Effect of a 90 g/day low-carbohydrate diet on glycaemic control, small, dense low-density lipoprotein and carotid intima-media thickness in type 2 diabetic patients: An 18-month randomised controlled trial

Aim This study explored the effect of a moderate (90 g/d) low-carbohydrate diet (LCD) in type 2 diabetes patients over 18 months. Methods Ninety-two poorly controlled type 2 diabetes patients aged 20–80 years with HbA1c ≥7.5% (58 mmol/mol) in the previous three months were randomly assigned to a 90 g/d LCD r traditional diabetic diet (TDD). The primary outcomes were glycaemic control status and change in medication effect score (MES). The secondary outcomes were lipid profiles, small, dense low-density lipoprotein (sdLDL), serum creatinine, microalbuminuria and carotid intima-media thickness (IMT). Results A total of 85 (92.4%) patients completed 18 months of the trial. At the end of the study, the LCD and TDD group consumed 88.0±29.9 g and 151.1±29.8 g of carbohydrates, respectively (p < 0.05). The 18-month mean change from baseline was statistically significant for the HbA1c (-1.6±0.3 vs. -1.0±0.3%), 2-h glucose (-94.4±20.8 vs. -18.7±25.7 mg/dl), MES (-0.42±0.32 vs. -0.05±0.24), weight (-2.8±1.8 vs. -0.7±0.7 kg), waist circumference (-5.7±2.7 vs. -1.9±1.4 cm), hip circumference (-6.1±1.8 vs. -2.9±1.7 cm) and blood pressure (-8.3±4.6/-5.0±3 vs. 1.6±0.5/2.5±1.6 mmHg) between the LCD and TDD groups (p<0.05). The 18-month mean change from baseline was not significantly different in lipid profiles, sdLDL, serum creatinine, microalbuminuria, alanine aminotransferase (ALT) and carotid IMT between the groups. Conclusions A moderate (90 g/d) LCD showed better glycaemic control with decreasing MES, lowering blood pressure, decreasing weight, waist and hip circumference without adverse effects on lipid profiles, sdLDL, serum creatinine, microalbuminuria, ALT and carotid IMT than TDD for type 2 diabetic patients.

( Adult Treatment Panel 3). The traditionally diabetic diet recommended 50-60% of intake from carbohydrates and less than 30% from fat. But there is not enough medical evidence, including: a systematic review or meta-analysis to support such a proposal [1].
The 2007ADA "Standards of Medical Care in Diabetes" did not recommend low-carb diets because there was not enough evidence to support the long-term effects and the risk of cardiovascular disease was still unclear [2]. But in 2008 ADA changed the discourse [3], they suggested short-term use of low-fat diet or a low carbohydrate diet for overweight diabetics or non-diabetics.( for one year in 2008, and 2 years in 2011 [4]. In fact, ADA published "Life With Diabetes" with documentation of "Carbohydrate, protein and lipid provide calories, but only carbohydrate affects blood glucose " [5] . In 2013, the ADA even removed the words that 130 grams of carbohydrates a day was required [6]. In 2014, ADA mentioned low fat diet can't achieve good blood glucose control and reduce the cardiovascular risk by metaanalysis and systematic review [7]. Although the low-carbohydrate diet is not mainstream, there is growing evidence that such treatments are low-risk and easy to follow [2] . The low carbohydrate diet, in accordance with the latest proposal [8]: classified as: (1) very low-carbohydrate ketogenic diet, VLCKD, the intake of carbohydrate per day of 20 -50 grams, or less than 10% of a daily diet of 2000 calories; (2) low-carbohydrate diet : the intake of carbohydrates is less than 130 grams per day , or less than 26% of total daily energy ; (3) medium carbohydrate diet : the intake of carbohydrate is 26-45% of the total daily energy per day ; (4) High-carbohydrate diet : The intake of glycogen is > 45% of the total daily energy of the daily diet.
A recent systematic review [1] found that low carbohydrate diet, low glycemic index diet, Mediterranean diet and high protein diet control blood glucose better than the control group. Low-carbohydrate diet and Mediterranean diet were most effective for weight loss.
Most of the review literatures [9][10][11][12][13][14][15][16][17] targeted on obese persons [9][10][11][14][15][16][17] , three articles included 30-40% of diabetic patients [9.10.16] and four focused on obese diabetic patients [11 , 14, 15, 17] , only two are not for obese persons, including a study for diet-controlled diabetic patient [12] and a study for general diabetes patients [13] . In the study of mixed diabetic and non-diabetic patients, the 6 -month study [9] showed significant improvement in body weight, blood glucose, and triglyceride; however, in the one-year study [10], low carbohydrate diet showed no significant weight loss effect, possible due to good diet control in control group, poor compliance of low carbohydrate diet , high drop out rate. The other one-year study compared low-carb diets and low-fat diets with weight loss pills "Orlistat " [16].There was no significant difference between the two groups. The low carbohydrate intake at 48 weeks for this study was 62 Gram / day (15% calories ) , showing good compliance.
In a 6-month study of patients with diabetes, 40% carbohydrate diet was not superior to traditional diabetic diet on glycemic control [12]. In another non-randomized controlled study in Japan for two years, 40% of carbohydrates showed, a better control on blood glucose, weight, total cholesterol and low density lipoprotein, reducing the types and dosage of oral hypoglycemic agents, especially sulfonylurea [13] . In a study for overweight diabetic patients, very low carbohydrate diet (20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30) grams/day) showed effectiveness in reduction of HbA1c and body weight in the first 3 months. But it turned to be 77 ± 44 g carbohydrate/day (24% of calories) after one year and showed no difference between low carbohydrate diet and control groups [14].
For renal function, a randomized controlled trial of three different diets ( low carbohydrate diet, Mediterranean diet, low-fat diet ) published in Diabetes Care [19] in August 2013 (n=318) , showed that all of the above diets were beneficial in weight loss and increased eGFR in chronic kidney disease( CKD) stage III or less patients in two years and without difference between groups.
For cardiovascular risk factors, such as lipid profile and CRP, a systematic review of low carbohydrate diets published in the Obes Rev in November 2012 [20], showed significant reduction of body weight, body mass index (BMI), blood pressure, Abdominal circumference , triglyceride (TG), HbA1c, insulin, and CRP . There was significant increase of HDL and no significant change of LDL and creatinine. Another study showed that a low carbohydrate diet reduced hs-CRP and increase total Adiponectin in three months [21].
Long-term randomized controlled trials revealed that high-protein low-carbohydrate diets didn't show significant advantage in HbA1c and lipid metabolism than high-carbohydrate diets for patients with type 2 overweight or obese diabetes in one year follow-up [22]. However, the high protein low-carbohydrate diet was 30% protein and 40% carbohydrate, compared with 15% protein and 55% carbohydrate in high carbohydrate diet. There may be behavioral changes in both groups, or similar self-reported dietary status and food choices. The sample size is 93 with adequate power. However, the carbohydrate intake was not shown in the study. A two-year research based on the same low-carbohydrate formula (n=419) [23] (200g carbohydrate/day in a daily energy of 2000 kcal) didn't show benefit than the traditional diabetic diet. It is possible that more carbohydrate restriction result in better control. Another two-year study in overweight or obese type 2 diabetes restricted 20% daily carbohydrate in men and women (total energy1800 kcal and 1600 kcal/day respectively) with small sample size 61. There is no significant difference in body weight, HbA1c , insulin and high-density lipoprotein.
In a small sample size study with half diabetes and non-diabetes (n=26) [25], low carbohydrate diet with 40 grams carbohydrate per day showed significant reduction in HbA1c and body weight in three months, but no significant benefit after two years of observation .
Literature 8 made a systematic review for low carbohydrate diet and proposed the 12 points of evidences: 1. Hyperglycemia is the most salient feature of diabetes. Dietary carbohydrate restriction has the greatest effect on decreasing blood glucose levels. In a prospective study [14]compared a VLCKD with a low-calorie diet over a 24-wk period in 102 diabetic and 261 nondiabetic individuals , blood glucose dropped more dramatically in the VLCKD group than in those given the low-calorie diet (HbA1c: VLCKD: 6.2%, low-calorie diet: > 7.5%). dietary carbohydrate in men rose from 42% to 49% of calories. For women, carbohydrate rose from 45% to 52%. The absolute amount of fat decreased for men during this period and showed only a slight increase for women. Data reveals the rise, during this period, of the incidence of type 2 diabetes to its current near epidemic proportions [27] Studies have shown that [28] the stimulation of insulin secretion increases the anabolism of triglycerides (TG) and other TG-rich lipoproteins. It is also believed that fat accumulation in the liver and pancreas will increase the production of VLDL and bring fat to the pancreas, which will have a negative effect on the function of beta-cell [29].
3. Benefits of dietary carbohydrate restriction do not require weight loss. Type 2diabetes took limited carbohydrate diet (30% carbohydrate (CHO), 30% protein, 40% fat, (30:30:40) , which didnnnot fulfill the low carbohydrate diet definition (<130g/d or <26% total energy) , got better glycemic control despite no significant weight loss after ten weeks [30]. 4. Although weight loss is not required for benefit, no dietary intervention is better than carbohydrate restriction for weight loss. One study [31] randomly allocated 26 people to either a low-carbohydrate diet (40 g/d carbohydrate) or a "healthy-eating diet" following Diabetes UK nutritional recommendations for 3 mo. Thirteen people with type 2 diabetes and 13 controls without diabetes were included. Weight loss was greater in the low-carbohydrate arm (6.9 versus 2.1 kg).
The Women's Health Initiative (WHI) is the most recent example. In the study [32], diet performance in 48,000 postmenopausal women was compared with usual behavior. The low-fat intervention group had modest weight loss (average 2.2 kg) occurred in the first year, but regained weight at the end of study. The low-carbohydrate diets like the Atkins diet [33,34] put no formal limit on caloric consumption on the assumption that the greater satiety of protein and fat will provide control of intake.
5. Adherence to low-carbohydrate diets in people with type 2 diabetes is at least as good as adherence to any other dietary interventions and is frequently significantly better [35][36][37]. The good adherence is because of easy to follow without calculation of energy, greater satiety of protein and fat, decreasing fluctuation of blood glucose, reducing the diabetic medication and insulin use. 6. Replacement of carbohydrate with protein is generally beneficial effect on weight loss, body composition, resting metabolic rate, and cardiovascular risk than fat-reduced diets [38][39][40].
7. Dietary total and saturated fat do not correlate with risk for cardiovascular disease.
8. Plasma saturated fatty acids are controlled by dietary carbohydrate more than by dietary lipids It is increasingly understood that plasma SFAs are associated with increased risk for CVD and insulin resistance [51] . In humans, plasma SFAs do not correlate with dietary saturated fat but, rather, are more dependent on dietary carbohydrates [52][53][54]. Elevated SFAs arise from increased production of TG-containing lipoproteins, reduced clearance, and the effect of dietary carbohydrate on de novo fatty acid synthesis. 9. The best predictor of microvascular and, to a lesser extent, macrovascular complications in patients with type 2 diabetes, is glycemic control (HbA1c) There was a 14% decrease in MI for every 1% reduction in HbA1c. There was a dramatic 37% decrease in these end points for microvascular risk for each 1% reduction in HbA1c [55][56][57]. 10. Dietary carbohydrate restriction is the most effective method (other than starvation) of reducing serum TGs and increasing high-density lipoprotein. The low-GI diet increases the high-density lipoprotein (HDL) levels [11,[58][59].
Total and/or LDL cholesterol are the most commonly assessed lipid markers for CVD risk despite the general recognition that they are not good predictors [60][61][62]. Several other parameters have been shown to provide stronger evidence of risk and these tend to be reliably improved by dietary carbohydrate restriction. These include apolipoprotein (apo) B , ratio of total cholesterol to HDL, higher populations of the smaller dense LDL known as pattern B , as well as the ratio of apoB to apoA1. The ratio of TG to HDL, which is also improved more by carbohydrate restriction is taken as a correlate of the smaller dense LDL, which is not routinely measured [63]. 11. Patients with type 2 diabetes on carbohydrate-restricted diets reduce and frequently eliminate medication. People with type 1 usually require lower insulin [64][65][66][67][68][69][70].
12. Intensive glucose lowering by dietary carbohydrate restriction has no side effects comparable to the effects of intensive pharmacologic treatment.The ACCORD (Action to Control Cardiovascular Disease in Diabetes) trial revealed the intensive-therapy group with more cardiovascular mortality and hypoglycemia compared with the standard-therapy group [71]. The low carbohydrate diet improves the glycemic control, reduces medications and decreases cost of pharmacy.
In Japan, Dr. Ebe at Takao Hospital in Kyoto finds the effectiveness of low carbohydrate diet after he has diabetes , does research on the nature of diabetes, and treats his patients with this special diet [72][73]. He promotes the low-carbohydrate diet to the Japanese medical system and society. He is well recognized and becomes a bestseller in Japan. Japanese restaurants follow to design low-carb menu.
In summary, the low-carbohydrate diet is gradually accepted in the guidelines of the ADA, and is also considered as the initial treatment for diabetes, but there is still no evidence of long-term ( more than one year ) effectiveness; the study is mostly limited to overweight or obese diabetic patients without data for the normal or low BMI diabetic patients in the real world; the best low carbohydrate restriction to take into account the effectiveness and durability is inconclusive; it is still lack of long-term effectiveness; the reduction of drug requirement should be taken into consideration of the effectiveness. For the long term diabetic control, too many drugs increase medical expenses and expose secondary drug failure risk. The easy principle of low carbohydrate diet is worthy to have further study to confirm the glycemic control in long-term.
In order to determine the long-term effect and safety of LCD on type 2 DM with a reasonable, effective and tolerable carbohydrate intake, a clinical trial will be designed and conducted for the effectiveness of glycemic control, reduction of polypharmacy, biochemical markers, body composition measurement, metabolic markers , functional assessment, atherosclerosis and quality of life.

Methods:
Type of study: a randomized controlled trial for 18 months. (Current evidence is

12-month)
First year: 1. Study design: a single-centered, parallel-designed, randomized controlled trial 2. Study population: Adults, 20-80 years of age, with type 2 DM will be recruited if they are diagnosed with diabetes ≥1 year ago, regardless of whether they receive treatment [oral hypoglycemic agents (OHA) and/or insulin treatment], and if they have a poorly controlled HbA1c ≥ 58 mmol/mol (7.5%) in the previous 3 months.
The patients will be identified for inclusion by physicians' referrals from outpatient clinics at a medical center and will be verified by a research assistant to conduct the screening. The exclusion criteria includes pregnant or lactating women, impaired renal function with a serum creatinine ≥ 132.6µmol/L (1.5 mg/dL), abnormal liver function (alanine aminotransferase (ALT), aspartate aminotransferase ≥3 times the normal upper limit) or liver cirrhosis, significant heart diseases (unstable angina, unstable heart failure), frequent gout attacks (≥3 times/year), participation in other weight-loss programs or the use of weight-loss drugs, eating disorders, and inability to complete the questionnaire .
3. Recruitment: The patients will be referred from outpatient clinics, health check-up, dietitian clinics, and poster announcement. We expect to recruit the patients within 4 months and do the baseline assessment.

Randomization:
The study population will be allocated sequence implementation using Taves covariate-adaptive randomization, stratified using sex and BMI (<24 and ≥24).If the members of the same family are recruited, they will be assigned in the same group. The study population will be assigned to the study group after they complete the baseline assessment.

Intervention:
For the LCD group, the daily carbohydrate intake will be limited to less than 90 g, without any restriction to total energy. For those with good dietary compliance, sulfonylurea and insulin injections will be reduced to half doses in advance to prevent hypoglycemia.
For the CRD group, the target total calorie intake will be calculated by multiplying the ideal body weight by 25 kcal/kg for those with a BMI between 18.5-24, 20 kcal/kg for obese subjects with a BMI >24 and 30 kcal/d for underweight subjects with a BMI <18.5. The macronutrient percentage will be 50-60% for carbohydrates, 1.0-1.2 g/kg for protein, and for fat it was ≤30%.
The medication for both groups will be adjusted every 6 months if HbA1c is more than 64 mmol/mol (8.0%) or lower than 48 mmol/mol (6.5%), with or without hypoglycemic symptoms. Exercise will be recommended for both groups and not a part of the intervention. There will be no limitation in medication for hypertension, hyperuricemia and aspirin for prevention of cardiovascular disease. The will be social connected with the "line" social network within groups to encourage the compliance.
The dietitian will educate every participant for 30 minutes and will follow up the participants at the 2 nd week and monthly for 3 months. The dietitian will follow up the patients every 3 months in person for 18 months and monthly telephone contact. The " line" social network within group provides sharing of appropriate food. According to a previous study, the estimated difference in the HbA1c reduction between the LCD and calorie-restricted diet (CRD) groups is 0.5%, with a standard deviation (SD) 0.408%. With a two-sided level of 5%, a power (1-ß) of 80%, and an assumed 20% loss to follow-up rate, the appropriate sample size was calculated to be 80 patients (20 patients in each group stratified by gender and BMI).
The estimation of sample size is 20 for a stratified cell. We will choose 30 if possible for a stratified cell because we use stratified random sample. According to the stratified layer by BMI and sex (BMI≦24 and >24; men and women). The total number will be 120, with control group and experimental group in parallel-designed. 9. Statistical analysis The analysis will be performed using an intention-to-treat analysis; baseline or last observations will be carried forward if the complete set of data for an individual is not available. The frequency or mean±SD will be presented for demographic data and single variant analysis. The repeated measures ANOVA wlll be conducted for different time group difference. The independent t tests or Wilcoxon rank tests will be used according to the normal distribution or not, for example, the different diet composition. A p-value <0.05 denoted a statistically significant difference.   (1) The first study of low carbohydrate diet for our country (2) Short-term effects of low carbohydrate diet on diabetes control