Hostname: page-component-5d59c44645-klj7v Total loading time: 0 Render date: 2024-02-23T07:35:15.579Z Has data issue: false hasContentIssue false

Impact of high-protein diets with either moderate or low carbohydrate on weight loss, body composition, blood pressure and glucose tolerance in rats

Published online by Cambridge University Press:  01 June 2007

Gerald E. Lobley*
Affiliation:
Obesity and Metabolic Health Division, Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, UK
David M. Bremner
Affiliation:
Obesity and Metabolic Health Division, Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, UK
Grietje Holtrop
Affiliation:
Biomathematics and Statistics Scotland (BioSS), Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, UK
Alexandra M. Johnstone
Affiliation:
Obesity and Metabolic Health Division, Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, UK
Christopher Maloney
Affiliation:
Obesity and Metabolic Health Division, Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, UK
*
*Corresponding author: Dr G. E. Lobley, fax +44 1224 716629, email g.lobley@rowett.ac.uk
Rights & Permissions [Opens in a new window]

Abstract

One approach to achieve weight loss and decrease both obesity and associated morbidities involves high-protein, low-carbohydrate (HPLC) diets. This study compares the impact on metabolic health of HPLC and high-protein, medium-carbohydrate (HPMC) diets offered to diet-induced obese (DIO) rats. Weanling male rats were fed either a 37 % fat diet (n 48) or stock pellets (n 12) for 22 weeks. Rats fed the 37 % fat diet accumulated more body fat (26·6 versus 14·8 % body weight, P < 0·001) compared with those on stock diet. The DIO rats had higher systolic blood pressure (+6·6 mmHg, P = 0·002), fasting insulin (+63 % P = 0·006) and areas under the glucose (+21 %, P < 0·001) and insulin (+81 %, P < 0·001) curves following an oral glucose tolerance test. DIO rats were then separated into four groups and offered for 8 weeks either: (1) the 37 % fat diet; (2) an HPLC or (3) HPMC diet; or (4) fed the 37 % fat diet to the intake of the HPMC group. Rats offered the 37 % fat or HPLC diets gained while those on HPMC lost body fat. Blood pressure was not altered by the dietary switch. Both HPLC and HPMC rats had lowered fasting insulin (P = 0·027) and improved homeostatic assessment (HOMA; P = 0·011) that was not different from those of stock animals. These improvements occurred despite differences in fat gain, and indicate that both weight loss and macronutrient intake can impact favourably on obesity-associated morbidities.

Type
Full Papers
Copyright
Copyright © The Authors 2007

Obesity is associated with co-morbidities, including dyslipidaemia, hypertension, cardiovascular disorders, and insulin insensitivity leading to type 2 diabetes. Combinations of these co-morbidities, loosely grouped under the term metabolic syndrome (Kahn et al. Reference Kahn, Buse, Ferrannini and Stern2005), impact on longevity and quality of life. Therefore, the inexorable rise in obesity in more affluent countries (World Health Organization, 2000) has associated burgeoning costs of health care (Avenell et al. Reference Avenell, Broom, Brown, Poobalan, Aucott, Stearns, Smith, Jung, Campbell and Grant2004; Kuller, Reference Kuller2006), In consequence, a large number of behavioural, surgical and nutritional strategies have been proposed to either halt or reverse the obesogenic trend. One popular approach with the general public involves use of low-carbohydrate (ketogenic) diets as these appear to induce satiety at intakes below energy maintenance (Stadler et al. Reference Stadler, Burden, McMurry, Gerhard and Connor2003; Johnstone et al. Reference Johnstone, Murison, Bremner, Horgan and Lobley2006). Such diets have provoked fierce debate amongst nutritionists, however, as the low carbohydrate intake is partly substituted by fat and this may increase the dyslipidaemia often observed in the obese (Grundy, Reference Grundy2005) and contribute to hypertension, plaque formation and other CVD (Lorenzo et al. Reference Lorenzo, Williams, Hunt and Haffner2006; Truesdale et al. Reference Truesdale, Stevens, Lewis, Schreiner, Loria and Cai2006). Nonetheless, in human trials, the effects of such diets on blood lipid parameters have been either favourable or neutral (e.g. Foster et al. Reference Foster, Wyatt, Hill, McGuckin, Brill, Mohammed, Szapary, Rader, Edman and Klein2003; Wood et al. Reference Wood, Volek, Liu, Shachter, Contois and Fernandez2006). This is probably because either the enforced (Foster et al. Reference Foster, Wyatt, Hill, McGuckin, Brill, Mohammed, Szapary, Rader, Edman and Klein2003) or voluntary (Stadler et al. Reference Stadler, Burden, McMurry, Gerhard and Connor2003; Johnstone et al. Reference Johnstone, Murison, Bremner, Horgan and Lobley2006) reduced intake results in negative energy balance and oxidation of both dietary and stored fat.

This raises the question – how do such low-carbohydrate diets lead to reduced voluntary intake? Often such diets contain high amounts of protein, the macronutrient with the greatest impact on satiety (e.g. Weigle et al. Reference Weigle, Breen, Matthys, Callahan, Meeuws, Burden and Purnell2005). Alternatively, low carbohydrate supply may alter insulin status and sensitivity, with impacts on appetite (Hellstrom et al. Reference Hellstrom, Geliebter, Naslund, Schmidt, Yahav, Hashim and Yeomans2004; Wynne et al. Reference Wynne, Stanley, McGowan and Bloom2005). Finally, during ketosis, regions of the brain switch from glucose as the primary fuel to oxidation of ketone bodies (Hawkins et al. Reference Hawkins, Mans and Davis1986). These regions include those critically involved in regulation of appetite (Archer et al. Reference Archer, Rayner and Mercer2004), and thus alteration in their fuel use may influence intake. A second question relates to the reported beneficial effects on metabolic health of low-carbohydrate diets (Foster et al. Reference Foster, Wyatt, Hill, McGuckin, Brill, Mohammed, Szapary, Rader, Edman and Klein2003; Samaha et al. Reference Samaha, Iqbal, Seshadri, Chicano, Daily, McGrory, Williams, Williams, Gracely and Stern2003; Johnstone et al. Reference Johnstone, Murison, Bremner, Horgan and Lobley2006) – are these due to the associated weight loss or are the improvements linked to the pattern of macronutrient intake (Volek & Feinman, Reference Volek and Feinman2005)? For example, a recent report (Boden et al. Reference Boden, Sargrad, Homko, Mozzoli and Stein2005) showed that insulin sensitivity in the obese, as determined by homeostatic assessment (HOMA), improved within 2 weeks on a low-carbohydrate diet, before there were appreciable changes in body weight or fat mass.

While these questions may be partly answered by direct studies in obese humans, some of the more detailed information requires invasive approaches and use of appropriate animal models that mimic the human situation. A number of dietary-induced obese (DIO) rodent models have been developed to examine impacts on either blood pressure (e.g. Velkoska et al. Reference Velkoska, Cole and Morris2005) or insulin sensitivity (e.g. Madsen et al. Reference Madsen, Guerre-Millo and Flindt2002), but rarely have both morbidities been studied at the same time in rodents. Furthermore, technical constraints have limited the ability to monitor longitudinal changes in body composition in rats or mice, and thus to link such changes with metabolic responses. Finally, many studies of rodent models have described the effect of rate of body fat gain, but few have attempted to look at the loss of lipid stores from obese animals under different dietary conditions. Yet this is exactly the situation that faces many dieters' having become obese, how do they lose weight and what is the impact on overall metabolic health?

The current study describes a DIO rat model, based on offering a diet similar in macronutrient energy composition to the average UK human diet, and then quantification of the gain in body fat. The impact of the resultant obesity on blood pressure and aspects of the glucose insulin axis (HOMA and an oral glucose tolerance test (OGTT)) was then assessed. This was followed by offering two high-protein diets, one moderate and the other low in carbohydrate, with changes in body weight and body composition monitored over an 8-week period, with intermediate measures of blood pressure and glucose tolerance.

Materials and methods

Rats and diets

Sixty male Rowett strain Hooded Lister rats were weaned at 19 d of age onto CRM (P) (Special Diet Services, Witham, Essex, UK) stock cubes, with measured digestible and estimated metabolisable energy (ME) of 13·20 and 12·13 MJ/kg, partitioned as Atwater fuel energy, protein 22 %, fat 9 %, carbohydrate 69 %. At 4 weeks of age, the rats were randomly separated into two groups of similar average weight. One group (n 12) were maintained on stock while the others (n 48) were transferred to a semi-synthetic 37 % fat diet (Table 1), with measured digestible energy of 18·96 MJ/kg and estimated ME of 17·98 MJ/kg as fed, partitioned by Atwater coefficients as protein 16 %, fat 37 %, carbohydrate 47 %. While this is similar in energy composition to the average UK human diet, the maize oil used (MP Biochemicals Inc., Solon, OH, USA) contains less saturated fat (10–12 %) and more MUFA plus PUFA (39–42 and 43–47 %, respectively) than the average human fat intake. Animals were housed four to a cage for the next 18 weeks and then were housed singly for the remainder of the study (total 30 weeks). Throughout the study, individual rats were weighed twice weekly, with food intake per cage measured continuously throughout. All procedures were approved by the Animal Ethics Committee of the Rowett Research Institute under the auspices of the UK Animals (Scientific Procedures) Act 1986.

Table 1 Compositions, as g/kg as fed, of the various diets used in this study

HPLC, high-protein. low-carbohydrate; HPMC, high-protein, medium-carbohydrate; AF, as fed; DE, digestible energy; ME, metabolisable energy.

* Values in parentheses are for stock diet, CRM (P), provided by the manufacturer

Contains 10 g of AIN-93G vitamin mixture, 35 g of AIN-93 VX minerals mixture (both from MP Biomedicals Europe, Illkirch, France), 2 g of choline chloride and 3 g of potassium phosphate.

Based on assumed N in urine (digested N – retained N) × 40 kJ/gN.

Measurements

Body composition

Whole body composition (as total fat and lean) of conscious animals was quantified by MRI (EchoMRI, Houston, TX, USA) at intervals during the trial. The machine was originally calibrated by the manufacturer based on chicken breast (for lean) and canola oil (for fat). The lean calibration was checked with weights (from 75 to 600 g) of chicken breast analysed for dry matter plus lean and fat content (the latter approximately 2 %), by Dumas combustion and extraction with petroleum spirit, respectively, and appropriate corrections applied. The relationship against the manufacturer's lean calibration was 0·947x+6·0, R 2 = 0·999. The lipid calibration was tested against freshly dissected rat fat (from 5 to 375 g), also analysed for N, fat and dry matter. The relationship between measured and observed calibration (based on the canola oil calibration) was 0·959x+2·31, R 2 = 0·999. These calibration factors were then used to correct the MRI output to obtain body composition as fat and lean mass (g).

First measurements on the rats were made at 14 weeks, and from 18 weeks onwards the animals (six rats per day) were scanned at 2-weekly intervals until the end of the study.

Blood pressure

At 19 weeks, blood pressure was measured using the tail cuff procedure (Rees et al. Reference Rees, Hay and Cruickshank2006). Rats were left in a cage over a plate heated to 28°C for 10 min and lightly restrained manually during the measurement. Data were recorded using an MP35 BSL system (BioPac Systems, Inc, Goleta, CA, USA) with software BSL Pro 7. Systolic pressure was determined from the pressure at which a pulse was first detected. The procedure was repeated until three acceptable traces were obtained; this required the rat to be motionless. The rats were measured again on the following 2 d and the composite data analysed.

Oral glucose tolerance test

Starting at week 20, over a 2-week period, six rats were selected daily at random for the 2 h measurement of OGTT. On the previous day, body composition was quantified by MRI scan and then the rats were fasted overnight (18.00 h onwards). At 08.00 h the next morning, 0·25 ml of blood was taken from the tail tip into Eppendorf tubes containing 10 μl of heparin solution (15 IU/ml). Animals were then given a gavage of a 50 % (w/v) glucose solution equivalent to 2 g of glucose per kg body weight. Samples of blood (150 ml) were then taken from the tail tip at 7, 15, 30, 60, 90 and 120 min. Animals were then returned to their normal cages and allowed free access to food and water. Insulin was assayed using a Mercodia enzyme-linked immunosorbent assay (Mercodia AB, Uppsala, Sweden) and glucose was quantified by an enzyme-linked procedure based on a glucose oxidase/peroxidase method with 4-amino-antipyrine as dye (Maloney et al. Reference Maloney, Gosby, Phuyal, Denyer, Bryson and Caterson2003). Both analyses were quantified spectrophotometrically by plate reader. Areas under the curve (AUC) for both plasma insulin and glucose were calculated for both total (tAUC) and incremental (iAUC; Wolever & Jenkins, Reference Wolever and Jenkins1986, 1997) values. Various indices were used to assess glucose tolerance, insulin sensitivity or insulin resistance, and these were based on either the fasting glucose and insulin concentrations or the responses to the oral glucose load. These methods included fasting insulin resistance index, quantitative insulin sensitivity check index, simplified formula for the HOMA method, insulin sensitivity index, insulin sensitivity index of glycaemia; insulin sensitivity index, oral glucose insulin sensitivity; and insulin sensitivity index (composite), composite insulin sensitivity index for the hepatic and peripheral tissues. These are all described and discussed for rats in Tran et al. (Reference Tran, Gupta, Goh, Naigamwalla, Chia, Koohestani, Mehrotra, McKeown-Eyssen, Giacca and Bruce2003). Many gave similar conclusions and, for simplicity, therefore, HOMA, calculated as (fasting insulin × fasting glucose)/22·5, and iAUC and tAUC, was used as the main indices.

Diet switches for rats fed the 37 % fat diet

At week 22, the 48 rats on the 37 % fat diet were allocated to four matched groups (each n 12), based on percentage body fat as assessed by the MRI scans. Each group was then allocated to one of four dietary treatments (Table 1), either: (1) remaining on the 37 % fat diet; (2) a high-protein, low-carbohydrate (HPLC) diet; (3) a high-protein, medium-carbohydrate (HPMC) diet; or (4) the 37 % fat diet but with each animal fed the ME intake of the weekly group average of the HPLC rats (GF). In practice, within the first week of the dietary switch, it became apparent that the GF animals could not match the energy intake of those on HPLC. This was because of the higher energy density of the HPLC diet compared with either the 37 % fat or HPMC diets (Table 1). In consequence, after 7 d, the GF rats were then matched to the ME intake of the HPMC group. The remaining twelve animals (stock) continued to be fed CRM pellets. Because only six animals could be measured by the OGTT each day, the starts for these dietary regimes were staggered over a 2-week (2 × 5 d) period. For the first 8 d, animals were allocated on a random basis from each of the stock, 37 % fat, HPLC and HPMC groups (i.e. one rat from two groups and two rats from the other two groups). Within each group, the start dates for the rats were based on descending order of percentage body fat. The GF animals were started on days 9 and 10. Time on treatment was set from the day of allocation and so all main measurements (MRI scans and OGTT) were staggered over 2-week periods, with six animals measured per day, and with the MRI scans performed the day before each OGTT. Animals remained on these diet allocations between weeks 22 and 30.

Faeces were collected over a 4 d interval during weeks 26 and 27 from stock (n 6) and 37 % fat, HPLC and HPMC (all n 8) animals so that dry matter, energy (bomb calorimetry) and N (Dumas combustion) digestibilities could be calculated (Table 1). These were used to estimate energy intakes throughout the trial. At the end of week 27, blood pressure was again measured on three successive days.

At the start of week 28, rats were fasted overnight and an OGTT was performed as for the pre-diet allocation procedure. At week 30 (i.e. 8 weeks after diet switch), animals were anaesthetised using isofluorane inhalation at 10.00 h, decapitated and trunk blood taken into two tubes containing either EDTA or heparin. Plasma samples (EDTA) were assayed for insulin and leptin using a Lincoplex dual assay kit (Linco Research, Inc., St Charles, MO, USA) measured on a Luminex 100 system (Luminex Corporation, Austin, TX, USA). Analyses were by enzyme-linked assays for plasma glucose, cholesterol, triglycerides (Thermo Scientific kits 981304, 981813, 981301, respectively; Labmedics Ltd, Salford, Manchester, UK), NEFA (kit 999–75 406; Wako Chemicals GmbH, Neuss, Germany) and 3-hydroxybutyrate (based on Li et al. Reference Li, Lee, MacGillivray, Schaefer and Siegel1980) using a Konelab 30 discrete clinical analyser (Thermo Electron Corp, Vantaa, Finland).

Statistical analysis

Data from the first stage of the study (stock versus 37 % fat diet) were analysed by a two-sample t test. For OGTT data, two-way ANOVA was conducted, with effects of diet, time and their interactions as fixed effects. Results from the five dietary groups during the last 8 weeks of the study were analysed by one-way ANOVA, as were differences between the four high-fat groups (37 % fat, HPMC, HPLC and GF). The status of the animals at the start of this second phase (e.g. fat content, blood pressures or OGTT) was also included as an appropriate covariate to allow for any differences established in the pre-diet switch period (weeks 1–22). For a number of variables, the covariate regression was significant but, despite this, inclusion in the analysis rarely altered the main findings. For simplicity, therefore, only the results of the one-way ANOVA are reported unless inclusion of a covariate altered the findings. Any significant differences in main effects were further assessed on group means by a post hoc t test. All analyses were performed with Genstat 8th Edition, Release 8·1 (VSN International Ltd., Hemel Hempstead, Herts, UK).

Results

Chow versus 37 % fat dietary treatments

Food intake and weight gain

Body weights and body weight gains were either greater or equal for stock compared with the 37 % fat diet for the first 6 weeks, but from 8 weeks onwards the latter rats were heavier (P < 0·05). By week 22 (diet switch), the 37 % fat diet animals weighed, on average, 80 g more than rats fed the stock diet (P < 0·001; Table 2). Average intakes over weeks 1–22 were 24·6 and 18·2 g/d for stock and 37 % fat diets (P < 0·001) with corresponding ME intakes of 298 and 327 kJ/d (P = 0·004), respectively.

Table 2 Body composition and blood pressure data from Hooded Lister rats (25 weeks old) fed for 20 weeks on either stock pellets (n 12) or a 37 % fat diet (15 % protein; 48 % carbohydrate; 37 % fat; n 48)

* Average data for days 2 and 3 of measurement, minimum of three measurements per day.

Body composition

By week 14, the stock and 37 % fat rats had mean weights of 505 and 549 g (P < 0·001); this included 70 and 145 g of fat (P < 0·001), respectively. Lean tissue was not significantly different (425 versus 417 g). At week 20, the time of the first OGTT, respective mean body weights had increased to 577 and 644 g (P < 0·001), fat to 86 and 172 g (P < 0·001) and lean tissue to 441 and 436 g (not significant; Table 2). Together, these accounted for 0·91–0·95 of live weight, with the remainder attributed to bone mass (which gives a low signal in the MRI scanner), glycogen and gut fill. Fat comprised 14·8 and 26·6 % of live weight (15·4 and 27·1 % of fasted weight) for stock and 37 % fat groups, with lean tissue as 76·5 and 67·7 % (79·5 and 68·9 % of overnight fasted weight).

Blood pressure

There was a significant day effect (P = 0·003) for systolic pressures, associated mainly with higher values for the first day; this may involve an animal ‘training’ effect. There were no differences between days 2 and 3, and these were then averaged to test for diet effects. Rats on the 37 % fat diet had higher systolic pressure (6·6 mmHg, P = 0·002; Table 2).

OGTT (Table 3 and Fig. 1)

Fasting glucose was not different between the two diets (Table 3). Within 60 min after the oral glucose dosing, plasma values had returned (P>0·05) to pre-dose concentrations for rats fed stock diet (Fig. 1a). In contrast, the post-dose values for the 37 % fat group remained higher (P < 0·001) than pre-dose concentrations throughout, and even at 120 min were still 24 % greater. From 30 to 120 min, the values for the 37 % fat group exceed those for the stock-fed animals by 24–31 % (P < 0·005). In consequence, both iAUC (P = 0·009) and tAUC (P < 0·001; Fig. 1a, Table 3) were greater for the 37 % fat group.

Table 3 Oral glucose tolerance test data from Hooded Lister rats (25 weeks old) fed for 20 weeks on either stock pellets (n 12) or a 37 % fat diet (15 % protein; 48 % carbohydrate; 37 % fat; n 48)

NS, not significant; HOMA, homeostatic assessment; iAUC, incremental area under the curve; tAUC, total area under the curve.

* Data also compared within the variable for effect of time (***P < 0·001).

Fig. 1 Temporal changes in plasma (a) glucose and (b) insulin in overnight fasted rats given an oral dose of glucose (2 g/kg body weight). Rats were offered either stock pellets (□) or a 37 % fat diet (■). Data are means with their standard error.

Fasting plasma insulin was 63 % higher (P = 0·006, Table 3, Fig. 1b) for rats fed the 37 % fat diet compared with those offered stock pellets, and this also resulted in higher HOMA values (P = 0·005). Rats on stock increased (P < 0·05) plasma insulin 15 and 30 min post-dose, but at all other times the values were not different from fasting concentrations. In contrast, for rats on 37 % fat, insulin concentrations remained above fasting from 15 min onwards (P < 0·001), and even at 120 min were still more than double the pre-dose values. Furthermore, in comparison with the stock group, insulin values for the 37 % fat rats were greater (P < 0·001) from 60 min onwards. Consequently, the AUC were greater for the 37 % fat group compared with stock-fed animals for both the incremental (+88 %; P = 0·007) and total (+80 %; P < 0·001) areas. The maximal insulin concentration (at 30 min) was greater for 37 % fat rats than those on stock; this indicates that the inability to lower glucose concentration to fasting concentrations when animals were fed the 37 % fat diet was not due to less pancreatic release of insulin.

The main parameters of the glucose–insulin axis correlated (P < 0·001) with body fat (either in absolute terms or as a percentage of body weight), e.g. HOMA (R = 0·56), glucose tAUC (R = 0·50), fasting insulin (R = 0·51), insulin iAUC (R = 0·42) and insulin tAUC (R = 0·54). Weaker relationships (P < 0·01) were observed between systolic blood pressure and body fat (R = 0·36) and between systolic blood pressure and fasting insulin (R = 0·42).

Responses of obese rats to dietary interventions

Food intake, weight gain, body composition and blood pressure (Table 4 and Fig. 2)

After the 8 weeks on the new diet allocations, all groups were heavier, except for rats on the HPMC diet who lost weight (average 25 g), and this was different (P < 0·001) from the other groups that originated from animals fed the 37 % fat diet. Rats on the HPLC diet ate less food (g/d, P < 0·001) than those on the 37 % fat and HPMC diets but, due to the greater energy density of this diet, they had the largest daily ME intake (kJ/d, P < 0·001). In contrast, the HPMC rats had the lowest ME intake. Protein intakes were more than doubled for the two high-protein groups compared with 37 % fat rats, while for the HPLC animals carbohydrate consumption was only approximately 16 %. Changes in body weight and fat gain were approximately linear during the 8 weeks following diet switch (Fig. 2). Over this period, most of the differences in body weight reflected changes in fat gain (groups 37 % fat, GF and HPLC) or fat loss (HPMC). The GF animals lost the most lean tissue, probably related to the restricted intake of the 16 % protein diet. While the ME intakes of the GF and HPMC groups were similar, energy retention was positive in the former and negative in the latter (11·0 versus − 15·4 kJ/d, respectively). This may reflect either differences in physical activity between the two groups or a higher thermic effect of high protein intake as observed in humans (Westerterp-Plantenga et al. Reference Westerterp-Plantenga, Rolland, Wilson and Westerterp1999). There were no changes in systolic pressure, although this remained lower (P = 0·02) for those rats offered stock compared with all other groups.

Table 4 Body weight, composition, food intake, macronutrient intake (protein, carbohydrate and fat), metabolisable energy intake and blood pressure in obese rats following 8 weeks on diets of either 37 % fat, HPLC, HPMC or the 37 % fat diet fed to the group intake mean of HPMC (GF)

HPLC, high-protein low-carbohydate; HPMC, high-protein, medium-carbohydrate; HF, high fat; BW, body weight; NS, not significant; FI, food intake; CHO, carbohydrate; MEI, metabolisable energy intake; BP, blood pressure.

* Analysis for all groups, including rats fed stock (sed refers to this analysis).

Analysed for four groups selected from rats fed on the original 37 % fat diet.

Data at the time (22 weeks) the rats fed from weaning on the 37 % fat diet were re-allocated to subsequent treatment diets.

§ Changes after 8 weeks on diets (difference between week 30 and week 22 data).

An additional group of animals was maintained on stock pellets. For all groups, n 12 except for stock and GF (n 11 in both cases).

Fig. 2 Body weight (g) and body fat (g) during the last 8 weeks of the study after diet switch. Rats remained on either ther 37 % fat diet (○) or stock pellets (□), or were switched to high-protein, low-carbohydrate (Δ), high protein-moderate carbohydrate (▲) or fed the 37 % fat diet to intake of the HPMC diet (●).

Glycaemia, insulinaemia and OGTT (Table 5)

At diet switch, animals originally fed the 37 % fat diet were allocated to the four matched groups based on percentage body fat and not on responses to the first OGTT. In practice, this selection introduced a bias between groups with lower initial glucose iAUC (197 versus 440 mm/2 h, P < 0·016) and tAUC (1003 versus 1242 mm/2 h) for rats maintained on the 37 % fat compared with the other three groups. Other parameters, including fasting glucose, all insulin parameters and HOMA, were not different between the original 37 % fat groups at diet switch. Nonetheless, all parameters were also tested with inclusion of a covariate to allow for any differences in starting values. Comparisons were made with results from rats fed stock both included and excluded; the latter allowed the impact of diet switch from 37 % fat to be analysed in more detail.

Table 5 Fasting glucose and insulin concentrations, HOMA calculation and OGTT data in obese rats following 6 weeks on diets of either 37 % fat, HPLC, HPMC or the 37 % fat diet fed to the group intake mean of HPMC (GF)

HOMA, homeostatic assessment; OGTT, oral glucose tolerance test; HPLC, high-protein low-carbohydate; HPMC, high-protein, medium-carbohydrate; HF, high fat; NS, not significant; iAUC, incremental area under the curve; tAUC, total area under the curve.

* Analysis for all groups (including rats fed stock); sed refers to this analysis.

Analysed for four groups selected from rats fed on the original 37 % fat diet.

An additional group of animals was maintained on stock pellets. For all groups, n 12 except for stock and GF (n 11 in both cases).

At 6 weeks post-diet switch (week 28), rats fed stock diet had lower fasting glucose (6·44 versus 7·10 mm, P = 0·002), iAUC (190 versus 288 mm/2 h, P = 0·006) and tAUC (960 versus 1137 mm/2 h, P < 0·001) for glucose than animals on the other diets, similar to the findings at week 20. In contrast, for insulin parameters, including fasting values, iAUC and tAUC, the stock animals were not significantly different from the other four groups combined, neither was HOMA.

These differences compared with findings at 20 weeks were related to changes within the four groups following the diet switch from 37 % fat. For example, fasting insulin concentrations were lower (P = 0·011) for the GF, HPLC and HPMC groups compared with 37 % fat. This difference remained even after covariate inclusion.

Furthermore, the GF, HPLC and HPMC groups were not different from rats that continued on the stock diet. In contrast, fasting glucose concentrations were similar for the 37 % fat, GF and HPMC diets and greater than for the HPLC diet (P < 0·05). HOMA (based on the product of insulin and glucose concentrations) reflected these responses, with greater values for rats maintained on 37 % fat diet compared with the GF, HPMC and HPLC groups (P ≤ 0·006).

Following the oral glucose dose, insulin iAUC and tAUC were lower for rats fed HPMC (P < 0·05) compared with those that continued on 37 % fat. This was due to higher insulin concentrations for the latter throughout the post-dose period, as reflected in the difference between the diets at 120 min (P < 0·05; Table 5). For glucose, again iAUC and tAUC were similar for 37 % fat, HPMC and HPLC, but the GF values were lower (P < 0·05) than for those that remained on the 37 % fat diet.

Several of the parameters that showed correlations with body fat (either absolute or as percentage body weight) prior to the diet switch were maintained (P < 0·01); these included HOMA (R = 0·34), glucose tAUC (R = 0·43), fasting insulin (R = 0·43), insulin iAUC (R = 0·49) and tAUC (R = 0·57). Again, a weaker relationship was maintained with systolic blood pressure (R = 0·37).

Terminal biochemical measurements (Table 6)

The terminal measurements were performed on non-fasted animals, except for the GF rats that were deprived of food for 9 h. Plasma leptin was lower (P < 0·001) in the stock-fed rats compared with the other groups, but there were no differences between plasma insulin. Differences in plasma glucose concentrations (P < 0·001) were due to the lower values for the fasted GF rats; the other groups of fed animals showed no differences. Plasma 3-hydroxybutyrate was greatest in the fasted GF rats, followed by those on the HPLC diet (both P < 0·001 compared with each other and the remaining groups). For lipid parameters, there were no differences in plasma cholesterol or NEFA. Stock animals had higher triacylglycerols (P < 0·001) than any of the other groups, with values for the 37 % fat rats also greater than those for the remaining groups, while the GF rats were the lowest (P < 0·05).

Table 6 Plasma concentrations of insulin, leptin, glucose, cholesterol, NEFA, triacylglycerols and 3-hydroxybutyrate after 8 weeks of diet switch

HPLC, high-protein low-carbohydate; HPMC, high-protein, medium-carbohydrate; HF, high fat; NS, not significant; 3-OHB, 3-hydroxybutyrate.

* Analysis for all groups (including rats fed stock); sed refers to this analysis.

Analysed for four groups selected from rats fed on the original 37 % fat diet.

These animals were not fasted, except for 9 h of food withdrawal for the GF rats. Animals were fed 37 % fat, ad libitum or pair fed (GF) to the HPMC group, HPLC or HPMC, with n 12 per group except for stock and GF (n 11 in both cases).

Discussion

The first objective of this study was to make rats obese through feeding a diet similar to the average macronutrient energy composition consumed by humans in the UK and to see if this resulted in aspects of metabolic syndrome, including raised systolic blood pressure and impaired insulin sensitivity. The second objective was to offer these obese rats diets high in protein, with either moderate or low carbohydrate, and assess impacts on food intake, body composition and parameters of metabolic syndrome. These results would be compared with published human data to assess the reliability of the rodent model for studies on underlying mechanisms in obese humans.

DIO rats

A large number of studies have shown that feeding rats or mice high-fat diets results in increased weight gain and adiposity (e.g. Levin et al. Reference Levin, Dunn-Meynell, Balkan and Keesey1997; Archer et al. Reference Archer, Rayner, Rozman, Klingenspor and Mercer2003; Alexander et al. Reference Alexander, Chang, Dourmashkin and Leibowitz2006). In the current study, body fat was approximately doubled from week 14 onwards and, in view of observations that early changes in fatness occur before differences in weight gain are apparent (Archer et al. Reference Archer, Rayner and Mercer2004), an obesogenic condition was probably prolonged over much of the current study. The slow development of obesity with the current protocol matches the situation in the human population.

Blood pressure

It has long been recognised that hypertension in both men and women correlates with obesity (Kannel et al. Reference Kannel, Brand, Skinner, Dawber and McNamara1967) and forms an important component of the health risks contained within the definition of metabolic syndrome (Kahn et al. Reference Kahn, Buse, Ferrannini and Stern2005). A number of rodent studies have also shown that obese animals have increased systolic blood pressure (e.g. Dobrian et al. Reference Dobrian, Davies, Prewitt and Lauterio2000; Wilde et al. Reference Wilde, Massey, Walker, Vollmer and Grekin2000; Yoshioka et al. Reference Yoshioka, Uemura, Tamaya, Tamagawa, Miura, Iguchi, Nakamura and Hotta2000; Velkoska et al. Reference Velkoska, Cole and Morris2005), although this has not always been observed (Beltowski et al. Reference Beltowski, Wojcicka and Jamroz-Wisniewska2006). Such changes have also correlated with the higher concentrations of plasma leptin and insulin (Velkoska et al. Reference Velkoska, Cole and Morris2005), but the correlation with fasting insulin was weak in the current study. Such comparisons may be confounded by inclusion of maize oil in the current diets, because this can lower blood pressure (Langley-Evans et al. Reference Langley-Evans, Clamp, Grimble and Jackson1996; Truett et al. Reference Truett, Borne, Monteiro and West1998) and may reduce the impact of obesity on hypertension.

OGTT and insulin sensitivity

Decreased insulin sensitivity is a characteristic feature of metabolic syndrome in humans, usually first detected as increased plasma glucose concentrations despite elevated insulin following an overnight fast. In the current study, the simpler approaches based on either fasting glucose and insulin concentrations (e.g. HOMA), or the OGTT were used rather than surgically invasive hyperinsulinaemic–euglycaemic clamps (Madsen et al. Reference Madsen, Guerre-Millo and Flindt2002; Tran et al. Reference Tran, Gupta, Goh, Naigamwalla, Chia, Koohestani, Mehrotra, McKeown-Eyssen, Giacca and Bruce2003). These simplified approaches are not perfect substitutes but, in rats, approximately 50 % of the variability (R 2) with glucose infusion rate during an insulin clamp could be accounted for by the insulin tAUC (taken as a measure of insulin resistance) or the tAUC for glucose/insulin (a marker of insulin sensitivity) during an OGTT (Tran et al. Reference Tran, Gupta, Goh, Naigamwalla, Chia, Koohestani, Mehrotra, McKeown-Eyssen, Giacca and Bruce2003). The HOMA approach has also been shown to correlate well with clamp data in humans (R = 0·62; Yeni-Komshian et al. Reference Yeni-Komshian, Carantoni, Abbasi and Reaven2000).

The current results support the myriad of observations that demonstrate obese rodents to have elevated fasting plasma insulin (e.g. Gao et al. Reference Gao, Ghibaudi, van Heek and Hwa2002; Madsen et al. Reference Madsen, Guerre-Millo and Flindt2002; Catalano et al. Reference Catalano, Bergman and Ader2005; Clegg et al. Reference Clegg, Benoit, Reed, Woods, Dunn-Meynell and Levin2005; Morris et al. Reference Morris, Velkoska and Cole2005) occasionally in combination with elevated fasting glucose (Molnar et al. Reference Molnar, Yu, Mzhavia, Pau, Chereshnev and Dansky2005; Thim et al. Reference Thim, Bentzon, Kristiansen, Simonsen, Andersen, Wassermann and Falk2006). Obese rats require lower glucose infusion to maintain euglycaemia during hyperinsulinaemic clamps (Chang et al. Reference Chang, Graham, Yakubu, Lin, Peters and Hill1990; Tran et al. Reference Tran, Gupta, Goh, Naigamwalla, Chia, Koohestani, Mehrotra, McKeown-Eyssen, Giacca and Bruce2003; Catalano et al. Reference Catalano, Bergman and Ader2005) or have a greater AUC for insulin during OGTT (Madsen et al. Reference Madsen, Guerre-Millo and Flindt2002; Tran et al. Reference Tran, Gupta, Goh, Naigamwalla, Chia, Koohestani, Mehrotra, McKeown-Eyssen, Giacca and Bruce2003). This was also observed here, but only 10–35 % of the observed variation in OGTT was explained by the overall animal fatness. Although this study had the advantage of longitudinal measures of body fat content, the MRI scan did not discriminate regional fat distribution and, in both humans (Coon et al. Reference Coon, Rogus, Drinkwater, Muller and Goldberg1992) and rodents (Barzilai & Rossetti, Reference Barzilai and Rossetti1996; Catalano et al. Reference Catalano, Bergman and Ader2005), decreased susceptibility to insulin was associated with increased intra-abdominal fat.

Effect of diet switch

Intake and body composition

Rats switched to the two high-protein diets (HPLC and HPMC) showed modest reduced food intake, by 9 and 3 % as g/d, less than the 13–16 % observed in other studies (Jean et al. Reference Jean, Rome, Mathe, Huneau, Aattouri, Fromentin, Achagiotis and Tome2001; Mithieux et al. Reference Mithieux, Misery, Magnan, Pillot, Gautier-Stein, Bernard, Rajas and Zitoun2005) with diets that included an even higher proportion of gross energy as protein (50 %, compared with 36 % in the current experiment). The modest responses in rodents compares with reductions of energy intake up to 35 % below maintenance requirements observed in obese humans offered high-protein diets (Skov et al. Reference Skov, Toubro, Ronn, Holm and Astrup1999; Stadler et al. Reference Stadler, Burden, McMurry, Gerhard and Connor2003; Weigle et al. Reference Weigle, Breen, Matthys, Callahan, Meeuws, Burden and Purnell2005; Johnstone et al. Reference Johnstone, Murison, Bremner, Horgan and Lobley2006). These reductions led to substantial loss of body weight (Skov et al. Reference Skov, Toubro, Ronn, Holm and Astrup1999; Stadler et al. Reference Stadler, Burden, McMurry, Gerhard and Connor2003; Johnstone et al. Reference Johnstone, Murison, Bremner, Horgan and Lobley2006), much of this as body fat (Johnstone et al. Reference Johnstone, Murison, Bremner, Horgan and Lobley2006).

Nonetheless, even small changes in intake (or metabolism) produce substantial changes over the long term. For example, for the rats fed the HPMC diet, body fat was reduced by 12 % over 8 weeks; this represents a difference in energy retention of 14 kJ/d, similar to the 5 % decrease in ME intake observed. Rats fed the HPLC diet had the lowest absolute intake (g/d) but gained the most fat due to the higher energy density of the ration. This raises the possibility that a combination of high protein and low carbohydrate in the diet has synergistic effects in terms of lowering appetite, at least for the absolute weight of food eaten. Recent data from humans offered iso-energetic rations (Johnstone et al. Reference Johnstone, Murison, Bremner, Horgan and Lobley2006) indicate that diets supplying 30 % of energy as protein with low (20 g/d) as opposed to moderate (170 g/d) carbohydrate intake produce an additional modest reduction in intake (approximately 0·7 MJ/d), coupled with feeling less hungry. Such findings may explain why HPLC diets can be effective in achieving weight loss.

Health

It is well established that risks associated with obesity-related metabolic syndrome e.g. insulin resistance, dyslipaemia, hypertension and CVD, are all improved by weight loss (Pasanisi et al. Reference Pasanisi, Contaldo, de Simone and Mancini2001). In the current study, the dietary changes did not improve systolic blood pressure, while, in human studies, effects of short-term weight loss induced by high-protein diets on blood pressure have been equivocal (Elliott et al. Reference Elliott, Stamler, Dyer, Appel, Dennis, Kesteloot, Ueshima, Okayama, Chan, Garside and Zhou2006). Reduced plasma cholesterol has been a specific target towards reducing cardiovascular risk, and concerns have been expressed about the consequences of high-fat diets (e.g. Nordmann et al. Reference Nordmann, Nordmann, Briel, Keller, Yancy, Brehm and Bucher2006). In practice, for many human trials with low-carbohydrate diets, the total fat consumption is equal or below normal due to either the enforced (Foster et al. Reference Foster, Wyatt, Hill, McGuckin, Brill, Mohammed, Szapary, Rader, Edman and Klein2003; Samaha et al. Reference Samaha, Iqbal, Seshadri, Chicano, Daily, McGrory, Williams, Williams, Gracely and Stern2003) or the voluntary (Stadler et al. Reference Stadler, Burden, McMurry, Gerhard and Connor2003; Johnstone et al. Reference Johnstone, Murison, Bremner, Horgan and Lobley2006) reduction of intake. Consequently, blood lipid parameters remain unchanged or improve with such diets (Foster et al. Reference Foster, Wyatt, Hill, McGuckin, Brill, Mohammed, Szapary, Rader, Edman and Klein2003; Volek et al. Reference Volek, Sharman, Gomez, Scheett and Kraemer2003; Crowe, Reference Crowe2005; Johnstone et al. Reference Johnstone, Murison, Bremner, Horgan and Lobley2006). This was also observed with the rats fed the HPLC diet that comprised 56 % of the dietary energy supplied as fat, much in saturated form, and where there was no increase in plasma cholesterol. Such comparisons need to be treated with caution, however, because there were no differences in cholesterol concentrations between the stock and 37 % fat groups either, and this may simply reflect the differences in fat metabolism between rodents and humans (Siguel, Reference Siguel1983; Nishina et al. Reference Nishina, Schneeman and Freedland1991).

The current study did show changes in the indirect measures of insulin sensitivity and resistance for rats switched from the 37 % fat diet. This was most clear for fasting plasma insulin concentrations that were reduced by 32–37 % in the HPLC and HPMC groups, and were not different from those of rats fed the stock diet. These changes were reflected in the 37 % lowering in calculated HOMA values for both the HPLC and HPMC groups and where, 2 h after the oral glucose load, plasma insulin concentrations or tAUCs were not different from those of rats fed stock pellets. Despite these insulin responses, plasma glucose concentrations 2 h after the oral dose were greater for all rats initially offered the 37 % fat diet compared with those on stock pellets, as were the incremental and total AUCs for glucose. Therefore, the apparent improvement under basal (fasting) conditions for both high-protein groups did not fully extend to the induced hyperglycaemic state and, therefore, full restoration of glucose tolerance (and thus insulin sensitivity) was not achieved. This is not too surprising when the relatively modest fat loss (12 %) is considered for the HPMC group. Nonetheless, in humans, even a relatively small weight loss, assumed to be mainly as fat, is associated with improvements in indices of insulin sensitivity (Diabetes Prevention Program Research Group, 2002; Foster et al. Reference Foster, Wyatt, Hill, McGuckin, Brill, Mohammed, Szapary, Rader, Edman and Klein2003; Samaha et al. Reference Samaha, Iqbal, Seshadri, Chicano, Daily, McGrory, Williams, Williams, Gracely and Stern2003; Johnstone et al. Reference Johnstone, Murison, Bremner, Horgan and Lobley2006).

Although fat loss is the preferred route to improve indices of insulin sensitivity in the obese, other options exist. For example, in the rats fed the HPLC diet, and where body fat increased by 12 % over the 8 weeks of dietary intervention, fasting insulin and HOMA improved. This probably relates to the low carbohydrate intake, because reduced fasting plasma glucose and insulin concentrations have been observed for humans in response to such diets within 4 d (Fery et al. Reference Fery, Bourdoux, Christophe and Balasse1982), with improved insulin sensitivity for obese patients by 2 weeks (Boden et al. Reference Boden, Sargrad, Homko, Mozzoli and Stein2005), before any substantial loss of weight. Furthermore, in obese volunteers on high-protein weight loss diets, the HOMA values were lowered more with low than moderate carbohydrate intakes (Johnstone et al. Reference Johnstone, Murison, Bremner, Horgan and Lobley2006). Together, these data suggest that insulin sensitivity may be improved independently by either fat loss or markedly lowered carbohydrate supply. Indeed, the current rodent data suggest that this latter improvement can occur in the face of increased fat deposition. While improvements in metabolic syndrome through diet alone but in the absence of weight loss is not the ideal scenario, there may be situations where improvements induced by changes in macronutrient supply are better than none at all, and may provide short-term approaches to reducing certain health risks.

Acknowledgements

Grateful thanks are extended to Mrs Donna Wallace and Mrs Wilma Strachan and their staff for expert care of the rats during this study. This work was funded as part of the grant from the Scottish Executive Environment and Rural Affairs Department (SEERAD) to the Rowett Research Institute

References

Alexander, J, Chang, GQ, Dourmashkin, JT & Leibowitz, SF (2006) Distinct phenotypes of obesity-prone AKR/J, DBA2J and C57BL/6J mice compared to control strains. Int J Obes 30, 5059.Google Scholar
Archer, ZA, Rayner, DV & Mercer, JG (2004) Hypothalamic gene expression is altered in underweight but obese juvenile male Sprague-Dawley rats fed a high-energy diet. J Nutr 134, 13691374.Google Scholar
Archer, ZA, Rayner, DV, Rozman, J, Klingenspor, M & Mercer, JG (2003) Normal distribution of body weight gain in male Sprague-Dawley rats fed a high-energy diet. Obes Res 11, 13761383.Google Scholar
Avenell, A, Broom, J, Brown, TJ, Poobalan, A, Aucott, L, Stearns, SC, Smith, WC, Jung, RT, Campbell, MK & Grant, AM (2004) Systematic review of the long-term effects and economic consequences of treatments for obesity and implications for health improvement. Health Technol Assess 8, iii-1v, 1182.Google Scholar
Barzilai, N & Rossetti, L (1996) Age-related changes in body composition are associated with hepatic insulin resistance in conscious rats. Am J Physiol 270, E930E936.Google Scholar
Beltowski, J, Wojcicka, G & Jamroz-Wisniewska, A (2006) Role of nitric oxide and endothelium-derived hyperpolarizing factor (EDHF) in the regulation of blood pressure by leptin in lean and obese rats. Life Sci 79, 6371.Google Scholar
Boden, G, Sargrad, K, Homko, C, Mozzoli, M & Stein, TP (2005) Effect of a low-carbohydrate diet on appetite, blood glucose levels, and insulin resistance in obese patients with type 2 diabetes. Ann Intern Med 142, 403411.Google Scholar
Catalano, KJ, Bergman, RN & Ader, M (2005) Increased susceptibility to insulin resistance associated with abdominal obesity in aging rats. Obes Res 13, 1120.Google Scholar
Chang, S, Graham, B, Yakubu, F, Lin, D, Peters, JC & Hill, JO (1990) Metabolic differences between obesity-prone and obesity-resistant rats. Am J Physiol 259, R1103R1110.Google Scholar
Clegg, DJ, Benoit, SC, Reed, JA, Woods, SC, Dunn-Meynell, A & Levin, BE (2005) Reduced anorexic effects of insulin in obesity-prone rats fed a moderate-fat diet. Am J Physiol 288, R981R986.Google Scholar
Coon, PJ, Rogus, EM, Drinkwater, D, Muller, DC & Goldberg, AP (1992) Role of body fat distribution in the decline in insulin sensitivity and glucose tolerance with age. J Clin Endocrinol Metab 75, 11251132.Google Scholar
Crowe, TC (2005) Safety of low-carbohydrate diets. Obes Rev 6, 235245.Google Scholar
Diabetes Prevention Program Research Group (2002) Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 346, 393403.Google Scholar
Dobrian, AD, Davies, MJ, Prewitt, RL & Lauterio, TJ (2000) Development of hypertension in a rat model of diet-induced obesity. Hypertension 35, 10091015.Google Scholar
Elliott, P, Stamler, J, Dyer, AR, Appel, L, Dennis, B, Kesteloot, H, Ueshima, H, Okayama, A, Chan, Q, Garside, DB & Zhou, B (2006) Association between protein intake and blood pressure: the INTERMAP Study. Arch Intern Med 166, 7987.Google Scholar
Fery, F, Bourdoux, P, Christophe, J & Balasse, EO (1982) Hormonal and metabolic changes induced by an isocaloric isoproteinic ketogenic diet in healthy subjects. Diabetes Metab 8, 299305.Google Scholar
Foster, GD, Wyatt, HR, Hill, JO, McGuckin, BG, Brill, C, Mohammed, BS, Szapary, PO, Rader, DJ, Edman, JS & Klein, S (2003) A randomized trial of a low-carbohydrate diet for obesity. N Engl J Med 348, 20822090.Google Scholar
Gao, J, Ghibaudi, L, van Heek, M & Hwa, JJ (2002) Characterization of diet-induced obese rats that develop persistent obesity after 6 months of high-fat followed by 1 month of low-fat diet. Brain Res. 936, 8790.Google Scholar
Grundy, SM (2005) A constellation of complications: the metabolic syndrome. Clin Cornerstone 7, 3645.Google Scholar
Hawkins, RA, Mans, AM & Davis, DW (1986) Regional ketone body utilization by rat brain in starvation and diabetes. Am J Physiol 250, E169E178.Google Scholar
Hellstrom, PM, Geliebter, A, Naslund, E, Schmidt, PT, Yahav, EK, Hashim, SA & Yeomans, MR (2004) Peripheral and central signals in the control of eating in normal, obese and binge-eating human subjects. Br J Nutr 92, Suppl. 1, S47S57.Google Scholar
Jean, C, Rome, S, Mathe, V, Huneau, JF, Aattouri, N, Fromentin, G, Achagiotis, CL & Tome, D (2001) Metabolic evidence for adaptation to a high protein diet in rats. J Nutr 131, 9198.Google Scholar
Johnstone, AM, Murison, SD, Bremner, DM, Horgan, G & Lobley, GE (2006) Hunger and appetite response to a high protein ketogenic diet in obese men feeding ad libitum. Proc Nutr Soc 65, 90A.Google Scholar
Kahn, R, Buse, J, Ferrannini, E & Stern, M (2005) The metabolic syndrome: time for a critical appraisal: joint statement from the American Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care 28, 22892304.Google Scholar
Kannel, WB, Brand, N, Skinner, JJ Jr, Dawber, TR & McNamara, PM (1967) The relation of adiposity to blood pressure and development of hypertension. The Framingham study. Ann Intern Med 67, 4859.Google Scholar
Kuller, LH (2006) Nutrition, lipids, and cardiovascular disease. Nutr Rev 64, S15S26.Google Scholar
Langley-Evans, SC, Clamp, AG, Grimble, RF & Jackson, AA (1996) Influence of dietary fats upon systolic blood pressure in the rat. Int J Food Sci Nutr 47, 417425.Google Scholar
Levin, BE, Dunn-Meynell, AA, Balkan, B & Keesey, RE (1997) Selective breeding for diet-induced obesity and resistance in Sprague-Dawley rats. Am J Physiol 273, R725R730.Google Scholar
Li, PK, Lee, JT, MacGillivray, MH, Schaefer, PA & Siegel, JH (1980) Direct, fixed-time kinetic assays for beta-hydroxybutyrate and acetoacetate with a centrifugal analyzer or a computer-backed spectrophotometer. Clin Chem 26, 17131717.Google Scholar
Lorenzo, C, Williams, K, Hunt, KJ & Haffner, SM (2006) Trend in the prevalence of the metabolic syndrome and its impact on cardiovascular disease incidence: the San Antonio Heart Study. Diabetes Care 29, 625630.Google Scholar
Madsen, L, Guerre-Millo, M, Flindt, EN, et al. (2002) Tetradecylthioacetic acid prevents high fat diet induced adiposity and insulin resistance. J Lipid Res 43, 742750.Google Scholar
Maloney, CA, Gosby, AK, Phuyal, JL, Denyer, GS, Bryson, JM & Caterson, ID (2003) Site-specific changes in the expression of fat-partitioning genes in weanling rats exposed to a low-protein diet in utero. Obes Res 11, 461468.Google Scholar
Mithieux, G, Misery, P, Magnan, C, Pillot, B, Gautier-Stein, A, Bernard, C, Rajas, F & Zitoun, C (2005) Portal sensing of intestinal gluconeogenesis is a mechanistic link in the diminution of food intake induced by diet protein. Cell Metab 2, 321329.Google Scholar
Molnar, J, Yu, S, Mzhavia, N, Pau, C, Chereshnev, I & Dansky, HM (2005) Diabetes induces endothelial dysfunction but does not increase neointimal formation in high-fat diet fed C57BL/6J mice. Circ Res 96, 11781184.Google Scholar
Morris, MJ, Velkoska, E & Cole, TJ (2005) Central and peripheral contributions to obesity-associated hypertension: impact of early overnourishment. Exp Physiol 90, 697702.Google Scholar
Nishina, PM, Schneeman, BO & Freedland, RA (1991) Effects of dietary fibers on nonfasting plasma lipoprotein and apolipoprotein levels in rats. J Nutr 121, 431437.Google Scholar
Nordmann, AJ, Nordmann, A, Briel, M, Keller, U, Yancy, WS Jr, Brehm, BJ & Bucher, HC (2006) Effects of low-carbohydrate vs low-fat diets on weight loss and cardiovascular risk factors: a meta-analysis of randomized controlled trials. Arch Intern Med 166, 285293.Google Scholar
Pasanisi, F, Contaldo, F, de Simone, G & Mancini, M (2001) Benefits of sustained moderate weight loss in obesity. Nutr Metab Cardiovasc Dis 11, 401406.Google Scholar
Rees, WD, Hay, SM & Cruickshank, M (2006) An imbalance in the methionine content of the maternal diet reduces postnatal growth in the rat. Metabolism 55, 763770.Google Scholar
Samaha, FF, Iqbal, N, Seshadri, P, Chicano, KL, Daily, DA, McGrory, J, Williams, T, Williams, M, Gracely, EJ & Stern, L (2003) A low-carbohydrate as compared with a low-fat diet in severe obesity. N Engl J Med 348, 20742081.Google Scholar
Siguel, EN (1983) Cancerostatic effect of vegetarian diets. Nutr Cancer 4, 285291.Google Scholar
Skov, AR, Toubro, S, Ronn, B, Holm, L & Astrup, A (1999) Randomized trial on protein vs carbohydrate in ad libitum fat reduced diet for the treatment of obesity. Int J Obes Relat Metab Disord 23, 528536.Google Scholar
Stadler, DD, Burden, V, McMurry, M, Gerhard, G & Connor, W (2003) Impact of a 42-day Atkins diet and energy-matched low-fat diet on weight and anthrometric indices. FASEB J 17, A745.Google Scholar
Thim, T, Bentzon, JF, Kristiansen, SB, Simonsen, U, Andersen, HL, Wassermann, K & Falk, E (2006) Size of myocardial infarction induced by ischaemia/reperfusion is unaltered in rats with metabolic syndrome. Clin Sci 110, 665671.Google Scholar
Tran, TT, Gupta, N, Goh, T, Naigamwalla, D, Chia, MC, Koohestani, N, Mehrotra, S, McKeown-Eyssen, G, Giacca, A & Bruce, WR (2003) Direct measure of insulin sensitivity with the hyperinsulinemic-euglycemic clamp and surrogate measures of insulin sensitivity with the oral glucose tolerance test: correlations with aberrant crypt foci promotion in rats. Cancer Epidemiol Biomarkers Prev 12, 4756.Google Scholar
Truesdale, KP, Stevens, J, Lewis, CE, Schreiner, PJ, Loria, CM & Cai, J (2006) Changes in risk factors for cardiovascular disease by baseline weight status in young adults who maintain or gain weight over 15 years: the CARDIA study. Int J Obes (Lond) 30, 13971407.Google Scholar
Truett, AA, Borne, AT, Monteiro, MP & West, DB (1998) Composition of dietary fat affects blood pressure and insulin responses to dietary obesity in the dog. Obes Res 6, 137146.Google Scholar
Velkoska, E, Cole, TJ & Morris, MJ (2005) Early dietary intervention: long-term effects on blood pressure, brain neuropeptide Y, and adiposity markers. Am J Physiol 288, E1236E1243.Google Scholar
Volek, JS & Feinman, RD (2005) Carbohydrate restriction improves the features of metabolic syndrome. Metabolic syndrome may be defined by the response to carbohydrate restriction. Nutr Metab 2, 31.Google Scholar
Volek, JS, Sharman, MJ, Gomez, AL, Scheett, TP & Kraemer, WJ (2003) An isoenergetic very low carbohydrate diet improves serum HDL cholesterol and triacylglycerol concentrations, the total cholesterol to HDL cholesterol ratio and postprandial pipemic responses compared with a low fat diet in normal weight, normolipidemic women. J Nutr 133, 27562761.Google Scholar
Weigle, DS, Breen, PA, Matthys, CC, Callahan, HS, Meeuws, KE, Burden, VR & Purnell, JQ (2005) A high-protein diet induces sustained reductions in appetite, ad libitum caloric intake, and body weight despite compensatory changes in diurnal plasma leptin and ghrelin concentrations. Am J Clin Nutr 82, 41–48.Google Scholar
Westerterp-Plantenga, MS, Rolland, V, Wilson, SA & Westerterp, KR (1999) Satiety related to 24 h diet-induced thermogenesis during high protein/carbohydrate vs high fat diets measured in a respiration chamber. Eur J Clin Nutr 53, 495502.Google Scholar
World Health Organization (2000) Obesity: Preventing and Managing the Global Epidemic. Report of a WHO consultation. WHO Technical Report Series 894. Geneva: World Health Organization.Google Scholar
Wilde, DW, Massey, KD, Walker, GK, Vollmer, A & Grekin, RJ (2000) High-fat diet elevates blood pressure and cerebrovascular muscle Ca(2+) current. Hypertension 35, 832837.Google Scholar
Wolever, TM & Jenkins, DJ (1986) The use of the glycemic index in predicting the blood glucose response to mixed meals. Am J Clin Nutr 43, 167–172.Google Scholar
Wood, RJ, Volek, JS, Liu, Y, Shachter, NS, Contois, JH & Fernandez, ML (2006) Carbohydrate restriction alters lipoprotein metabolism by modifying VLDL, LDL, and HDL subfraction distribution and size in overweight men. J Nutr 136, 384389.Google Scholar
Wynne, K, Stanley, S, McGowan, B & Bloom, S (2005) Appetite control. J Endocrinol 184, 291–318.Google Scholar
Yeni-Komshian, H, Carantoni, M, Abbasi, F & Reaven, GM (2000) Relationship between several surrogate estimates of insulin resistance and quantification of insulin-mediated glucose disposal in 490 healthy nondiabetic volunteers. Diabetes Care 23, 171–175.Google Scholar
Yoshioka, S, Uemura, K, Tamaya, N, Tamagawa, T, Miura, H, Iguchi, A, Nakamura, J & Hotta, N (2000) Dietary fat-induced increase in blood pressure and insulin resistance in rats. J Hypertens 18, 18571864.Google Scholar
Figure 0

Table 1 Compositions, as g/kg as fed, of the various diets used in this study

Figure 1

Table 2 Body composition and blood pressure data from Hooded Lister rats (25 weeks old) fed for 20 weeks on either stock pellets (n 12) or a 37 % fat diet (15 % protein; 48 % carbohydrate; 37 % fat; n 48)

Figure 2

Table 3 Oral glucose tolerance test data from Hooded Lister rats (25 weeks old) fed for 20 weeks on either stock pellets (n 12) or a 37 % fat diet (15 % protein; 48 % carbohydrate; 37 % fat; n 48)

Figure 3

Fig. 1 Temporal changes in plasma (a) glucose and (b) insulin in overnight fasted rats given an oral dose of glucose (2 g/kg body weight). Rats were offered either stock pellets (□) or a 37 % fat diet (■). Data are means with their standard error.

Figure 4

Table 4 Body weight, composition, food intake, macronutrient intake (protein, carbohydrate and fat), metabolisable energy intake and blood pressure in obese rats following 8 weeks on diets of either 37 % fat, HPLC, HPMC or the 37 % fat diet fed to the group intake mean of HPMC (GF)

Figure 5

Fig. 2 Body weight (g) and body fat (g) during the last 8 weeks of the study after diet switch. Rats remained on either ther 37 % fat diet (○) or stock pellets (□), or were switched to high-protein, low-carbohydrate (Δ), high protein-moderate carbohydrate (▲) or fed the 37 % fat diet to intake of the HPMC diet (●).

Figure 6

Table 5 Fasting glucose and insulin concentrations, HOMA calculation and OGTT data in obese rats following 6 weeks on diets of either 37 % fat, HPLC, HPMC or the 37 % fat diet fed to the group intake mean of HPMC (GF)

Figure 7

Table 6 Plasma concentrations of insulin, leptin, glucose, cholesterol, NEFA, triacylglycerols and 3-hydroxybutyrate after 8 weeks of diet switch