Effectiveness of a Ketogenic Diet in the Treatment of Type 2 Diabetes Mellitus

Health and Social Care Research: Methods and Methodology

Research Proposal

Word Count: 3009

1. Title

Effectiveness of a ketogenic diet in the treatment of type 2 diabetes mellitus: a systematic review

2. Background

Type 2 diabetes (T2DM) is a metabolic disorder characterized by continual hyperglycaemia and eventually causing a failure of insulin secretion by the pancreas (Packer and Castro, 2015). Diabetes affects around 382 million people worldwide (Diabetes UK, 2014). T2DM is strongly linked to obesity and physical activity (Packer and Castro, 2015). Patients with T2DM are at risk of complications such as renal impairment, retinopathy and peripheral never damage. They are also at high risk complications of stroke and cardiovascular diseases (Fowler, 2018).

KD is defined by elevated circulating ketone bodies, achieving a state of ketosis. This state is achieved by restricting carbohydrate to <50g per day and the predominant of calories coming from fats of 85% (Aragon et al., 2017). A study by Phinney et al. (1983) strongly emphasizes a diet low in protein is needed so that the reduction in carbohydrate calories are entirely compensated by fat. This puts the patient in a complete ketogenic state thus reducing blood glucose levels. A study conducted by Gumbiner et al. (1996) shows greater levels of circulating ketones when having a lower carbohydrate diet (24g) compared to higher carbohydrate diet which is very favourable in glycaemic control in diabetics. Haemoglobin A1C (HbA1c) provides a reliable measure of chronic hyperglycaemia and a good predictor of high risk complications (Sherwani et al., 2016). The diagnosis of T2DM is set by a cut-off point of 48 mmol/mol or more (6.5%) recommended by the expert advisory group (Farmer, 2012).

A number of studies have shown positive effects of KD on T2DM patients in reducing HbA1C levels (Boden et al., 2005; Scherger, 2017; McKenzie et al., 2017; Saslow et al., 2017; Hussain et al., 2012). Metformin, a popular and leading medication for the treatment of T2DM has similar effects of reducing HbA1c compared with the KD (Rojas and Gomes, 2013; Berstein, 2010). However, pharmacological interventions are often accompanied by reduced quality of life, economic burden and side effects (García-Pérez et al., 2013). This implies the need for alternative interventions without these negative effects. In the short term, rigorous lifestyle interventions tend to improve health outcomes for people living with diabetes. In a longer study participants following the KD did not present with any major side effects, in addition to better glycaemic control it also aided in significant improvements in both weight loss and metabolic parameters (Dashti et al., 2006). However more long-term studies are necessary to determine the possible adverse effects and the long-term acceptability of the diet.

Hypoglycaemia is the major side effect of diabetes medications in the treatment of T2DM. Hypoglycaemia can present with severe complications such as seizure, coma and even death. Prevalence of hypoglycaemia patients have increased due to more focus on tighter glucose control (Kalra et al., 2013).  A cross-sectional study of patients with T2DM treated with metformin and sulfonylurea found that patients had poorer medication adherence upon reporting a previous hypoglycaemic in fear of getting complications again. Non-adherence to medication can results in serious in health issues, with acute and chronic complications, more hospitalization and higher costs to healthcare (Lee et al., 2005). Moreover, different avenues of disease management would need to be considered because of increased T2DM prevalence and high costs of medication (Sokol et al., 2005).The removal of diabetes medications combined with clinically significant weight loss has been shown to generate health care cost savings (Cawley et al., 2014) .

In studies it was revealed that KD has improved glycaemic control in patients with T2DM such that medications were discontinued or even eliminated in participants (Westman et al., 2008; William et al. 2005; Boden et al., 2005; Nielsen and Joensson, 2008). This is evident that the KD is optimal in treating T2DM. Taking diabetes medications can inhibit ketosis and increase weight which is aimed for in the earliest phases of KD, therefore, reducing the effectiveness of being on the KD. However, it is evident through studies that being on the KD can override effects of diabetes medications enough for patients to come off them. Furthermore, patients following the KD should be under close medical supervision when adjusting dosages to prevent hypoglycaemia (William et al. 2005).

Obesity is a high risk factor in causing T2DM and can cause further cardio metabolic complications which are major causes of morbidity and mortality in T2DM (William et al. 2004). However, patients instructed to limit their carbohydrate intake to low levels have decreased levels of caloric intake. This overall reduced caloric intake may reduce their weight loss due to increased satiety (Johnstone et al., 2008). In previous research with non-diabetic patients have shown a positive correlation between level of ketonuria and weight loss success (Westman et al., 2002).  However, studies have shown that elevated ketosis improves glycaemia in T2DM patient’s independent of weight loss (Gumbiner et al. 1996; Westman et al. 2008).

In a study conducted by OK et al., (2018) it was found that patients had a greater compliance to the KD than a generalized diet. In order to stay compliant patients can inspect for ketosis by daily testing of their urine (Urbain and Bertz, 2016).   However, the induction of ketosis is influenced by the participant’s motivation to eat the meals found by Schmidt et al. using a single- arm pilot study. For participants to sustain motivation medical professionals should have an individualized approach to each patient maximising the effects of KD. Factors would be need to be considered such as age and socioeconomic status and having a shorter duration of follow ups (William et al. 2005).

Ketogenic diet being a favourable intervention compared to standard interventions makes the basis of this systematic review. It would be valuable to collate this firm evidence to support the need to bring in this intervention as there is limited research on this field. There are no reviews specifically on KD for the treatment of T2DM. The aim is addressed below:

The aim of this study is to evaluate the evidence to see whether the ketogenic diet can reduce HbA1c to the point that it can treat T2DM.

With the in-depth reading and evidence, I have moulded a question for this systematic review;

How effective is the ketogenic diet in reducing Hba1c compared to pharmacological interventions?

3. Methods

There is evidence of primary research and basic literature when conducting searches on Google scholar around the topic in question. Extraction and synthesise of data will allow a critical analysis of the studies. This methodological section will discuss a rigorous search design extracting and synthesising the most relevant studies that will be transparent and reproducible.  https://ebn.bmj.com/content/ebnurs/14/3/64.full.pdf

3.1   Search strategy

Evidence based medicine (EBM) often uses a specialized framework called PICO which stands for Patient, Intervention, Comparison and Outcome. This framework is outline by the Cochrane Collaboration (Higgins and Green 2011). Using search terms from table 1 using PICO framework enables an effective search strategy (Snowball et al, 1997). In table 2, there are compilation of a list of words that were identified that authors have used in their studies which includes synonyms, acronyms and rephrasing of key words. This increases the scope of studies for my inclusion and exclusion criteria which will be discussed below.

Table 1. PICO framework

Table 2. Key words derived from searches using elements of PICO framework (table 1)

Conducting a systematic review, I can use a variety of relevant databases (table 4) to get as many papers as possible using elements of the PICO framework using the Boolean operators. Of the studies identified in the search I will check for reference lists and citation search for relevant and included studies. I will use a reference management software namely Mendeley to save the search results in an organized manner and get rid of duplicate studies.

Table 3. Boolean operators; each search is assigned an identifier i.e S1

Table 4. Databases will be used for using elements by using the Boolean operators.

3.2 Eligibility criteria

Text word searches will be performed using the searches from table 3 combining the terms in the set using operator ‘OR’ and combining sets using  operator ‘AND’.  This will focus a search down to as many relevant articles as possible. This first screening method will be on the title and abstracts exclusively. If answered No to any of the 3 eligibility criteria than they will not be reviewed for full screening as seen on Figure 2. The studies that have answered ‘Yes’ to all questions will be tabulated under the heading ‘passed’. Upon retrieval of full manuscripts they will be analysed against the inclusion and exclusion criteria. This will be completed by myself and a supervisor. We will have an in-detail discussion about our selection and settle any discrepancies. We will assess agreement using the agreement percentage;

$\frac{\mathrm{Agreements}}{\mathrm{Agreements}+\mathrm{Disagreements}}\mathit{ x }100$