Childhood obesity is a global health concern. Conventional nutrition guidelines have come under scrutiny in helping to achieve long-term healthy weight. An alternative carbohydrate-restricted, higher fat approach has shown to be effective in adults, but research is limited in youth.
To assess the feasibility of a 12-week whole-food, carbohydrate-restricted diet on weight loss and metabolic health.
Overweight children aged 8–13 years.
In this single-arm study, 25 overweight children were provided with whole-food, carbohydrate-restricted dietary guidelines. Primary outcomes – dietary acceptability, adherence and affordability – were assessed qualitatively weekly (telephone) and post-intervention (focus groups). Secondary outcomes – Body mass index (BMI), waist circumference, lipids and glycaemic control measures – were assessed at 0 and 12 weeks. Change scores were analysed using the
Overall, dietary acceptability was mostly positive, and reports of affordability by parents were mixed. Attrition rates were high (48%); adherence was influenced, positively and negatively, by levels of support from friends and family. Completing children reduced BMI by 2.1 ± 1.5 kg.m2 (
Children achieved some weight loss and health outcome success using this dietary approach. For sustainable weight loss maintenance, full family and health professional support, particularly on a more intensive level at the start, may be required.
Childhood obesity is one of the greatest global public health challenges of our time. Over the past 30 years, it has more than doubled in children and quadrupled in adolescents worldwide.
Despite the effort invested, globally, to alleviate childhood obesity, it continues to rise incessantly.
More recently, an alternate option for dietary guidance has emerged,
The study took place in Hawkes Bay, New Zealand. It was a single-arm trial (no control group), which aimed to determine the feasibility from several perspectives, that is diet acceptability, adherence, affordability and efficacy, of applying a 12-week whole-food LCHF intervention on overweight children aged 8–13 years and their families. The trial was approved by the Northern A Health and Disability Ethics Committee, New Zealand, No. 14/NTA/99. All caregivers provided written consent on behalf of their children and for the use of their data in the study. The trial was retrospectively registered with the Australian New Zealand Clinical Trials Registry on 28 November 2016 (Registration reference: ACTRN12616001640493).
Our target sample size for this feasibility study was 25; 28 children were invited, and three declined to participate. Children were included if they met the criteria for being overweight according to the Centers for Disease Control (CDC) definition of ≥85th percentile for BMI-for-age,
All participants and their caregivers or families were invited to attend a 90-min workshop delivered by a dietitian and an endocrinologist, during which the LCHF nutrition approach was explained. It focused on a moderate level of carbohydrate restriction and was based on a whole-food approach, discouraging consumption of processed food wherever possible. Participants were not provided with an individualised calorie-controlled plan, or prescribed a macronutrient breakdown. Instead, they received a food guide (i.e. preferred foods and foods to avoid). It was anticipated that this style of eating reflected a moderate carbohydrate restriction, low glycaemic index diet, with intake approximating less than the lower range specified in mainstream New Zealand food and nutrition guidelines, that is 45% of total energy; a higher fat intake, approximating more than 33% of total energy, and a moderate protein intake, approximating 15% – 25% of total energy.
Prior to starting the intervention, participants attended a dietitian’s clinic in a fasted state, where they had their baseline anthropometry measures taken, and thereafter attended the hospital for their blood tests. This was repeated at the conclusion of the intervention at 12 weeks. Participants’ caregivers were contacted by telephone each week to assist with any questions they might have and to assess progress using the feasibility aspects as prompting questions (see
The primary outcome variables were diet acceptability, adherence and affordability, which were measured qualitatively in two ways: firstly, as part of the individual interviews with caregivers during the weekly telephone conversations. Parents were asked open-ended questions about these three aspects, along with other more general questions to assess overall progress; data were documented in written format. Secondly, as part of the focus groups conducted with parents or caregivers and children at the end of the intervention. In addition to the parent focus groups, before they commenced, we also asked each of them, individually, about the affordability of the diet, and documented this accordingly in written format. We did this to provide another, more private, opportunity to discuss the financial aspect with a member of the research team, with whom they had built rapport over the course of the study, in case they felt uncomfortable talking about their finances in a group of mixed socio-economic individuals.
Secondary outcome variables were anthropometric measures (BMI – calculated as weight [kg] / height2 [m], weight, waist circumference), fasting lipid profiles (HDLc, LDL cholesterol [LDLc] and triglycerides) and glycaemic control (serum glucose, serum insulin and HbA1c). Body weight was measured using weighing scales (Tanita-410, Tanita Corporation America Inc, Arlington Heights, IL), height using a stadiometer (Tanita HR-200, Tanita Corporation America Inc) and waist circumference using a Lufkin W606PM tape measure. BMI
Qualitative data were analysed according to standard qualitative data analysis protocols as follows. The focus groups were recorded and transcribed. All data from the focus groups and from the weekly interviews were uploaded into qualitative data analysis computer software (NVIVO 11, QSR International, Victoria, Australia) and analysed using thematic analysis. We examined the transcripts for emerging themes guided by the three categories; diet acceptability, adherence and affordability. Data were coded by and discussed between two members of the research team, and are presented in tabular format as key themes with supporting transcripts. The data collected, analysed and presented for the qualitative aspect of the study represent that of both the non-completers (from the telephone conversations prior to drop out) and the completers (those that attended the focus groups) to avoid any potential bias of presenting data from only the most motivated participants.
All secondary outcomes of participants (completers only) are presented as means and standard deviations for the pre- and post-measures. Because of the explorative nature of this study and our small sample size, quantitative data are presented as individual responses. We elected to apply a probability statistic using a paired
A total of 25 children were recruited into the study, and 13 of them completed (attrition was 48%).
Baseline participant characteristics and blood marker variables.
Participant characteristics | Total |
---|---|
Female | 15 (60) |
Male | 10 (40) |
New Zealand or other European | 11 (44) |
Maori | 12 (48) |
Pacific Island | 2 (8) |
Age, mean ± SD (years) | 10.52 ± 1.66 |
All | 68.4 ± 16.6 |
Female | 65.9 ± 17.0 |
Male | 72.0 ± 16.2 |
All | 150.6 ± 9.6 |
Female | 148.5 ± 9.7 |
Male | 153.8 ± 8.9 |
All | 29.7 ± 4.1 |
Female | 29.4 ± 4.3 |
Male | 30.1 ± 4.0 |
All | 2.2 |
Female | 2.2 |
Male | 2.4 |
All | 100.1 ± 9.5 |
Female | 98.2 ± 7.7 |
Male | 103.0 ± 11.7 |
Waist:height | 0.7 ± 0.1 |
HDLc (mmol/L) | 0.8 ± 0.4 |
LDLc (mmol/L) | 1.5 ± 0.5 |
Triglyceride (mmol/L) | 0.9 ± 0.5 |
Triglyceride:HDLc | 1.3 ± 0.8 |
HbA1C (mmol/mol) | 29.9 ± 2.5 |
Glucose (mmol/L) | 4.4 ± 0.9 |
Insulin (pmol/L) | 106.6 ± 86.9 |
HOMA-IR | 3.27 ± 3.5 |
SD, standard deviation; HOMA-IR, Homeostatic model assessment – Insulin resistance; HDL, high-density lipoprotein; LDL, low-density lipoprotein.
Themes and supporting transcripts from caregivers and children.
Themes | Supporting transcripts of caregivers |
---|---|
Acceptability | ‘I found breakfasts really easy because my daughter just made her own; she’s old enough to cook her own bacon and eggs or whatever, she’s started making omelettes for us which has been really good’. |
Satiety | ‘One thing I definitely notice is that when I’m not eating like normal, you don’t need as much food, you actually don’t feel like as much’. |
Being prepared | ‘I just found it incredibly stressful if I am not organised. Because I hate cooking, I don’t have anything at home that I can just throw together, so that means I have to go to the supermarket.’ |
Adherence | ‘It was very easy for the 3 months with strict boundaries. My daughter was probably the same, she was very determined, so there were no problems’. |
Support | ‘My daughter’s teacher noticed she’d changed because she was bringing almonds. So the teacher used to have chocolates as treats but she started bringing almonds for the kids’. |
Affordability | ‘I found a bit of saving in terms of buying food on the run’. |
Acceptability | ‘I liked it. My favourite bit would have been having cheeses because I like cheese. And the cauliflower mash, that was yum as well’. |
Adherence-linked to support | ‘Well it was a bit hard because my dad would occasionally have some chips because he likes those’. |
Satiety | ‘At the start I would have eaten all the sections in my lunch box for the first month, then it kinda just got that I wasn’t eating as much anymore’. |
Individual responses for anthropometric and metabolic health outcome variables.
Change in metabolic health markers of completing participants.
Variables | Baseline Mean ± SD | Week 12 Mean ± SD | Effect size | |
---|---|---|---|---|
Weight (kg) | 62.9 ± 10.7 | 59.7 ± 12.4 | −0.3 | |
BMI (kg.m2) | 28.1 ± 3.1 | 26.0 ± 3.8 | −0.7 | |
BMI |
2.2 | 1.9 | - | - |
Waist (cm) | 97.8 ± 8.5 | 92.5 ± 9.9 | −0.8 | |
Waist:height | 0.7 ± 0.1 | 0.6 ± 0.1 | −0.9 | |
Glucose (mmol/L) | 4.6 ± 0.9 | 4.8 ± 0.4 | 0.2 | |
Insulin (pmol/L) | 119.0 ± 89.0 | 85.0 ± 52.6 | −0.4 | |
HbA1C (mmol/mol) | 30.4 ± 2.9 | 31.0 ± 2.0 | 0.2 | |
HOMA-IR | 3.9 ± 4.2 | 2.6 ± 1.7 | −0.3 | |
LDLc (mmol/L) | 1.6 ± 0.5 | 2.0 ± 0.6 | 0.8 | |
HDLc (mmol/L) | 0.7 ± 0.2 | 0.9 ± 0.2 | 1.5 | |
Triglycerides (mmol/L) | 1.0 ± 0.5 | 0.8 ± 0.2 | −0.3 | |
Triglyceride:HDLc | 1.5 ± 0.8 | 0.9 ± 0.5 | 0.7 |
HDL, High-density lipoprotein; LDL, low-density lipoprotein; HOMA-IR, HOMEostatic model assessment – Insulin resistance; SD, standard deviation;
, Significant finding.
This is the first study to qualitatively investigate the feasibility of a whole-food, LCHF nutrition approach for weight loss in children. Overall, the findings suggest that there is potential, particularly with support from health professionals, friends and family as these factors impact adherence and ultimate success.
Parents and children found the majority of the food choices to be acceptable. Not every food or meal suggested was enjoyed; however, it is likely that whatever the nutrition paradigm, children have unique food likes and dislikes with ‘fussy eating’ being a relatively common phenomenon during childhood in general.
There was mention that children became bored with the lack of variety and the restrictions put on food choices, particularly towards the end of the intervention. Again, while this might not necessarily be different from boredom experienced with usual food intake, it does emphasise the need for more support in this area. Despite a plethora of resources on the Internet, it might be that a structured, comprehensive package of meal ideas and recipes alongside some practical cooking class guidance is required to better support families and to help prevent boredom. Kirk et al. attributed their long-term weight loss success in all three dietary intervention groups, to the initial intensity of the guidance applied.
Adherence was directly related to the level of social support. Both parents and children reported LCHF eating easy to adhere to; however, when friends and family members were not supportive, adherence was challenged and often compromised. Children expressed that with family support, LCHF eating would not only have been easier to adhere to but also to continue with. Some mentioned that when friends and acquaintances understood why they were eating this way, they became supportive and even aligned their food choices. Parents also attributed their children’s success to their determination and willingness to participate. Bailes et al. describe two children reporting difficulty complying with their low carbohydrate diet at school as the reason for not completing their 8-week intervention. While the theme of support is similar to our study, in this case, the diet was extreme (< 30 g carbohydrate per day),
We propose that the satiety that usually accompanies LCHF diets may have an important influence on adherence, which ultimately affects sustainability and success. Research evidence suggests that protein is the most satiating macronutrient.
In the only systematic review of weight loss randomised controlled trials in children comparing diets with differing macronutrient profiles, authors refer to a low carbohydrate diet as being hard to adhere to long-term because of restriction of foods.
Adherence was also assessed quantitatively by anthropometric measures. One cannot assume that weight loss was achieved as a direct result of adherence to LCHF eating; however, we are confident in saying that adherence was good in these children, as a dietary history (24-h dietary recall and food frequency questionnaire) was measured during weekly phone calls and during the two face-to-face consultations. The data for these measures are not presented here as they were incomplete; researchers found inconsistencies with availability of caregivers, many of which did not have voicemail services on their telephones or did not return calls. The attrition rate was considered high in this study (48%). It was assumed that adherence was compromised in the non-completers, apart from two families, who despite documented weight loss progress for the first 8 weeks could not logistically get to their final set of measures. Despite our best efforts, we were unable to contact these families to establish reasons for discontinuation. High attrition rates are not uncommon in weight loss studies in general, and our rate is consistent with those reported in similar studies, with both LCHF (22% – 70%) and low-fat approaches (0% – 70%) showing wide variation.
Mixed reports of affordability included cost savings when family members supported LCHF eating because of alternative foods not being purchased, and extra costs spent if these foods were extra purchase items. Some families reported vegetables being expensive during some months. This, however, would not differ from mainstream guidance, where a high intake of vegetables is also advocated. Most parents reported that new foods added extra cost initially, but then savings were made on foods that were no longer being purchased. Some commented that the cost increase was outweighed by the benefits and was not a barrier to sustainability. These findings indicate that further work emphasising lower cost items, LCHF budgeting strategies and incorporating cost data would be valuable.
Our study demonstrated significant improvements in BMI and waist:height ratio at three months. We do not know whether this would be sustainable long-term; however, even with this short intervention, we were able to demonstrate the significant weight loss potential of this dietary approach. This finding, along with the significant finding of improved HDLc, compares favourably with that of similar studies.
Two children in our study increased LDLc above the reference range, a finding that has been reported in similar studies.
Our study has several limitations. It was a short-term study with a small sample, no control group and was not controlled for exercise. In particular, the lack of a control group prevents us from drawing causal inferences. We also do not have a comprehensive set of diet data to objectively verify adherence to this way of eating over the 12 weeks. A further limitation of the study is using HOMA-IR to determine insulin resistance. We acknowledge the emerging evidence, showing that insulin resistance and the use of fasting insulin measures to determine insulin resistance may no longer be the best methods to determine disease risk.
Key strengths of this work were that it profiled a real-life translation of a carbohydrate-restricted style of eating amongst a mixed ethnic group of families with overweight children. The qualitative exploration into how both children and their families fared when adopting an LCHF way of eating is a further strength as this type of information provides us with valuable insights from both caregivers’ and children’s perspectives as to how best to move forward with the application of LCHF nutrition, in research and in practice. We would recommend that future work of this nature should progress this by capturing qualitative outcomes in addition to establishing dietary efficacy. Finally, an additional strength of this work was the range of health professional expertise and their input in this study, from recruitment through to study implementation, that is, GPs, an endocrinologist, two registered dietitians and a registered nutritionist.
Our study was primarily a feasibility study to investigate the translation of the LCHF dietary approach in the homes and lives of children and their families. As such, we are encouraged by the favourable outcomes and the stories of influence that ensued. We urge researchers to pursue the field by assessing the long-term sustainability of the whole-food LCHF approach for overweight children and their families.
The authors declare that they have no financial or personal relationships which may have inappropriately influenced them in writing this article.
C.Z. and G.S. were involved in the development of the study design. C.Z., M.W., J.M., D.S. and O.S. were involved in the implementation and data collection. C.Z., M.W. and N.H. conducted the analysis and interpretation of the quantitative data. C.Z. and J.M. were involved with the interpretation of the qualitative data. C.Z. prepared the manuscript. All authors read and approved the final manuscript.
How is everything going in general on this diet? What’s working well? What’s not working well?
How is your child feeling overall?
Can I help you with anything that is not working well?
Can I help with any queries you might have about any aspect of the diet?
FOR SUBSEQUENT PHONE CONVERSATIONS: The last time we talked, you found [issue] to be difficult. How has that been this week?
(if necessary) What might need to be changed in order to help you follow the new way of eating?
(if necessary) What are your ideas about making these changes?
Can you take me through yesterday’s food intake for your child?
Breakfast
Lunch
Dinner
Other
Food frequency questionnaire (11 questions)
In the past 7 days, on how many occasions did you consume bread/toast/ wraps/bread rolls?
(Interviewer to encircle the answer)
0 1 2 3 4 5 6 7 8 9 10 > 10
In the past 7 days, on how many occasions did you consume pasta or noodles? This includes all pasta dishes, lasagne and noodles such as 2 min noodles, Vietnamese noodles, etc.
(Interviewer to encircle the answer)
0 1 2 3 4 5 6 7 8 9 10 > 10
On how many occasions over the past 7 days did you consume rice? This includes brown or white rice and sushi.
(Interviewer to encircle the answer)
0 1 2 3 4 5 6 7 8 9 10 > 10
On how many occasions over the past 7 days did you eat fast food or takeaways from places like McDonalds or Burger King?
(Interviewer to encircle the answer)
0 1 2 3 4 5 6 7 8 9 10 > 10
On how many occasions over the past 7 days have you consumed crackers, or snacks such as crackers, potato chips, corn chips and pies?
(Interviewer to encircle the answer)
0 1 2 3 4 5 6 7 8 9 10 > 10
On how many occasions over the past 7 days did you drink fruit juices, soft drinks, sports drinks or energy drinks? Do not include diet varieties.
(Interviewer to encircle the answer)
vii 1 2 3 4 5 6 7 8 9 10 > 10
On how many occasions over the past 7 days did you consume lollies, sweets, chocolates or confectionary?
(Interviewer to encircle the answer)
0 1 2 3 4 5 6 7 8 9 10 > 10
On how many occasions over the past 7 days did you consume baked goods such as manufactured or homemade biscuits or cakes?
(Interviewer to encircle the answer)
0 1 2 3 4 5 6 7 8 9 10 > 10
How easy or difficult was it for you / your child to stick to eating in such a way for the duration of the study?
What were the factors that made it easy?
What were the factors that made it difficult / barriers?
Were there any meal or snack times / occasions that you found particularly easy/difficult to follow the new way of eating?
What would make it easier for your child to stick to this way of eating?
Can you comment on whether you and / or your child enjoyed eating this way and the reasons for this.
How did this way of eating fit in with your family life?
How did this way of eating fit in with your child’s school / social life?
Do you think your child (and you) will continue with this way of eating?
Why / why not?
Can you describe how eating like this affected your budget? (asked separately and in the focus group)
How easy or hard was it for you to stick to eating like this?
What was easy about it? / When were the easy times?
What was hard about it? / When were the hard times?
How did other kids at school react?
What would make it easier for you to stick to this way of eating?
Did you enjoy eating this way? Why? / Why not?
Which foods did you like the most?
Which foods did you like the least?
How did this way of eating fit in with your friends at school / family life?
Do you think you will carry on eating like this?
Why / why not?