NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
National Clinical Guideline Centre (UK). Obesity: Identification, Assessment and Management of Overweight and Obesity in Children, Young People and Adults: Partial Update of CG43. London: National Institute for Health and Care Excellence (NICE); 2014 Nov. (NICE Clinical Guidelines, No. 189.)
Obesity: Identification, Assessment and Management of Overweight and Obesity in Children, Young People and Adults: Partial Update of CG43.
Show details6.1. Introduction
Obesity is a serious public health issue and the most common interventions are diet and exercise. The use of very-low-calorie diets (VLCDs) is sometimes considered for weight management in the NHS and in commercial programmes. There is a need for long-term comparison with conventional dietary interventions to assess clinical effectiveness. VLCDs are defined as hypocaloric diets which provide between 450 to 800 kcal per day and are relatively enriched in protein of high biological value. They must contain the full complement of vitamins, minerals, electrolytes and fatty acids. They are usually in a liquid formulation and are intended to completely replace other food intake in a weight loss programme for a specific period of time.
NICE CG43 (2006) reviewed and approved the short-term use of VLCDs in treatment of the obese person. The provision of very-low-calorie diets on the NHS is an emerging intervention option, and often used by people with a high BMI, despite a lack of published literature on the use of VLCDs in this group. However, it is noted that a large number of individuals purchase these limited calorie total diet replacements from commercial providers.
CG43 recommended the use of VLCDs in people who are obese and have reached a plateau in weight loss. The care and management of these people is likely to be delivered within specialist services. Moreover, the GDG recognise that attendance at NHS care will vary but that it is likely that people with comorbidity would attend at a GP surgery. It is important to assess whether the provision of VLCDs in this population is of added benefit compared to the usual care provided at this level.
The GDG wished to determine the long-term efficacy of VLCDs as it is currently unknown. The potential increased use of VLCD in the NHS requires evidence of improved patient outcomes without compromising patient safety and quality of care. This includes assessing evidence of reduction in comorbidities with long-term maintenance of weight loss and the impact on the patient's quality of life.
To this end the GDG posed 3 questions that aimed to provide the evidence for effectiveness and safety of VLCDs as well as establishing what maintenance strategies maximised weight loss in the long-term. Each of these questions and the evidence are presented in this chapter. The GDG interpretation of the body of evidence considered, together with recommendations made, conclude the chapter. This guidance amended the definition of VLCD to ≤ 800 calories per day in line with current practice, whereas the previous guideline had used a definition of 1000 calories or less. The search for this review question was repeated without any date cut-offs to ensure that no relevant evidence was missed.
6.2. Review question: In people who are overweight or obese, what is the clinical and cost effectiveness of very-low-calorie diets in reducing weight?
For full details see review protocol in Appendix C.
Table 7
PICO characteristics of review question.
6.2.1. Clinical evidence
Seven studies were included in the review.51,63-65,67,68,70 Evidence from these studies is summarised in the clinical evidence summary below (Table 8). See also the study selection flow chart in Appendix D, study evidence tables in Appendix G, forest plots in Appendix I, GRADE tables in Appendix O and excluded studies list in Appendix J.
Table 8
Summary of studies included in the review.
Three studies reported on specific populations: 1 with all male participants in an occupational setting 51 and 2 with all female participants.64,65 Furthermore, 3 studies reported on people with type 2 diabetes. 63,67,70
All included studies compared VLCD to standard dietary advice, but the length of VLCD (8-50 weeks), the number of calories included (400-800 kcals), and the length of treatment and follow-up (6 months-24 months), varied. Three papers63,64,68 reported intermittent VLCD. Six studies 51,64,65,67,68,70 included behavioural therapy (or educational sessions) as part of the treatment in both arms and 4 of these studies included an exercise component as well.51,64,65,70
No studies reported the weight change outcomes in the GDG preferred units of percentage loss of initial weight. However, the studies did report BMI or kg change so this was used instead. One study did report percentage ‘ideal’ weight loss, which was included.51
None of the studies reported weight change (BMI or kg) from end of VLCD to end of maintenance period. However, studies reported weight change from start of study to end of VLCD and from start of study to end of maintenance period which indirectly provided the same information, so this was reported instead.
Numbers of participants withdrawing from the study was extracted for all 7 studies. None of the included studies reported on the following outcomes from the protocol: health -related quality of life and improvement in physical activity.
Table 9
Summary of studies included in the review: weight in kg.
Table 10
Summary of studies included in the review: withdrawals.
Table 11
Clinical evidence profile: VLCD versus standard dietary advice for overweight and obese people.
6.2.2. Economic evidence
6.2.2.1. Published literature
No relevant economic evaluations were identified.
See also the economic article selection flow chart in Appendix E.
6.2.2.2. Unit costs
Relevant unit costs are provided below to aid consideration of cost effectiveness.
The sections below detail the costs borne by the NHS for providing a VLCD diet. These costs have been calculated in conjunction with GDG members who have experience in running VLCDs services within NHS weight loss management programmes and therefore reflect what is seen in current NHS led VLCD programmes. Note that the level of care provided as part of a VLCD varies across the UK, therefore the following costs are intended to provide only an example of the costs faced; local costs will vary. This analysis focuses on NHS-run VLCDs, though consideration is given to commercially run VLCDs. Two cost analyses were conducted:
- VCLDs that individuals with and without comorbidities with a BMI over 40 kg/m2 may receive.
- VLCDs that individuals with no comorbidities and a BMI 30 - 40 kg/m2 may receive.
Tier 3 service VLCD costs (individuals with and without comorbidities with a BMI over 40 kg/m2)
The following represents the costs per individual for undertaking a VLCD for 1 year. Where appropriate, separate costs are displayed for individuals with comorbidities, such as hypertension and type-2 diabetes. These individuals are likely to receive a different level of care to ensure their comorbidities are properly monitored and controlled.
Initial medical assessment
When an individual first undertakes a VLCD they undergo an initial assessment in which their suitability is assessed, and they receive advice on how to undertake the diet properly. The individual is also discussed at a separate multidisciplinary-team (MDT) meeting which comprises of medical professionals who will be involved in the VLCD. The costs of these assessments are shown in Table 12.
Table 12
Initial medical assessment.
Undertaking the diet
After the initial medical assessment the individual undergoes the diet for 12 weeks. During this period they are likely to receive intense monitoring to identify any medical or compliance issues. Individuals with comorbidities are likely to receive additional monitoring from different healthcare professionals to ensure their comorbidities are being correctly treated. They may also receive additional telephone support from a clinical nurse specialist. Monitoring costs for individuals with no comorbidities are shown in Table 13, and monitoring costs for individuals with comorbidities are shown in Table 14.
Table 13
Costs of undertaking the diet for individuals with no comorbidities.
Table 14
Costs of undertaking the diet for individuals with comorbidities.
End of programme
After the diet has been completed the individual will have their blood tested again and undertake a final medical assessment to discuss a follow-up plan. It is assumed that if the individual has comorbidities then this final assessment will take place with a medical consultant, the costs of which are shown in Table 15. If the individual has no comorbidities then the final assessment takes place with a dietitian, the costs of which are shown in Table 16.
Table 15
Final medical assessment for individuals with comorbidities.
Table 16
Final medical assessment for individuals with no comorbidities.
Follow-up
It is assumed that after the diet the individual will continue to receive regular monitoring, in the form of monthly follow-up visits with a dietitian, to ensure weight is maintained and any medical issues are identified and addressed. These costs are shown in Table 17.
Table 17
Follow-up.
Additional costs
During the time the VLCD is being undertaken there are additional pressures placed on NHS services to accommodate people undertaking a VLCD. This includes additional time spent by an administrator and longer MDT meetings. Individuals on a VLCD will be seen more and therefore more time will be spent by an administrator making appointments, entering additional information into a database and sending more letters to the individual's GP. MDT meetings are weekly meetings which are used to discuss the progress and issues of patients on the service, and individuals who are undertaking a VLCD will require more time to discuss any safety or compliance concerns. Individuals with comorbidities will have an even longer MDT discussion to discuss issues related to medication titration for example. The additional costs for patients with no comorbidities are shown in Table 18, and the additional costs for patients with comorbidities are shown in Table 19. Note that transport costs have been excluded from this analysis as most services do not cover transport costs and this cost is not specific to VLCDs.
Table 18
Additional costs borne by the centre for individuals with no comorbidities.
Table 19
Additional costs borne by the centre for individuals with comorbidities.
The comparator
In this analysis VLCDs are compared to a tier 3 weight management service. This decision was made given that:
- The majority of people placed on VLCDs in the NHS are those already on a tier 3 weight management service
- These individuals have a BMI over 40 kg/m2, therefore they are likely to receive intensive treatment.
Some of the costs detailed above would be incurred as part of a tier 3 weight management service, even if the patient had not gone on the VLCD. The cost components that are unique to a VLCD are identified in the following section.
Incremental costs
The incremental cost of providing a VLCD in comparison to a tier 3 weight management service is calculated in Table 20. This assumes:
Table 20
Incremental costs for undertaking a VLCD.
- After 1 year people who underwent a VLCD will receive the same standard of care as those on a tier 3 service.
- People undertaking a VLCD will also receive the level of care provided within a tier 3 service, for example the same number of physiotherapist and psychologist visits.
Sensitivity analysis
To reflect the differential level of care patients may receive, sensitivity analyses were run on the costs by considering the following additional scenarios.
Firstly, in some centres, an MDT may not be run to discuss individuals with no comorbidities on a VLCD. Instead, medical support may be offered to these patients if they choose to access it. An assumption was made, based on GDG opinion, that 20% of individuals would access this additional medical support. The cost of this additional support is calculated in Table 21, and would be instead of the additional MDT cost calculated in Table 18 for people with no comorbidities. For people with comorbidities it was assumed in this sensitivity analysis that the MDT would last the same amount of time as in a tier 3 service.
Table 21
Cost of additional medical support.
Secondly, in some centres, blood tests may also be taken for people as part of the tier 3 service, and therefore do not represent an incremental cost.
Table 22
Incremental costs for undertaking a VLCD (low end estimate).
Primary care VLCD costs (individuals without comorbidities with a BMI 30 – 40 kg/m2)
The GDG also recognised that a small group of individuals with a BMI between 30 – 40 kg/m2 with no comorbidities may receive a much less intensive care package during their VLCD. These individuals would only receive an initial medical assessment from a dietitian, then weekly monitoring visits during the time the individual is on the diet. They would have access to medical support, provided by a GP, and an assumption was made that only 20% of patients would access this service. The costs of this service are provided in Table 23.
Table 23
Costs of reduced care package VLCD in people with no comorbidities.
As these people will not require intense treatment the relevant comparator is standard dietary advice. This consists of visits with a dietitian the costs of which are provided in Table 24.
Table 24
Cost of standard dietary advice.
Therefore the incremental costs of providing a VLCD for these people is £158.
Commercially run VLCD
A number of people choose to fund their own VLCD and complete this outside of NHS care. Even though the individual pays for the service, there are still costs borne by the NHS. These costs are mainly relevant to individuals with comorbidities and are related to medical assessments that are required by the commercial providers before and during the diet. The NHS will incur the cost of a professional assessment to ensure the VLCD is safe to complete and also the cost professional monitoring via telephone to alter medication and ensure no complications arise. Finally, the costs of any adverse effects that arise from undertaking the VLCD will also fall on the NHS.
Economic considerations
Using the incremental cost of providing a VLCD, the QALY increase which would be required for VLCDs to be considered cost effective at a £20,000 per QALY threshold can be calculated like so:
The change in BMI required to achieve this increase in QALYs can then be calculated using data on the health related quality of life (HRQoL) increase expected from a 1 point reduction in BMI:
Evidence from the literature17,34 suggests that HRQoL increases between 0.0079- 0.0164 per unit decrease in BMI. Therefore if a unit decrease in BMI was sustained for a year this would mean an increase of 0.0079 – 0.0164 QALYs. It is worth noting that the estimate of 0.0164, derived from Dixon et al,17 is unadjusted for age and whether or not the individual has type-1 or type-2 diabetes. The estimate of 0.0079, derived from Lee et al,34 is for non-diabetics and adjusted for age. Therefore we would expect the true value to be closer to 0.0079 for patients without diabetes.
Using the information above, the BMI changes needed for VLCDs to be cost effective at a £20,000 per QALY threshold are displayed in Table 25 taking into account the range of costs and quality of life values.
Table 25
BMI change needed for VLCD to be cost effective at a £20,000 per QALY threshold (no comorbidities).
Literature suggests that the HRQoL improvements associated with BMI reductions may be higher for individuals with type-2 diabetes (T2D).34 Using the T2D specific estimate detailed in Lee et al,34 Table 26 details the BMI change needed for VLCDs to be cost effective at a £20,000 per QALY threshold for individuals with T2D.
Table 26
BMI change needed for VLCD to be cost effective at a £20,000 per QALY threshold (people with T2D).
Other considerations
A study by Lean et al32 calculated the costs of running a low-energy-liquid-diet of 810–833 kcal/day and found the cost to the NHS to be £861 per patient. To be comparable with the above analysis, by excluding the costs of the formula diet, Orlistat and the costs to Counterweight, this cost falls to £266 per patient. However this cost is averaged across all individuals that entered the study. Due to dropout rates the cost of an individual that completed the diet would likely be higher. As the study's exclusion criteria is fairly strict the individuals who completed the diet in the study are unlikely to represent those seen in a tier 3 service, therefore these costs are more comparable to the costs calculated for primary care run VLCDs above. It is worth noting that that the healthcare staff participating in the study were already familiar with the counterweight programme in Scotland. This means their skills would not be representative of current skill levels in primary care in the rest of the UK. This would affect the amount of time staff would spend with the individual on the VLCD and therefore the cost of the intervention. Finally although the calorie intake just falls above this guideline's definition of 800 kcal/day for a VLCD the GDG noted that the costs could still be relevant. With all things considered, this study suggests that the £236 cost calculated in Table 23 could be seen as an underestimate of the true cost of the intervention to the NHS.
Secondly, this analysis only considers the costs associated with the VLCD, and only HRQoL changes associated with weight change. In reality the probability of a person experiencing comorbidities is likely to decrease as they lose weight. By reducing these comorbidities the individual will achieve a higher HRQoL, for example weight loss may improve osteoarthritis related knee pain or improve obesity related sleep apnoea, and costs to the NHS will fall. For example reducing BMI results in lower prescriptions costs as outlined in a study by the counterweight project team.10 The study showed that unit BMI decreases from 40 kg/m2 can reduce prescription costs by £5 - £8 per year.
Thirdly, in this analysis, the HRQoL increases per unit change in BMI reflect an assumed linear relationship between BMI and HRQoL. A study by Hunger et al22 shows that HRQoL is sensitive to changes in BMI between 30 – 40 kg/m2 however is fairly unresponsive to changes between a BMI of 40 – 45 kg/m2. If the weight change from a VLCD was the same for all BMI values then this analysis would suggest that they are less likely to be cost effective for individuals with a BMI above 40 kg/m2.
Finally the above analysis does not take into account any adverse effects that could arise from a VLCD. These could reduce HRQoL and increase costs to the NHS.
6.2.3. Evidence statements
6.2.3.1. Clinical
- Low to very low quality evidence showed that there may be no clinical difference between VLCD and standard dietary advice in:
- % ideal weight loss (1 study, n=110)
- Withdrawals (7 studies, n=487)
- Weight in BMI change from start of study to end of weight maintenance period (1 studies, n=33)
- Weight in kg change from start of study to end of weight maintenance period (7 studies, n=373)
- However, low to very low evidence showed that there may be a clinical benefit for VLCD in:
- Weight in kg change from start of study to end of VLCD period (5 studies, n=265)
- Weight in BMI change from start of study to end of VLCD period (2 studies, n=112)
- No evidence was found for health related quality of life outcome.
6.2.3.2. Economic
- No relevant economic evaluations were identified.
- An original comparative cost analysis showed that:
- For obese individuals with comorbidities:
- Using the lowest cost estimate for VLCDs and the highest quality of life estimate for unit decreases in BMI, a VLCD would need to reduce BMI by 2.31 kg/m2 and sustain this for 1 year, to be considered cost effective at a £20,000 per QALY threshold.
- Using the highest cost estimate for VLCDs and the lowest quality of life estimate for unit decreases in BMI, a VLCD would need to reduce BMI by 7.84 kg/m2 and sustain this for 1 year, to be considered cost effective at a £20,000 per QALY threshold.
- For obese individuals without comorbidities:
- Using the lowest cost estimate for VLCDs and the highest quality of life estimate for unit decreases in BMI, a VLCD would need to reduce BMI by 0.48 kg/m2 and sustain this for 1 year, to be considered cost effective at a £20,000 per QALY threshold.
- Using the highest cost estimate for VLCDs and the lowest quality of life estimate for unit decreases in BMI, a VLCD would need to reduce BMI by 6.29 kg/m2 and sustain this for 1 year, to be considered cost effective at a £20,000 per QALY threshold.
- This analysis was considered directly applicable with potentially serious limitations.
6.2.4. Review question: In people who are overweight or obese, what is the safety of very-low-calorie diets when used to reduce weight and maintain weight loss?
For full details see review protocol in Appendix C.
Table 27
PICO characteristics of review question.
6.2.5. Clinical evidence
Six studies were included in the review.4,8,9,20,56,65,66,69 These are summarised in Table 28 below. Evidence from these studies is summarised in the clinical evidence summary below (Table 29). See also the study selection flow chart in Appendix D, study evidence tables in Appendix G, forest plots in Appendix I, GRADE tables in Appendix O and excluded studies list in Appendix J.
Table 28
Summary of studies included in the review.
Table 29
Clinical evidence summary: VLCD versus LCD (both with or without behavioural and/or exercise therapy) for overweight and obese people.
Mean BMI ranged between 31.9 and 39 kg/m2 in most studies4,8,9,20,56,65,69 but 1 study 66 had mean BMI of 40 and 44 kg/m2 in VLCD and LCD arms, respectively. Two studies reported on specific populations: one on participants with type 2 diabetes,69 and another on participants with osteoarthritis.8,9,56
All included studies compared VLCD to LCD but the length of VLCD varied (from 8 to 12 weeks) as well as the number of calories included (ranging from 400-520 kcal/d). The range of calories included for a LCD varied from 810 kcal/d to 1500 kcal in 1 study. The studies also varied in whether or not the VLCD treatment period included some time on a LCD (before or after the VLCD).
Furthermore, 4 8,9,56,65,66,69 studies included behavioural therapy as part of the treatment in both arms and 2 of these studies included exercise as well.
One study reported on binge eating scores (with the Binge Eating Scale), 1 aspect of disordered eating scale 65. However, no studies reported on other aspects of disordered eating such as night eating syndrome and bulimia nervosa.
Three studies65,66,69 reported final depression scores on Beck's Depression Inventory at different time points ranging from 4 months to 1 year. Since depression scores were expected to be different sooner after treatment when patients have achieved the most weight loss (18 weeks as in one study) than at 1 year when they may have gained back some weight (as in another study), results were presented separately for different time points.
One study8,9,56 reported the proportion of people with depressive tendencies at the end of the study but it was not clear how many of these individuals had depressive tendencies at the start of the study.
One study each reported on gallstones20, uric acid levels (gout)4, diarrhoea8,9,56, and constipation8,9,56 (an additional study 69 reported levels of uric acid but only in the VLCD group).
However, none of the included studies reported comparative data from the following outcomes from the protocol: postural hypotension, bone mineral density, or hypoglycaemia.
6.2.6. Economic evidence
See review on the effectiveness of VLCD for economic evidence.
6.2.7. Evidence statements
6.2.7.1. Clinical
- VLCDs may result in more binge eating after 1 year compared to LCDs (1 study, n=45, very low quality).
- It is unclear if depression is worse after VLCDs than after LCDs (3 studies, n=100, very low quality).
- There were more participants with ‘depressive tendencies’ after VLCDs than after LCDs (1 study, n=192, very low quality).
- Constipation (1 study, n=192, very low quality) and diarrhoea (1 study, n=192, very low quality) may be more frequent after VLCDs than LCDs.
- Gallstones occurred in some people during or after VLCDs but not with LCDs; however this was only symptomatic in 1 individual (1 study, n=13, low quality).
- There were similar serum uric acid levels in VLCDs and LCDs; however, more participants with VLCDs had ‘marked’ increases in serum uric acid at some point during treatment but these were not correlated with episodes of gout (1 study, n=45, very low quality).
- There was no evidence on postural hypotension, bone density, or hypoglycaemia.
6.2.7.2. Economic
- No relevant economic evaluations were identified.
6.2.8. Review question: What are effective management strategies for maintaining weight loss after very-low-calorie diets in people who are overweight or obese?
For full details see review protocol in Appendix C.
Table 30
PICO characteristics of review question.
A minimum study duration of 1 year was required for study inclusion because that was considered necessary to assess long-term maintenance of weight loss.
6.2.9. Methods summary
The protocol included a range of interventions (behavioural therapy, anti-obesity drugs, diet), which should not be combined in a meta-analysis and were reported separately. The protocol also treated separately comparisons with no treatment and placebo.
The various interventions were clustered into 6 main comparisons:
- behavioural therapy and different re-feeding techniques [Agras 1996]
- hypocaloric diet (1600 kcal/day), with additional 238 kcal VLCD diet (total 1838 kcal/day) [Ryttig 1997] or including VLCD sachet (total 1600 kcal/day) [Ryttig 1995]
- dietary counselling with exercise versus dietary counselling only [Borg 2002, Fogelholm 2000]
- Orlistat with or without dietary and lifestyle counselling versus dietary and lifestyle counselling or meal replacement [LeCheminant 2005, Richelsen 2007]; and, Orlistat with dietary and lifestyle counselling versus dietary and lifestyle counselling only [Richelsen 2007]
- high protein diet versus high carbohydrate diet [Delbridge 2009]
- interventions compared to no treatment: Sertraline versus placebo [Wadden 1995], high protein diet versus no treatment [Lejeune 2005], fibre diet versus no treatment [Pasman 1997]
The duration of VLCD varied from 1 to 6 months and this may influence the effectiveness of maintenance regimens. Similarly, the length of maintenance regimen and follow-up also varied (1 year was inclusion criteria, maintenance ranged up to 3 years).
The protocol specified the outcomes, % weight change (kg) from end of VLCD to end of study and % change in body mass index (BMI) from end of VLCD (called ‘baseline’) to end of study. The GDG considered as alternative outcomes, the weight at the end of the study or the weight change from before VLCD to the end of study or from the end of VLCD to the end of study. The GDG considered the outcomes in relation to the weight at baseline (that is, at the end of VLCD), end of treatment and end of follow-up. Another protocol outcome was withdrawals at baseline, end of treatment and end of follow-up. It was important to distinguish between withdrawals due to maintenance strategy or due to other circumstances.
In addition to the forest plots for each individual pairwise comparison, 2 forest plots summarising all comparisons in the literature were presented to the GDG: 1 including all head-to-head trials and 1 including all trials with interventions against either no treatment or no placebo.
6.2.10. Clinical evidence
10 papers were included in the review.2,5,13,19,33,35,50,55,58,59
4 studies reported on sub populations: 3 with all female participants (Agras 1996, Fogelholm 2000, and Pasman 1997) and 1 with male participants (Borg 2002). Furthermore, 5 studies did not have a minimum weight loss for inclusion in the maintenance phase of the study. There were differences in weight (kg) at baseline between treatment arms; for example, Ryttig 1995 reported a 12 kg difference between the 2 separate intervention arms.
Only 1 study reported the protocol % change outcomes,50 and the rest reported the alternative outcomes. Most studies reported withdrawals from maintenance; however, it was often unclear whether they were withdrawals due to maintenance strategy or due to other circumstances. None of the included studies reported on health-related quality of life. Two studies reported physical activity as an outcome 5,19 but as both studies were exercise interventions and this outcome was only reported during the study, the results are more about adherence to the exercise regimen rather than improved physical activity after the maintenance programme had completed.
Table 31
Summary of studies included in the review.
Table 32
Summary of studies included in the review: weight in kg.
Table 33
Summary of studies included in the review: withdrawals.
Table 34
Clinical evidence summary: head-to-head trials.
Table 35
Clinical evidence summary: intervention versus no treatment.
6.2.11. Economic evidence
See review on the effectiveness of very-low-calorie diets for economic evidence.
6.2.12. Evidence statements
6.2.12.1. Clinical
- Behavioural therapy and re-feeding: maintenance strategy did not result in clinical benefit, with greater weight regain (1 study, n= 201, very low quality)
- Orlistat: when compared with both meal replacement and dietary and lifestyle counselling only, the use of orlistat resulted in less weight regain (2 studies, n= 540, moderate to very low quality)
- Dietary counselling with or without exercise: dietary counselling, plus exercise (100 kcal or 200 kcal burn) was a more effective weight loss maintenance strategy than dietary counselling alone (2 studies, n= 175, very low quality)
- Diet, including:
- Hypocaloric and meal replacement: when compared with a hypocaloric diet only, hypocaloric diet which included a 220 kcal VLCD diet resulted in greater weight loss over the course of the weight maintenance phase; meal replacement (hypocaloric diet, with an additional 238 kcal VLCD) when compared to hypocaloric diet only, resulted in less weight regain also but this was not considered to be of clinical benefit (2 studies, n= 195, low to low quality)
- High fibre versus no treatment: more weight regain during the high fibre maintenance strategy, when compared to no treatment (1 study, n= 48, very low quality)
- High protein versus no treatment: less weight regain during the high protein maintenance strategy, when compared to no treatment (1 study, n= 140, low quality)
- High protein versus high carbohydrate: high protein weight maintenance strategy resulted in great weight loss than the high carbohydrate weight maintenance strategy(1 study, n= 180, very low quality)
- There was no evidence on ‘health related quality of life’ and ‘improvement in physical activity’ outcomes.
6.2.12.2. Economic
- No relevant economic evaluations were identified.
6.2.13. Recommendations and link to evidence
| Recommendations |
|
|---|---|
| Relative values of different outcomes | VLCD effectiveness When considering the effectiveness of VLCDs, the critical outcomes which the GDG wished to examine in the literature were % weight change (kg) at end of maintenance period, health related quality of life and withdrawals (to assess adherence). VLCD safety The GDG considered that, for the safety of VLCDs, the most critical outcomes which should be examined in the evidence were disordered eating, depression and postural hypotension. They wished to examine the evidence on disordered eating because they were particularly concerned from their clinical experience that the dramatic calorie reduction in diet with VLCD, even though for a short period, may create or worsen pre-existing unhealthy eating patterns, or disordered eating, such as binge eating, bulimia nervosa, or night eating syndrome. They were also particularly concerned to examine the evidence on depression because they noted that depression may be associated with increased weight gain and obesity and may be particularly exacerbated where weight loss and regain occurs over a shorter time period and through more radical changes and adjustments to eating habits such as those involved in the use of VLCD's and following a VLCD. Postural hypotension, if it occurred, was considered to have a significant impact upon an individual's life. While it was noted that a proper VLCD with vitamin/mineral supplementation may actually improve a person's vitamin D and iron levels, the GDG identified other important outcomes which may cause a number of problems which they chose to search the evidence for. These were reduction in bone density, constipation, gallstones, gout, diarrhoea, and hypoglycaemia. The GDG felt that reduction in bone density may occur because of insufficient calcium in the diet and may result in fragility and fracture. Constipation, which may be caused by insufficient fibre in the diet, is particularly troubling for individuals and they wished to see if there was evidence to substantiate these concerns. They also wished to see the evidence on the occurrence of gallstones (the result of inactivity of the gallbladder from low fat in VLCD diets) and gout (associated with rapid weight loss) after VLCDs as they considered them to be very painful experiences for people. As gout will be rarely reported, raised serum uric acid levels, which are associated with gout, were felt to be an appropriate surrogate outcome of gout. Diarrhoea, which the GDG thought might occur for several factors including undiagnosed lactose intolerance, was felt to be an important outcome to examine in the literature as it has a significant impact on patients. Hypoglycaemia, while rare, is a serious adverse event that may result from lower energy intake without concomitant review of diabetic medication so they also wanted to examine the incidence in the literature. VLCD maintenance The GDG considered, with respect to strategies for maintaining weight loss after VLCD, the most critical outcomes to examine in the literature to be % weight change (kg) from end of VLCD to end of study, health related quality of life and withdrawals. The GDG were particularly interested in looking at weight change from end of VLCD to end of study, because their clinical experience indicated that initial weight loss in people using VLCDs was not sustained. |
| Trade-off between clinical benefits and harms | The GDG noted the recommendation from CG43 which indicated that the main requirement of a dietary approach to weight loss is that total energy intake should be less than energy expenditure. They also recognised the challenges in engaging people with long-term behavioural change strategies to support and maintain weight loss. Clinicians and people wishing to lose weight may find the benefit offered by immediate and rapid weight loss when adhering to a very-low-calorie diet (a maximum of 12 weeks continuously or intermittently with a low-calorie diet) attractive. The GDG were very much aware that these diets are also available to individuals delivered by the commercial sector (at a cost to the individual). They also noted from their clinical experience that weight regain following a VLCD was common, although weight regain may be slower with proper support in returning to a balanced diet and change in lifestyle. Weight regain in people who have tried VLCDs may cause depression and perpetuate a sense of failure in those people who are trying to manage weight. If weight increases following a VLCD then this would also be undesirable for the person. The GDG noted that this review was an update of the evidence review from the original 2006 guideline but noted that a safety review had previously not been clearly available from the evidence considered in CG43 and wished to identify any concerns associated with such a restricted calorie intake. The GDG noted that early formulations of VLCDs, used in lower BMIs and with limited associated support, had previously been linked to some incidences of mortality in the past. The GDG felt however that the current formulations and associated support made such an outcome unlikely, and therefore did not prioritise this as an outcome in the safety review. Gathering and interpreting this evidence would allow the GDG to determine whether the benefit of any rapid initial weight loss, offset against safety and maintenance of weight loss (and consequently any harms), would require any changes to the existing CG43 recommendation for the NHS. |
| Trade-off between net health benefits and resource use | No relevant economic evaluations were identified. The incremental cost of providing a VLCD relative to a tier 3 service without a VLCD (standard dietary advice) was estimated to be £432 - £994 for individuals without comorbidities and a BMI over 40kg/m2. This cost increases to £759 - £1567 for individuals with comorbidities. At this cost, individuals without comorbidities would need to reduce their BMI by 1.32 – 6.29 kg/m2 and sustain this for a year for a VLCD to be considered cost effective at a £20,000 per QALY threshold. Individuals with comorbidities would need to reduce their BMI by 2.31 – 7.84 kg/m2 and sustain this for a year to be considered cost effective at a £20,000 per QALY threshold. The GDG felt that it was unlikely that this level of weight loss, relative to a tier 3 service without a VLCD, would be achieved or sustained, and therefore agreed that VLCDs are unlikely to be cost effective for this population. They did note that no clinical evidence was available for VLCDs in this population. The GDG did not consider that it was appropriate to extrapolate the effectiveness of a VLCD from one BMI population to another. For individuals with less complex clinical issues (those with no comorbidities and a BMI between 30 – 40 kg/m2), the incremental costs of using VLCDs are estimated to be lower, with an additional cost of £158 compared to standard dietary advice alone. Using these costs, it was estimated that an individual would need to reduce their BMI by 0.48 – 1 kg/m2 and sustain this for 1 year for the VLCD to be considered cost effective at a £20,000 per QALY threshold. The clinical evidence showed that at the end of the clinical trials individuals are expected to lose a very small amount of weight relative to standard dietary advice. This small decrease in weight is unlikely to provide any quality of life benefits. Therefore the GDG agreed that it was unlikely that VLCDs would be cost effective in this population at a £20,000 per QALY threshold. The GDG also noted that the small weight loss achieved from a VLCD measured at the end of the clinical studies was unlikely to result in any improvements in comorbidities. Therefore no additional cost savings or health benefits would be realised (note that such additional benefits have not been incorporated into this analysis). The GDG noted that the evidence on adverse events was weak and that there could be other adverse events which would increase costs and reduce quality of life, making VLCDs even less likely to be cost effective. It is assumed that weight loss achieved through the VLCD will have no long-term impact on quality of life unless it is sustained. The GDG felt that, in some cases, rapid initial weight loss may be of increased clinical benefit to some individuals, particularly those who need to meet clinical requirements for further medical or surgical interventions, for example, orthopaedic surgery and fertility treatment. Therefore a VLCD may be cost effective for this group of individuals as there will be a large benefit derived from the rapid initial weight loss which has not been captured in this analysis. For individuals seeking sustained long-term weight loss, the initial fall and then subsequent rise in weight often seen with VLCDs will not offer sustained improvements in quality of life, therefore use of VLCDs is not considered to be cost effective for long-term weight loss. |
| Quality of evidence | VLCD effectiveness The evidence for the effectiveness review was low to very low quality evidence. Evidence was typically downgraded for risk of bias or imprecision. Most of the studies did not report allocation concealment. There were differences in weight (kg) at baseline between treatment arms which would warrant caution when interpreting the weight change outcomes. The GDG felt it important to note that most of these included studies looked at populations with BMIs that the GDG would consider too low to warrant the use of VLCD on the NHS. There was only evidence in some BMI categories and it was not clear that the data would be relevant to the people with higher BMIs who, in practice, are the most likely to be using VLCDs under NHS care. The GDG discussed that there were no recent studies on effectiveness and the most recent was from 2000. It was also noted that 4 of these 6 studies were from the United States which had carefully selected population inclusion and this could lead to selection bias. The outcome weight in kg change from start of study to end of weight maintenance period had heterogeneity but this was explained by sub-grouping studies by intermittent versus non-intermittent VLCD diets. The GDG believed that this was clinically appropriate and sufficiently explained the heterogeneity. No studies reported on percentage weight loss, but one study did report percentage ‘ideal’ weight loss. Data on weight change in kg and BMI (from start of study to end of VLCD and start of study to end of maintenance period) was extracted and presented to the GDG. Numbers of participants withdrawing from the study were extracted for all 7 studies; however, it was often unclear whether they were withdrawals due to individual patients unable to tolerate the VLCD or due to other circumstances. None of the included studies reported on the following outcomes from the protocol: Health Related Quality of Life and improvement in physical activity. Three studies reported on specific populations: 1 with all male participants and 2 with all female participants. Furthermore, 3 studies reported on people with type 2 diabetes. All included studies compared VLCD to standard dietary advice but the length of VLCD, the number of calories included, the length of treatment, and the length of follow-up, varied. Four papers reported intermittent (3 papers) or short-term (1 paper) use of VLCDs. Six studies included behavioural therapy (or educational sessions) as part of the treatment in both arms, and 4 of these studies included an exercise component as well; however it was not possible to extract any useful data in these areas to inform recommendations. No evidence was found to inform recommendations in the pre-specified subgroups from the protocol: type 2 diabetes, ethnicity, diet and men and women with learning disabilities, people with osteoarthritis, sleep apnoea, and those giving up smoking or BMI thresholds. VLCD safety There was very low to low quality evidence on the safety of VLCDs compared to LCDs. Six RCTs were identified which reported on the pre-specified safety outcomes for VLCDs, with many of the outcomes considered critical or important to decision-making only being reported by 1 study. Most of the studies did not report allocation concealment and some included participants with different baseline characteristics (for example, in depression scores and baseline weight) which may indicate inadequate randomisation or be an indication that allocation was not concealed appropriately. Participants and most of the assessors in the studies did not appear to be blinded and this may have had an impact on the more subjective outcomes, for example depression or binge eating. Furthermore, some of the outcomes such as ‘depressive tendencies’ and ‘marked’ serum uric acid levels were not well-defined and the results were very imprecise. Heterogeneity was found and investigated for the one outcome where meta-analysis was possible: final depression scores at 4 and 5 months. One study66 which showed higher (worse) depression scores for VLCD than LCD was a very small study and there were problems with using the final values for the later study69 which reported higher values for VLCD at follow-up. For the later study, baseline values were also higher for VLCD. Consequently, change scores for this study (not reported) are likely to have shown no difference between VLCD and LCD. There was a similar problem with the results in final depression score reported at 1 year 65 where baseline values were also different and, if change scores were reported, may show little difference between VLCD and LCD. Another possible reason for this heterogeneity could be that the study favouring VLCD was reported at 18 weeks follow-up, whereas the studies favouring standard dietary advice were 1 year follow up. The GDG agreed that a period of rapid initial weight loss during the first phase (at 18 weeks) would be more likely to lead to reduced depression, whereas by the later follow-up times, individuals could have experienced weight regain, as they transition to solid food, and would be more likely to have higher depression scores in the VLCD group. The GDG highlighted that the majority of evidence identified was in an indirect population of people who would be unlikely to receive a VLCD in NHS clinical practice (that is, the evidence is in people with a BMI less than 40). No evidence was identified on the use of VLCDs in people with a BMI over 40 with complex comorbidities. There was evidence from one study to suggest that binge eating may be worse in those who have VLCD compared to LCD at 1 year. However, the GDG noted that final binge eating scores were better than at baseline; which was considered to be likely due to the psychological support element of the treatment that the patients received. However, there were no outcomes reported related to other aspects of disordered eating, such as bulimia nervosa or night eating syndrome. Regarding other possible adverse events, one study reported on gallstone formation detected by ultrasound. However, this was a small study which specifically looked at biliary physiology during VLCDs and LCDs and reported a very high rate of (asymptomatic) gallstone formation associated with VLCDs, leading to the study terminating early. One study reported on constipation and diarrhoea, reporting that there was a slightly higher occurrence of these effects with VLCDs compared to LCDs but there was little certainty in the results. No studies reported on the occurrence of gout but one small study reported the proportion of participants with ‘marked’ serum uric acid levels. As the study did not define this, it was difficult to draw any conclusions. The same study reported the mean levels of serum uric acid to be similar between VLCDs and LCDs. VLCD maintenance Most of the evidence for the maintenance review was very low to low quality evidence. The GDG noted that the studies also consisted of small sample sizes. Evidence was typically downgraded for risk of bias or imprecision and outcomes on weight (change or final scores) were downgraded for indirectness as they were not reported in the GDG's preferred way, % change from baseline. Most of the studies did not report allocation concealment. Some outcomes had very wide confidence intervals reflecting large levels of uncertainty in results. There were differences in weight (kg) at baseline between treatment arms. The GDG noted that the evidence presented was not applicable for relevant patients with a BMI above 50 as the majority of the studies included populations with lower BMIs. Only one study presented % change: the remaining studies reported the alternative outcomes. There was no clinical evidence found relating to health related quality of life. It was possible to determine withdrawals for all included studies; however once again, it was often unclear whether they were withdrawals due to the maintenance strategy or due to other circumstances. Two studies reported physical activity as an outcome but as both studies were exercise interventions and this outcome was only reported during the study, the results are more about adherence to the exercise regimen rather than improved physical activity after the maintenance programme had completed. Four studies reported on sub populations: 3 with all female participants and 1 with male participants. Furthermore, 5 studies did not have a minimum weight loss for inclusion in the maintenance phase of the study. No evidence was found to inform recommendations in the pre-specified subgroups (type 2 diabetes: expected weight maintenance strategies following VLCD to have different outcomes in people with T2DM; ethnicity (white (over 80%); Asian (over 80%); black (over 80%): expected maintenance strategies following VLCD to have different outcomes in different ethnicities; learning disabilities; adherence to weight maintenance strategies following VLCD may be challenging for men and women with intellectual disabilities; and, diet: expected weight maintenance strategies involving diet to work better in those who followed a supervised diet) or strata (BMI thresholds - intervention thought likely to have a different effect on different BMI thresholds) in the protocol. |
| Other considerations | The GDG considered the applicability of these studies and discussed the challenges in maintaining weight loss following any diet including very-low-calorie diets. Overall, the GDG felt that there was little evidence of effectiveness of VLCDs compared to LCDs (standard dietary advice) in the long-term, but they are relatively safe. The GDG noted that there was some evidence that VLCD worked in the short-term, but outcomes for weight loss at end of maintenance periods found no difference between VLCD and standard dietary advice. Evidence demonstrated that VLCDs achieve slightly greater weight loss over the short period of the intervention compared to LCDs; however this loss is not likely to be maintained. This was also supported by evidence to suggest a benefit in weight reduction at the start of a VLCD was not maintained over a long period of time. The GDG noted that the provision of VLCDs is relatively new in the NHS (other than pre-surgery); however they are provided extensively by commercial providers at cost to individuals. It is outside the remit of this guideline to consider issues related to the use of VLCDs purchased by the individual, although the GDG recognise that commercial companies advise their customers to seek medical advice or monitoring in certain conditions (such as hypertension or type 2 diabetes) and as such are not without cost to the NHS as it is usual for people to get this advice from their NHS general practitioner. The GDG agreed that VLCDs showed a clinical benefit over harms in the initial phase of the diet but felt that these benefits were rarely evident long-term. They also noted from their experience that many patients regain the weight lost during the initial VLCD period and often end up gaining weight. The GDG discussed the harmful nature of weight cycling and associated morbidity as well as the potential dissatisfaction and sense of failure that this weight regain may cause the individual. The GDG observed that there was no difference in withdrawals which was being used to determine adherence to the VLCD, and as such were unable to highlight any supportive measures to encourage adherence to any VLCD. The GDG noted with interest the incidence of gallstones in the VLCD population. The size of this study was acknowledged as a particular weakness, but the GDG were not particularly surprised by the high incidence. It was the opinion of the GDG that gallstones are common after rapid weight loss, but they were unsure whether this was increased in VLCD compared to rapid weight loss by other means. They also noted that they would expect other studies which actively looked for gallstones to discover high rates of gallstones associated with VLCDs. Whilst in practice, their experience demonstrated that the incidence of gall stones in the population was generally increasing probably caused by rapid changes in weight (increase or decrease), there was little evidence to suggest that these were a particular safety issue, especially as, in the evidence reviewed, only one case of gallstones was symptomatic. The GDG also noted that the included study reported using a VLCD with less than 2g of fat per day while the EU SCOOP report62 recommends that there is little risk of gallstones if VLCD include at least 5g per day of fat and virtually no risk with diets that include at least 7g per day. This was considered to be more reflective of current diets. For these reasons, the GDG did not choose to make a research recommendation to further investigate the cause and association and impact of gall stones in the population of people using VLCDs. The GDG considered the evidence related to depression. While the study showed that those with a VLCD may have higher (worse) depression scores than those with a LCD, scores were also higher in this group at baseline. Given this, the GDG did not feel able to make any conclusions about this. The GDG did not consider overall that there were benefits to providing VLCDs to the majority of obese people who wish to lose weight and recommended that it should not be routinely offered. However, the GDG considered that there were likely to be some benefits to providing VLCDs to selected people who need to lose weight quickly for clinical reasons, for example, those people who are being considered for surgical procedures such as orthopaedic surgery or women who wish to conceive. The GDG defined ‘rapid weight loss’ appropriately as greater than that which can be achieved with dietary and lifestyle changes. The GDG recommended the use of VLCDs in patients clinically assessed by the health care professional as likely to benefit from rapid weight loss in these or analogous circumstances. The GDG noted that there is concern that weight loss is not too rapid (that is, likely to result in excess loss of lean body mass and increase risk of gallstones) and that it is important to manage safely. The usual recommendation is 0.5-1kg/week (that is 0.5-1% for a 100kg person). Weight loss may be more rapid in the first couple of weeks due to fluid loss. The GDG noted that the maximum, safe, recommended rate without losing significant lean mass is approximately 1.5% per week, noting that the maximum weight loss per week required to not exceed this needs to be recalculated as weight is lost. The most successful VLCD trials achieve a total weight loss of 8-12% over 12 weeks which is in keeping with a 1% per week loss. The GDG reported that from personal experience there is a group of individuals who may do very well on a VLCD despite the poor evidence. The GDG agreed that patient choice would also be very important as some patients may not want to try a VLCD while others would be very keen. They also indicated in their recommendation a set of activities that should take place before a VLCD was offered that reflects the importance of ensuring that people are aware of the challenges with ensuring long-term sustained weight loss with this type of diet. They felt it important to ensure that the use of VLCDs was only used as part of a multicomponent intervention to ensure weight loss (as already recommended in CG43), and chose to highlight this in their recommendations. Furthermore, the GDG considered it important to note that diets less than 800kcal, were carried out under clinical supervision. The GDG also discussed the evidence related to disordered eating. The GDG felt from their clinical experience that individuals without a previous history of disordered eating behaviours may be more at risk of developing these after use of a VLCD, although no evidence that fit the criteria specified in the protocol was found. They agreed by consensus that careful review and screening of people being considered for a VLCD diet for binge eating and other disordered eating behaviours was necessary as from their clinical experience, VLCDs may trigger these behaviours and they chose to include this specifically in their recommendation. They particularly wished to ensure that the re-introduction of normal or solid food was carefully discussed to ensure that the person was reintroduced to a balanced diet rather than return to previous eating habits to maximise the chance to maintain weight loss following the intervention. Although no formal review of the evidence to identify the support that an individual should receive before initiating a VLCD, they felt that it was important to highlight that individuals being considered for a VLCD as part of a multi component weight strategy should be considered for counselling and assessed for eating disorders and other psychopathology. The GDG recognise that in any service of this kind, an appropriate pathway should be in place. It is beyond the scope of this guideline, however, to define that pathway or which professional groups support any pathway. The GDG noted that, from their tier 3 service experience, relevant psychological skills support, usually from a clinical psychologist, was available for the necessary assessment for eating disorders or similar. The GDG also wished to highlight that a discussion regarding the risks and benefits of VLCDs, and the concept of VLCD as a short-term weight strategy should take place before initiation of the diet. The GDG particularly felt that the re-introduction of healthy eating habits was more likely if the reintroduction of food had been discussed before the start of the diet. The GDG also wished to note that they considered it may also be important to discuss potential side effects and to monitor in order to ensure any side effects are only transient. The GDG used informal consensus to develop a recommendation to reflect this. The GDG wished to highlight that all people who are being given a VLCD should be monitored and reviewed regularly and provided with support to help maintain weight loss. The duration of this support should be tailored to individual need as outlined in recommendation 38. The GDG noted that the review did not consider evidence on the use of VLCD prior to bariatric surgery and believed this to be a separate issue. The GDG wanted to highlight that they did not want standalone VLCDs to be linked to the use of VLCD as a 2 week pre-surgery treatment. It was the opinion of the GDG that these diets should not be seen as a way to avoid surgery as although they do work for some people they are in general not as effective as bariatric surgery. The GDG were concerned that clinicians would use VLCDs routinely instead of bariatric surgery, and that the patient's treatment may be delayed as a consequence. The GDG also discussed the importance of highlighting that the use of VLCDs should not be considered as an alternative to surgery or a gateway to surgery. It was also noted that the definition of a VLCD diet has been amended since CG43 (see Glossary) which defined them as 1000 calories or less. This guideline has clarified and defined them as 800 calories or less, which is an accepted definition within current practice. The GDG selected adults and children over 2 years to be included within these VLCD reviews. However, no evidence was found for children for any of the VLCD reviews on effectiveness, safety or maintenance. VLCDs are not used on children in current practice and the GDG did not believe the evidence on adults could be extrapolated to make any recommendations on children. Similarly, it was not possible to make recommendations for any other of the identified subgroups or groups warranting special consideration. In line with advice from the co-opted expert for learning disabilities, the GDG noted that people with learning disabilities across the lifespan and carers supporting them should have access to specialist advice on overweight and obesity. They also felt that clinicians undertaking health checks for people with learning disabilities should be aware of referral pathways to services for individuals with obesity and thateducational programmes for health, education and social care staff should include training on assessing and managing overweight and obesity in people with learning disabilities. |
- Surveillance report 2018 - Obesity: identification, assessment and management (2014) NICE guideline CG189 and BMI: preventing ill health and premature death in black, Asian and other minority ethnic groups (2013) NICE guideline PH46
- NICE Clinical Guideline 43: Obesity prevention
- NICE Clinical Guideline 43: Obesity: The Prevention, Identification, Assessment and Management of Overweight and Obesity in Adults and Children
- Very-low-calorie diets - ObesityVery-low-calorie diets - Obesity
Your browsing activity is empty.
Activity recording is turned off.
See more...