Barnard RJ.
Am Coll Cardiol 2006c: 47: 1093-1100.
stroke, and type 2 diabetes mellitus.

Source: http://www.ostrc.no/upload/Nyheter/dokumenter/Anderssen_2007_SJMSS_Combined%20diet%20and%20exercise%20intervention%20reverses%20teh%20metabolic%20syndromein%20middle-aged%20males.pdf