In this study both HP and HC energy-restricted, isocaloric diets combined with regular moderate intensity exercise provided substantial improvements in weight, glycaemic control, and cardiometabolic risk factors in overweight and obese adults with T2DM. Furthermore, these improvements were maintained during a subsequent 12 weeks period when weight remained stable, allowing for two distinct phases to compare outcomes.
Improvements in glycaemic control were the primary outcome of this study. Our primary analysis showed substantial reductions in HbA1c, fasting blood glucose levels, and fasting insulin with no differences between the groups which is consistent with similar weight loss studies [3
] even when adjustments for changes in medication have been made [5
]. Another study reported that additional benefits for a lower-carbohydrate diet were only seen in participants whose baseline mean HbA1c levels were higher (>7.8%) even though the reduction in MES in that group was two-fold greater [14
]. Diabetes medication is often titrated in response to fluctuations in glycaemic control. The change in medication observed in our participants was small, as indicated by the MES, but may have obscured any benefits for a HP diet. After conducting a sub-group analysis, including only those who did not change their diabetes medication, we found that the HP group achieved a 0.7% greater reduction in HbA1c than the HC group following phase 1 with no difference between the groups for weight loss. Interestingly, this supports a previous study where a 0.7% greater reduction in HbA1c was found after following a hypocaloric lower-carbohydrate diet (14%Carbohydrate:27%Protein:54%Fat) compared to an isocaloric diet (50%C:19%P:25%F) for 24 weeks [14
]. In contrast to that study, our dietary intervention consisted of a moderate carbohydrate difference between the groups of 15% of total energy (or approximately 50 g carbohydrate) suggesting that even modest carbohydrate restriction can elicit glycaemic improvements. Following weight maintenance, the difference between the means (0.6%) in the sub-group analysis did not reach statistical significance suggesting we were under-powered to detect this level of change. However, this result is of clinical significance and is consistent with a five weeks cross-over study that reported their HP diet (30%P:40%C:30%F) achieved a 0.5% greater reduction in HbA1c than the HC (15%P:55%C:30%F) where participants were not treated with diabetes medication and were weight stable [15
]. The benefits we found in our sub-group analysis demonstrate the need for caution when interpreting results in T2DM studies where diabetes medication dosage fluctuates. Nevertheless, including participants with or without medications in studies is necessary to generalize the results.
While most studies focus on weight loss there is currently limited evidence available to support prescriptive higher-protein diets with higher-carbohydrate diets in adults with T2DM which incorporate a concurrent exercise protocol and report outcomes following successful weight stabilization over an appropriate duration to see changes. Our study was designed to promote weight loss initially and then re-evaluate the outcomes following 12 weeks of energy-balance without the influence of weight change so as to better isolate the effects of diet composition independent of any effects of weight loss. The 12 weeks duration was considered sufficient time to allow for any changes in HbA1c to be evident. The present study showed that there was no significant change in HbA1c when energy intake was increased for weight maintenance while weight was maintained for 12 weeks. Both dietary patterns increased carbohydrate similarly intake by 17–20 g. This is in contrast to findings of a previous study where HbA1c rebounded after a nine month weight maintenance phase [5
]. Although this previous study had a longer weight maintenance period, we achieved a 2.5-fold greater weight loss prior to the weight maintenance period (−8 kg vs.
−3 kg) resulting in a greater reduction in HbA1c (−1.5% vs.
−0.5%) which provided us with two very distinct phases with which to compare results. The sustained improvement in HbA1 reported in our study (−1.4%) is of clinical significance as it has been reported that the risk of a stroke is reduced by 12%, myocardial infarction by 14%, and diabetes-related deaths by 21% for each 1% sustained decrease in HbA1c [16
]. The magnitude of HbA1c reduction we observed is similar to the magnitude of reduction that is expected using oral antihyperglycaemic medication in monotherapy [17
]. Apart from the obvious financial benefits, glycaemic control through lifestyle modifications has many advantages over pharmacological usage. Side effects associated with antihyperglycaemic medication, including gastrointestinal upsets, hypoglycemia, renal impairment, and weight gain, can be reduced or eliminated as medication is lowered or stopped. Coupled with the improvement for HbA1c there was a significant reduction in antihyperglycaemic medication use over the course of the study. This was a smaller reduction than seen in other studies which have utilized the same MES method that have reported greater reductions (−0.5 to −1.24), but these studies have used diets with substantially lower-carbohydrate intakes (14%–20% carbohydrate) [9
There were no differences between the groups for changes in body composition: total fat mass (FM) and fat free mass (FFM) accounted for 79% and 20% of weight loss respectively. This is consistent with a previous 16 weeks study of adults with T2DM who followed a HP or HC diet with or without resistance training (RT) which showed that FM loss accounted for 80% of weight lost in the RT groups and 77% for the diet only groups and FFM loss accounted for 20% of weight loss in the RT groups and 23% for the diet only groups [3
]. Another 20-week study of overweight and obese women with polycystic ovary syndrome, which displays similar cardiometabolic pathology as T2DM (abdominal obesity, insulin resistance, hypertension, and dyslipidemia), also concluded that the inclusion of exercise to a hypocaloric diet provided greater FM loss and preservation of some FFM [19
]. Our covariate analysis did not identify any additional benefits in body composition from exercise. This may be due to our shorter weight loss phase of 12 weeks whereas the above mentioned studies were of 16–20 weeks duration. Furthermore, with both groups achieving similar weight loss and improvements in fitness, as evidenced from improved treadmill speed at the same RPE, it may be possible that the moderate weight loss of 8 kg may have masked any independent effect that exercise may have had. Importantly, we showed that body composition benefits remained when weight lost was sustained.
Apart from weight loss and adequate glycaemic control, improving BP, insulin resistance, and CRP and lipid profiles can also play an important role in lowering CVD risk in diabetes. Mean BP values at baseline indicate that our participants were within the recommendations set out in the American Diabetes Association’s (ADA) guidelines for T2DM management [20
]. Nevertheless, BP improved following weight loss with both diets which is consistent with previous findings [3
]. BP medication was adjusted during the study and a sub-group analysis for those who did not change their medication revealed the HP group reduced DBP to a greater degree than the HC group. Notably, the significance was lost when adjusted for the change in treadmill speed suggesting that exercise may have provided additional benefits beyond the weight loss or dietary patterns for those who did not change their blood pressure medication. Moreover, BP remained stable during phase 2. A previous study showed that BP improvements following a hypocaloric weight loss diet did not rebound in a weight stabilization phase when energy intake was increased in overweight and obese adults without T2DM [21
]. The energy-balance phase of that study was short (4 weeks) so our study extends this evidence to include a 12 weeks weight maintenance phase and individuals with T2DM. The overall reduction in SBP (−11 mmHg) is clinically noteworthy as a 10 mmHg decrease is associated with a 11% reduction in risk for myocardial infarction and 15% reduction in risk for diabetes-related deaths with benefits also seen in normotensive individuals [22
One of the independent predictors of CVD is insulin resistance through its association with hyperglycemia, hyperinsulinemia, hypertension, and dyslipidemia [23
]. Using the HOMA2-IR cut-off value for insulin resistance of >1.8, as determined by the BRAMS study [24
], our participants were substantially insulin resistant at baseline. However, following weight loss in phase 1, HOMA2-IR indices had markedly decreased which supports previous findings [25
]. It has been reported that the risk of CVD increases by 56% with a 1 unit increase in HOMA-IR [26
], therefore the overall reduction of −1.44 and −1.04 for the HP and HC diets respectively is considered clinically important. Following phase 2 in our study, HOMA2-IR scores indicated that insulin resistance had normalized (<1.8 for both groups), denoting a reduction in CVD risk.
CRP, a major acute-phase protein associated with chronic systemic inflammation, has been associated with obesity (particularly abdominal adiposity) and insulin resistance and may predict coronary heart disease risk [27
]. In our study, CRP significantly decreased following weight loss, thereby supporting the current evidence [28
] of there not being any difference between the HP or HC diets, which is in agreement with other studies regarding T2DM [3
]. Increases in CRP are also related to worsening glycaemic control in T2DM [29
] and we report a strong positive correlation between reductions in CRP levels and reductions in HbA1c. Improvements in CRP remained during phase 2.
After an initial reduction in phase 1, total cholesterol and LDL levels rebounded in phase 2 when weight was stabilised, whereas HDL improved during phase 2. Interestingly, the pattern of change observed is similar to studies previously reported in T2DM individuals [4
] and overweight and obese adults [30
] that show lipids decline with caloric restriction but rebound with weight stabilization and energy balance. At baseline, triglycerides level were above current ADA recommendations, but decreased in phase 1 to within recommended levels with no difference between the diets. This result is consistent with the findings from a meta-analysis of T2DM studies [31
] but not from a meta-analysis of overweight and obese adults where HP diets resulted in a significant 0.23 mmol/L greater reduction in triglyceride levels than standard protein diets [32
]. A previous study reported that a carbohydrate restricted diet provided greater reductions in triglycerides than a calorie restricted diet (−55% vs.
−28%) despite a similar 4.3% weight loss [33
], which is supported in other studies where carbohydrate content was more severe [18
]. Triglyceride levels remained stable during phase 2. It was interesting to note the additional benefits exercise appeared to exhibit on HDL results. This supports the findings from a meta-analysis of 35 studies which reported that regular aerobic exercise (for at least 8 weeks) modestly increased HDL by 2.53 mg/dL (95%CI 1.36 to 3.70) (equates to 0.065 mmol/L) more than non-exercising participants, but exercise duration needed to be ≥120 min/week [34
]. However, this is contrary to another meta-analysis of 17 studies which found no difference between HDL between the aerobic exercise groups and non-exercising groups [35
Comparable weight loss and the participants’ good compliance to their allocated macronutrient profile, energy intakes, and physical activity adds strength to our study as we can attribute metabolic changes to the differences between the intervention groups. The effective stabilization of weight in our 12-week maintenance phase is another strength of this study as it enabled further evaluation of outcomes without the confounders of energy imbalances and weight regain. Dietary compliance can be ascribed to the intensive dietetic and professional support given throughout the study, as evidence suggests that achieving weight goals is more successful when instructions are led by a registered dietitian [36
]. We provided two-weekly individual visits, prescriptive diet plans, recipe books, regular monitoring of weight, and core study foods. However, it may limit the potential translation of results from the study into clinical practice where such intensive dietetic support may not be realistic. To ensure our results would be of clinical interest to health professionals, our dietary patterns were designed to incorporate all core food groups with the HP diet including a moderate carbohydrate intake and low to medium glycaemic index carbohydrates.
The lack of difference seen between the diets may have been partly due to the macronutrient composition. Circulating blood glucose as a product of carbohydrate metabolism could have had an impact on HbA1c, glucose, insulin, and triglyceride levels and our diets were not planned to have the differences in carbohydrate intake which are often seen in studies which report greater benefits for a HP diet over a HC diet [14
]. However, it is likely that weight loss from the lifestyle intervention is the driving force for the metabolic changes as it was in this phase that most variables demonstrated significant changes. A 2%–10% reduction in body weight is associated with reductions in HbA1c, BP, and triglycerides and increased HDL cholesterol [2
]. In this study both diets achieved a moderate weight loss of ~8% and met the exercise protocol which has been reflected in the results for these outcomes.
Limitations in this study should be considered when interpreting the results. Firstly, our sample size may not have provided the statistical power to detect smaller changes between groups, particularly for sub-group analysis. Secondly, the differences in the outcomes cannot be entirely attributed to differences in the protein/carbohydrate ratio between the two groups and other dietary factors other than carbohydrate and protein could have also influenced the results. The total fat and saturated fat (SFA) intakes in the HP group were 7%–8% and 3% higher, respectively. This was a result of larger meat portions and the inclusion of reduced-fat dairy and almonds in the HP group. Fibre intake was higher in the HC group (~3.3 g/day). However, we are comparing two dietary patterns with concurrent exercise rather than specific nutrients. Although our diets were prescriptive, participants had choices within some types of food (e.g., non-starchy vegetables, fruit). It would be impossible to assign changes in outcomes to a specific nutrient when food is a combination of nutrients, all of which may impact health outcomes. Current Australian dietary guidelines recommend total fat intake to account for 20%–35% of energy, SFA < 10% of energy intake, and an adequate intake of fibre being 25–30 g/day [37
]. Therefore our dietary patterns are within the guidelines, although SFA was ~10% during weight maintenance. The POUNDS LOST trial of overweight /obese adults (n
= 424) compared four isocaloric, energy-restricted diets: two were low fat (26% en (energy)–28% en) with either low protein (18% en–20% en) or high protein (21% en–23% en) and two were high fat (33% en–35% en) with either low protein (18% en–20% en) or high protein (21% en–23% en) and saturated fat intake was 7% en–11% en [38
]. Their results led to the conclusion that the significant decreases in fat mass and abdominal fat were dependent on energy intake rather than macronutrient distribution as there were no dietary differences. Another study of obese adults with T2DM (n
= 115) following hypocaloric diets that were either low-carbohydrate (14%C:27%P:54%F:10%SFA) or a HC (50%C:19%P:25%F:7.5%SFA) over 24 weeks found that both diets improved body composition, BP, HOMA2-IR, and CRP with the LC diet exhibiting greater improvements in glycaemic control, triglycerides, and HDL [14
]. Thirdly, with both groups completing the same exercise protocol, it preludes the ability to separate the effects of diet and physical activity on the outcomes examined. Our aim was to compare the effects of the two dietary patterns when administered as part of a holistic lifestyle modification program with both diet and exercise being the key elements to the management of T2DM.