1. Introduction

The continuum of the degree of carbohydrate restriction that exists in contemporary clinical

practice is illustrated in Table 1^[1] [23], with general agreement that less than 20 g per day is a “very low”

carbohydrate intake (though some use a threshold of less than 50 g), equivalent to about 10% of total

energy intake. The threshold for “low” carbohydrate intake is usually accepted as less than 130 g/day,

equivalent to less than 26% of total energy from carbohydrates. Consumption greater than 230 g per

day is consistent with no restriction of carbohydrate, although we find that many of our patients with

diabetes or obesity exceed several times this amount on a daily basis. In general, the greater the degree

of carbohydrate restriction, the greater the degree of ketogenesis, such that carbohydrate intakes of

more than 50 g per day are not usually sufficient for ketogenesis^[2] [10]. Hence, “low-carbohydrate” and

“ketogenic” are not synonymous dietary terms, but do overlap.

2. Guidelines Endorse Low-Carbohydrate and VLCKD Diets.

8.1. Guidelines Endorse Low-Carbohydrate and VLCKD Diets.

Doctors, nurses and other health care professionals ought to be aware that the use of

low-carbohydrate and VLCKD diets in patients with obesity or type 2 diabetes is in fact supported by

several sets of guidelines from international bodies and professional groups, as outlined in Table 2.

For example, the most recent guidance from the American Diabetes Association is unequivocal in

stating that “Reducing overall carbohydrate intake for individuals with diabetes has demonstrated

the most evidence for improving glycemia and may be applied in a variety of eating patterns that

meet individual needs and preferences”.^[3] [86]. In the UK, pragmatic infographic resources based on

the glycaemic load of various foods are available from the National Institute for Health and Care

Excellence (NICE) (created by one of the authors (D.U.)^[4] [92]. See Figure 2 for an example. These help

patients understand the glycemic “consequences” of their dietary choices. For example, a 150 g bowl of

boiled rice has approximately an equivalent impact on blood glucose levels as ten standard teaspoons

of table sugar.

For example, the most recent guidance from the American Diabetes Association is unequivocal in

stating that “Reducing overall carbohydrate intake for individuals with diabetes has demonstrated

the most evidence for improving glycemia and may be applied in a variety of eating patterns that

meet individual needs and preferences.”^[3] [86]. 

3. Early De-Prescription is Important

8.2. Early De-Prescription is Important

Although research data are relatively scarce on optimal patterns of medication usage early in

low-carbohydrate and ketogenic diets, we have found that early and intensive de-prescribing is often

required, particularly in patients with diabetes^[5] [93]. In particular, rapid titration of insulin is

obviously important in order to prevent potentially serious hypoglycemia. (This clearly applies only

to patients with an established diagnosis of insulin requiring type 2 diabetes, as opposed to type 1

diabetes: Low-carbohydrate diets have been shown to reduce adverse events and improve control in

observational studies in patients with type 1 diabetes^[6] [94], but we have not considered this further

here.) In general, we tend to stop all fast-acting insulin at the time of initiation of VLCKD and, if not

stopping basal insulin completely, by then reducing the dose by between 50% and 80%. This

mandates four-times-daily monitoring of capillary blood glucose levels in the hours and days after

significant decreases in carbohydrate intake. We have found that it is essential that these patients

have immediate access to a diabetes nurse, primary care doctor, consultant or dietitian with

experience of low-carbohydrate diets during this time. In addition, we tend to stop sulphonylurea

drugs completely at initiation of the diet because of the risk of hypoglycemia. Conversely, we tend to

continue metformin given its insulin-sensitizing effects, cardiovascular benefits. and very low risk of

hypoglycemia. We take an individualized approach to titrating gliptins or glitazones, informed by

baseline HbA1c and patient preference. We often continue glucagon-like peptide-1 (GLP1) receptor

agonists. Given the potential risk of euglycemic diabetic ketoacidosis^[7] [95] in patients taking sodiumglucose cotransporter-2 (SGLT2) inhibitor drugs, we always stop these if the diet is initiated.

4. Calorie Counting Is Not Required

8.3. Calorie Counting Is Not Required

Rather than emphasizing the need for patients to quantify their calorie intake, we ask them to

focus on eating to comfortable satiation and then stopping. The effectiveness of calorie counting

has been questioned^[8][9] [97,98] with well-described physiological “recidivism” with this approach^[10] [99].

We take a more mechanistically intuitive approach, emphasizing to the patient that metabolic changes

associated with their reduced carbohydrate intake may adjust their “hunger set-point” as outlined

above. We often try to back this engagement and education up with graphical aids, emphasizing

the physiological mechanisms underlying insulin resistance and their reversal with carbohydrate

restriction, as shown in Figure 3. Our anecdotal experience of patients reporting significantly reduced

hunger and increased satiety is consistent with studies on higher fat and protein diet influence on the

physiological drivers of feeding behavior^[11][12] [100,101].

The second group of drugs that need consideration during a low-carbohydrate diet is

antihypertensive medications. This is because the higher circulating insulin levels in insulin resistant

type 2 diabetes patients can cause renal sodium retention, which may be reversed quickly with a

reduction in insulin levels (as part of a low-carbohydrate diet), leading to enhanced renal sodium

(and water) excretion and a lower blood pressure^[13] [96]. We (D.U.) have described these changes in a

cohort of 128 patients with type 2 diabetes on a low-carbohydrate diet for an average of two years^[14]

[75], where there was a reduction in systolic and diastolic blood pressure of 10.9 and 6.3 mmHg,

respectively, despite a 20% reduction in anti-hypertensive medication usage. The risk of hypotension

mandates cautious and frequent monitoring of patients’ blood pressure and vigilance for symptoms

of postural hypotension.

5. Monitor Cardiovascular Risk Factors

8.4. Monitor Cardiovascular Risk Factors

While all patients with obesity and type 2 diabetes should have cardiovascular risk factors

monitored periodically, the potential increase in fat consumption that arises on a low-carbohydrate diet,

in the context of historical epidemiological concerns that dietary fat might increase cardiovascular risk,

makes the issue more pertinent. We have found that individuals concerned about the appropriateness

of low-carbohydrate diets are either unfamiliar or don’t accept recent nutritional epidemiological

discoveries around the ineffectiveness of low fat diets to prevent cardiovascular disease^[15] [29] and the

benefits of low-carbohydrate intake^[16] [102] and certain saturated fats, such as those from dairy, in reducing

cardiovascular (and diabetes) risk^[17][18] [103,104]. A recent meta-analysis suggested low-carbohydrate

diets are superior to low fat diets in improving the lipid profile^[19] [105]. While current guidelines on

saturated fat are overdue a revision^[20][21] [106,107], it seems reasonable to reassure patients undertaking

a low-carbohydrate diet, and their health care providers, that saturated fats from foods that are not

ultra-processed are unlikely to do them harm, especially if they are losing weight and improving

glycaemia while undertaking a low-carbohydrate approach. A second consideration is the increase

in LDL-cholesterol that is described with some^[22] [27] but not all^[23] [108] low-carbohydrate interventions,

but the fact that this appears limited to the large LDL subfraction^[24] [76] suggests it is unclear whether it

increases cardiovascular risk. Nonetheless, we routinely measure blood pressure, lipid profile and

HbA1c in patients adhering to a low-carbohydrate or VLCKD and treat abnormal findings as we would

in routine clinical practice, where they have not improved over time.

6. Ensure Adequate Fiber Intake

8.5. Ensure Adequate Fiber Intake

While the recommended intake of dietary fiber is 30 g per day (in the UK) the average intake is

closer to 18 g per day. The reduction in wholegrain consumption associated with a low-carbohydrate

diet could accentuate that deficit and low intake of dietary fiber is associated with an increased risk

of metabolic^[25] [109] and colonic^[26] [110] disease. However, we have found in practice that adopting

a low-carbohydrate diet which limits ultra-processed foods and includes nuts, seeds, non-starchy

vegetables and low-carbohydrate fruits tends to lead to a net gain rather than a reduction in patients’

dietary fiber intake compared to baseline.

Table 2. Summary of current guidelines and consensus statements on the use of low-carbohydrate diets.

Table 1. Summary of current guidelines and consensus statements on the use of low-carbohydrate diets.