Uptake and early outcomes of a meal replacement intervention in an ethnically diverse adult population living with obesity and significant comorbidity

Abstract

Aims

Structured meal replacement low energy diets (MR‐LED) are effective in obesity management, but evidence for individuals of minority ethnic groups or those with BMI >35 kg/m² and significant comorbidities is limited.

Materials and methods

We evaluated the early effectiveness of a weight management program delivered by a multidisciplinary team within a specialist weight management service. We enrolled three groups: those who had been declined bariatric surgery, young adults with type 2 diabetes (T2D), and people awaiting corneal transplant who had their procedure deferred due to their weight. Individuals were primarily treated with a 12‐week MR‐LED. Adjunctive obesity medications were available under pre‐specified criteria. The present study reports outcomes following the 12‐week MR‐LED, which was the first phase of a 12 month programme.

Results

Of the 1172 people screened, 482 met inclusion criteria and 128 commenced MR‐LED. Most participants (111/128, 87%) had previously been declined bariatric surgery; 11/128 (9%) were young adults with T2D and 6/128 (5%) were awaiting corneal transplant surgery. Ninety‐three individuals (73%) were from Pacific or indigenous Māori ethnic groups. The cohort had mean (SD) baseline weight of 145.0 (38.6) kg (BMI 49.9 (11.3) kg/m²); 94 (73%) had T2D on treatment, with a mean HbA1c of 67.6 mmol/mol (11 young adults and 83 from the bariatric surgery declined list). Psychological distress was common with 41% and 30% having moderate depressive and anxiety symptoms, respectively. Ninety‐two people (72%) completed the 12‐week MR‐LED, achieving a mean (SD) weight loss of 10.9 (7.8) kg. Young adults with T2D were less likely to complete the programme. Among people with diabetes, 63/94 (67%) completed the 12‐week intervention, with 21/63 (33%) achieving an HbA1c <50 mmol/mol on no diabetes medication. Psychological and quality of life scores improved.

Conclusions

A 12‐week MR‐LED intervention delivered in a specialist service achieved clinically meaningful weight loss and improved metabolic and psychological outcomes in a high‐risk, multi‐ethnic cohort living with higher body weight.

1. INTRODUCTION

Obesity reduces life expectancy and increases the risk of multiple health conditions including type 2 diabetes (T2D), vascular disease, and several cancers. ¹ Recent years have seen a significant increase in rates of obesity around much of the globe, including New Zealand.^{2, 3} South Auckland is an ethnically diverse and deprived region of New Zealand's largest city,^{4, 5, 6} with obesity rates well above the national average (37.6% vs. 31.7%). ⁷

Public healthcare in South Auckland is delivered through Counties Manukau Health (CMH) which offers a small number of public‐funded bariatric surgeries each year.^{8, 9} Due to the limited resource, access to bariatric surgery is based on a locally developed prioritisation score, ¹⁰ rather than those meeting international guideline criteria. ¹¹ Until recently, there were no funded non‐surgical specialist weight management services in the CMH region beyond community providers of lifestyle programmes or primary care dietetic support.

Meal replacement low energy diets (MR‐LEDs) are one obesity treatment option, whereby individuals consume low calorie meal replacement products in place of food. In 2018, the landmark Diabetes Remission Clinical Trial (DiRECT) demonstrated robust efficacy of a structured 12‐week MR‐LED intervention delivered alongside behavioural support. Those who completed the 12‐week intervention lost a mean 14.5 kg in weight, with 46% achieving remission of T2D at 12 months. ¹² Similar results have subsequently been observed in a number of MR‐LED clinical trials in a variety of international settings.^{13, 14, 15, 16} Real‐world data describing the outcomes of a structured MR‐LED delivered by a specialist weight management service demonstrated comparable efficacy to that seen in trials. ¹⁷ MR‐LEDs are an attractive option for obesity management due to their high efficacy, potentially durable effects and relatively lower cost compared to other interventions. Despite the broad evidence base for MR‐LED, there are limited data regarding their acceptability and efficacy in populations with very high body weights (body mass index [BMI; weight in kg divided by the square of height in metres] above 35 kg/m²) or in ethnically diverse cohorts.

This study is a service evaluation of an in‐person, group‐based MR‐LED programme in individuals living with very high body weights. We describe the feasibility, recruitment, acceptability, and health outcomes after the initial 12‐week phase of a 12‐month MR‐LED programme delivered in the secondary care setting in Auckland, New Zealand between July 2023 and September 2024.

2. MATERIALS AND METHODS

2.1. Study population

Te Mana Ki Tua (TMKT) is a publicly funded specialist weight management service established in South Auckland by CMH in 2023. After stakeholder and consumer consultation, in its first year of operation, TMKT delivered a 12‐month MR‐LED based weight management programme.

Inclusion criteria for invitation to the TMKT service were: age 18–70 years, BMI ≥35 kg/m², and a significant obesity‐related medical condition which is likely to improve with an approximately 10 kg weight loss (e.g., type 2 diabetes, obstructive sleep apnoea, etc.). TMKT exclusion criteria were: type 1 diabetes, New York Heart Association stage 4 heart failure, end‐stage kidney disease, currently pregnant or breastfeeding, not able or willing to give informed consent, significant active psychiatric disorder, active eating disorder or the concurrent use of medication known to cause significant weight gain.

2.2. Recruitment

For the first 12 months of TMKT operation, the decision was made to manage participant numbers through direct invitation from clinical lists as opposed to allowing external referrals. Potential participants for TMKT were triaged from these lists on the basis of eligibility criteria and personally invited by phone call to a monthly group information session describing the MR‐LED intervention. The main source of recruitment was through a list of those previously referred for bariatric surgery at CMH but had their referral declined due to an insufficient local prioritisation score. This group was chosen as they were felt likely to wish to undergo obesity management and also to have significant obesity‐related comorbidity. Two small pilot groups were also invited: young adults (aged 18–27 years) with type 2 diabetes (YAD), with a haemoglobin A1c (HbA1c) >75 mmol/mol, and those awaiting corneal transplant but who had their procedure deferred due to their weight (target BMI <50 kg/m² to proceed). Those who wished to begin the MR‐LED programme following the information session were then offered an appointment with TMKT endocrinologists and dietitians to assess suitability before starting the intervention.

2.3. Intervention

During the study period, TMKT offered a 12‐month weight loss programme delivered in cohorts of 10–15 people which started approximately monthly. The programme structure was modelled on that used in the DiRECT study, ¹² beginning with a 12‐week MR‐LED intervention during which 7 group sessions were offered.

The first cohort during the study period received Optifast meal replacement (MR) products while later cohorts were treated with Counterweight Plus MR products. Those using Optifast were advised to take three MR products per day as well as two cups of non‐starchy vegetables. People using Counterweight products took four MR products per day as well as a fibre supplement. In general, no other food was allowed though additional MR products were used to meet protein requirements in people with higher body weights (calculated as 0.8 g protein/kg adjusted body weight). Optifast MR options were bars, shakes, soups, and desserts while counterweight offered shakes, porridge, and soups. The macronutrient composition of the MR products was similar. Per serve, Optifast products offered 201–203 kcal, 20–21 g protein, 18.2–22.1 g carbohydrate, and 4.3–7.6 g fat. ¹⁸ Counterweight MR products offered 200–212 kcal, 13–21 g protein, 20–29 g carbohydrate, and 2.8–6.0 g fat. ¹⁹ MR products were provided at no cost to participants.

After the 12‐week MR‐LED phase, there was a 12‐week period of food reintroduction (FR) followed by a six‐month weight loss maintenance (WM) phase. The programme was largely offered in group sessions, though 1:1 individual sessions were available if patient factors made group work unsuitable. Participants under the care of TMKT were offered a total of 18 group sessions and 5 medical contacts across their 12 months under the service, with 7 group sessions and 2 medical contacts being in the first 12 weeks (see Figure 1).

FIGURE 1.

Schedule of patient contacts in the TMKT service. Those observed to not meet 5% body weight loss at week 4 at the group sessions, or those experiencing weight regain at later stages of the programme were referred for medical review and consideration of obesity medication support on the discretion of the health coach/dietitian team. Obesity medication support was discussed at the medical appointments. The time period covered by this study has been shaded grey for clarity. LED, low‐energy diet; MDT, multi‐disciplinary team.

Health psychologist support was offered in both group and 1:1 settings. The TMKT health psychologist led three of the seven group sessions during the first 12 weeks, focusing on eating behaviours, stress management, and goal setting. In those with disordered eating or mood disorders, 1:1 health psychology appointments were also available.

Obesity management medications (OMMs) were offered as adjuncts at certain times for individuals based on pre‐specified criteria of <5% weight loss after 4 weeks MR‐LED, <10% weight loss after 12 weeks MR‐LED or a regain of >4 kg during FR or WM. For those who required OMMs, naltrexone/bupropion or liraglutide (3.0 mg) was provided for up to 12 months by the service. The choice between offered OMMs was based on mechanism of action, side effects, and patient preference regarding route of administration.

At the initiation of the MR‐LED, sulphonylureas and short‐acting insulins were ceased due to the risk of hypoglycaemia and sodium glucose cotransporter 2 inhibitors were withheld due to the risk of ketoacidosis. Decisions about the continuation or cessation of other diabetes medications were made on the basis of current glycaemic control. Due to the potential for hypotension with calorie restriction, antihypertensives were reduced at the commencement of MR‐LED based on baseline blood pressure (BP) and underlying comorbidities. This approach was a modification of that used in the DiRECT trial. ¹²

2.4. Outcomes

This was a non‐randomised, non‐controlled intervention study. The primary outcome of this study was change in weight at 12 weeks, following the completion of the MR‐LED. Data were collected and analysed for “completers,” defined as those who completed the 12‐week MR‐LED intervention and attended a 12‐week clinic appointment between weeks 10 and 14. A sensitivity analysis was performed for weight outcomes only with missing data imputed using the baseline observation carried forward (BOCF) approach. Additional outcomes assessed at 12 weeks were: change in glycaemic control, blood pressure, quality of life, and psychological screening scores. Recruitment flow at each stage of the intake process and retention rates at 12 weeks were analysed.

2.5. Data collection

At each patient contact, weight was measured with the same set of electronic scales and BP was measured using an electronic BP cuff. Height was assessed at baseline with a stadiometer.

Severity of obesity‐related complications was assessed using the King's Obesity Staging (KOS) criteria. Nine domains (airway, BMI, cardiovascular, diabetes, economic, functional, gonadal, health status, and body image) were graded 0–3 in severity with higher scores representing more severe complications and a total score reflecting the degree of obesity‐related comorbidity.

Routine metabolic blood tests including HbA1c were performed at baseline and again at 12 weeks.

Validated psychological screening questionnaires were administered at the beginning of the MR‐LED then again at 12 weeks. We administered the Patient Health Questionnaire‐9 (PHQ‐9)—scored 0–27 with higher scores representing more depressive symptoms, ²⁰ the General Anxiety Disorder‐7 (GAD‐7)—scored 0–21 with higher scores representing higher anxiety, ²¹ and the Binge Eating Disorder Screener‐7 (BEDS‐7)—scored 0–17 with higher scores representing higher likelihood of binge eating. ²² We also asked participants to complete the Hua Oranga questionnaire which is a nationally developed tool approved by the New Zealand Ministry of Health that assesses holistic wellbeing across four domains: physical, spiritual, whanau (family) and mental. Each domain is scored 4–20 with higher scores indicating better health.^{23, 24}

2.6. Statistical analysis

Descriptive statistics were used to summarise baseline demographics and clinical characteristics of study participants. Continuous data were described as mean (standard deviation [SD]) and categorical data were summarised as frequencies and percentages.

Differences between means were assessed for statistical significance with t tests or ANOVA for continuous data and chi square tests for categorical data. Statistical significance level was set at 5% (two‐sided).

Statistical analysis was conducted with R (version 4.5.0, R core team 2025).

2.7. Ethics

This study was approved by the Auckland Health Research Ethics committee, reference number AH28855. Localities approval was provided by the Counties Manukau Health research office, reference number 2056. The ethics committee granted a waiver of the requirement for individual informed consent on the basis that analysis was solely of routinely collected clinical data and there was minimal risk presented to participants. Extracted data were stored on password‐protected spreadsheets housed in institutional servers.

3. RESULTS

3.1. Recruitment

Between July 2023 and June 2024, 1172 people were assessed for potential TMKT eligibility. Of these, 974 names (83%) were retrieved from a list of those recently declined bariatric surgery, 186 (16%) from a list of YAD aged 18–27 years with a HbA1c >75 mmol/mol, and 12 (10%) individuals were referred for weight loss prior to corneal transplant surgery.

In total, 690 people did not meet initial screening criteria, most commonly due to the absence of a significant obesity‐related medical condition that would improve with 10 kg weight loss (n = 473), or due to a condition that meant MR‐LED was contraindicated (such as unstable mental health, advanced heart failure or pregnancy) (n = 124).

After initial screening, 482 (41%) people met eligibility criteria screening and were invited to information sessions using phone calls and emails. Of these, 74 (15%) were unable to be contacted and 47 (10%) declined to participate in the programme. Of the 255 (53%) who agreed to attend an information session, 155 (61%) attended. Of those eligible on initial screening, 27% (128/482) began the TMKT intervention (Figure 2).

FIGURE 2.

TMKT participant flow diagram. Of the original 1172 selected for triage, 974 came from the bariatric surgery declined list, 186 from the YAD list and 12 were referrals before corneal transplant. The 482 who met initial screening criteria were: 395/974 bariatric declined list, 75/186 YAD, and 12/12 awaiting corneal transplant. The 255 who agreed to attend an information session composed: 214/395 bariatric surgery declined, 30/75 YAD, and 11/12 awaiting corneal transplant. Ultimately, 111 from the bariatric declined list, 11 from the YAD list and 6 from the corneal transplant list started the MR‐LED. MR‐LED, meal replacement low‐energy diet.

Of the 128 who joined the programme, 124 began a group‐based MR‐LED programme with modifications as required based on protein requirements and comorbidities. Three began a 1:1 MR‐LED programme due to either patient preference to avoid a group setting (n = 2) or language barriers (n = 1). One person was treated with a real food 1:1 programme with liraglutide support due to very high body weight and patient preference.

Overall, 128 people started the programme, of whom 111 (87%) had been declined public‐funded bariatric surgery, 11 (9%) were YAD, and 6 (5%) were from the corneal transplant waitlist.

3.2. Baseline characteristics

Of all the 128 starting participants, 84 (66%) were female, 65 (51%) were Pacific with another 28 (22%) being Māori. Baseline characteristics are shown in Table 1 for the whole cohort as well as by referral source. The mean (SD) weight was 145.0 (38.6) kg, with a mean (SD) BMI of 49.9 (11.3) kg/m². Baseline psychological distress was high: 52 people (41%) had a PHQ‐9 score ≥10, indicating at least moderate depression, and 39 people (30%) had a GAD‐7 score ≥10, indicating at least moderate anxiety. T2D control was poor with a mean (SD) HbA1c of 67.6 (19.5) mmol/mol. Two people were using insulin at baseline.

TABLE 1.

Baseline characteristics by referral source.

	Whole cohort, n = 128	Bariatric decline, n = 111	Corneal transplant, n = 6	YAD, n = 11
Age, years	46.3 (12.0)	48.9 (10.4)	35.0 (8.9)	26.1 (2.2)
Gender
Female	84 (66%)	78 (70%)	1 (17%)	5 (45%)
Male	44 (34%)	33 (30%)	5 (83%)	6 (55%)
Ethnicity
Māori	28 (22%)	27 (24%)	0 (0%)	1 (9%)
Pacific	65 (51%)	51 (46%)	6 (100%)	8 (73%)
NZ European	21 (16%)	21 (19%)	0 (0%)	0 (0%)
Indian	11 (9%)	9 (8%)	0 (0%)	2 (18%)
Other	3 (2%)	3 (3%)	0 (0%)	0 (0%)
Weight, kg	145.0 (38.6)	143.0 (36.4)	207.0 (42.2)	131.0 (30.6)
BMI, kg/m2	49.9 (11.3)	49.8 (11.0)	64.8 (6.5)	43.0 (8.9)
KOS	7.4 (2.7)	7.5 (2.7)	6.3 (1.8)	6.8 (3.5)
PHQ‐9	9.3 (5.6)	9.2 (5.7)	8.8 (3.5)	9.6 (5.2)
GAD‐7	6.4 (5.5)	6.4 (5.6)	5.7 (3.8)	7.4 (5.6)
BED‐7	4.2 (4.7)	4.2 (4.8)	5.7 (4.7)	3.2 (4.0)
T2D, n (%)	94 (73%)	83 (75%)	0 (0%)	11 (100%)
HbA1c in those with T2D, mmol/mol	67.6 (19.5)	66.3 (18.5)	N/A	77.2 (24.5)
# of T2D medications	2.0 (1.3)	2.0 (1.3)	N/A	2.0 (1.3)

Note: Data are presented as mean (SD) or n (%). The KOS is scored 0–27, with higher scores representing more obesity‐related comorbidity. The PHQ‐9 is scored 0–27, with higher scores representing more depressive symptoms. The GAD‐7 is scored 0–21, with higher scores representing more anxiety symptoms. The BED‐7 is scored 0–17, with higher scores representing higher likelihood of binge eating disorder.

Abbreviations: BED‐7, Binge Eating Disorder 7; GAD‐7, General Anxiety Disorder 7; KOS, King's Obesity Scale; PHQ‐9, Patient Health Questionnaire 9; T2D, type 2 diabetes.

3.3. Retention

At 12 weeks, 92 (72%) people had completed the intervention and 36 (28%) had withdrawn. Of the 36 people who did not complete the 12‐week MR‐LED, 26 were lost to follow‐up, 6 withdrew due to life events, and 4 withdrew due to medical events (one each for mental health crisis, vomiting, oedema, and acute cholecystitis).

Those who did not complete the intervention were younger than completers (mean (SD) age 40.4 (11.9) years vs. 48.6 (11.3) years, p < 0.001) and were more likely to be from the YAD cohort (22% vs. 3%, p = 0.002). They had poorer baseline glycaemic control (mean (SD) HbA1c 76.8 (25.0) mmol/mol vs. 62.7 (14.9) mmol/mol, p = 0.009). There were no differences between the two groups in terms of baseline weight (mean 147.0 (33.7) kg in non‐completers vs. 144.2 (40.6) kg in completers, p = 0.66) nor gender (69% female in non‐completers vs. 64% in completers, p = 0.72). There were no significant differences in baseline KOS nor psychological screening test scores.

Retention rates by ethnicity were 21/28 (75%) for Māori, 41/65 (63%) for Pacific, 19/21 (90%) for European, 8/11 (73%) for Indian. A chi‐square test comparing the ethnicity of completers versus non‐completers was not significant (p = 0.12). Retention rates by referral source were 85/111 (77%) for the bariatric decline cohort, 3/11 (27%) for the YAD group, and 4/6 (67%) for those awaiting corneal transplant. A chi‐square test for completion by referral source was significant at 0.002.

Sixty‐nine people were referred and offered a health psychologist appointment, of whom 37/69 attended their first assessment and 26/37 attended a follow‐up psychology session.

3.4. Twelve‐week outcomes

For programme completers, mean (SD) weight fell from 144.2 (40.6) kg to 133.2 (38.2) kg, with a mean (SD) reduction of 10.9 (7.8) kg or 7.7 (4.7) % (p < 0.001). Weight change over time during the MR‐LED phase is shown in Figure 3. Among completers, 65% achieved at least 5% weight loss, 30% achieved at least 10% weight loss, and 6% achieved at least 15% weight loss. Using BOCF for missing data, the mean (SD) weight change at 12 weeks on all participants was −7.9 (8.2) kg (p < 0.001).

FIGURE 3.

Weight loss during MR‐LED for completers. This 12‐week time period is the first phase of a 12‐month weight management programme. The vertical bars represent the standard error of the mean for each measurement. MR‐LED, meal replacement low‐energy diet.

Changes in weight and secondary outcomes at 12 weeks are presented in Table 2. Quality of life and psychological metrics significantly improved for all measures aside from the Hua Oranga mental health domain, which was unchanged from a high baseline. Glycaemic control improved with a significant reduction in HbA1c and the number of diabetes medications used. At the completion of the MR‐LED phase, 21 people (33%) with baseline T2D had a 12‐week HbA1c <50 mmol/mol while taking no diabetes medications. There were small but significant reductions in both systolic and diastolic blood pressure.

TABLE 2.

Twelve‐week outcomes for those who completed the MR‐LED (n = 92).

	Baseline	12 weeks	p‐value
Weight, kg	144.2 (40.6)	133.2 (38.2)	<0.001
BMI, kg/m2	50.0 (11.9)	46.2 (11.4)	<0.001
KOS	7.3 (2.8)	5.6 (3.1)	<0.001
PHQ‐9	9.4 (5.8)	5.4 (4.9)	<0.001
GAD‐7	6.6 (5.4)	4.1 (4.4)	<0.001
BEDS‐7	4.2 (4.8)	2.3 (3.8)	0.002
Hua Oranga physical	14.9 (2.9)	16.8 (2.3)	<0.001
Hua Oranga spiritual	15 (4.3)	17.0 (2.8)	<0.001
Hua Oranga family	17.0 (3.4)	17.6 (3.0)	0.04
Hua Oranga mental	17.4 (2.7)	17.6 (2.6)	0.14
HbA1c in those with T2D, mmol/mol	62.7 (14.9)	55.8 (14.8)	<0.001
# of T2D medications	2.0 (1.3)	0.9 (1.1)	<0.001
Systolic BP, mmHg	131.4 (15.3)	129.4 (15.3)	0.04
Diastolic BP, mmHg	83.6 (8.3)	81.4 (8.3)	0.04

Note: Data are presented as mean (SD) or number (%). The KOS is scored 0–27, with higher scores representing more obesity‐related comorbidity. The PHQ‐9 is scored 0–27, with higher scores representing more depressive symptoms. The GAD‐7 is scored 0–21, with higher scores representing more anxiety symptoms. The BED‐7 is scored 0–17, with higher scores representing higher likelihood of binge eating disorder.

Abbreviations: BED‐7, Binge Eating Disorder 7; BP, blood pressure; GAD‐7, General Anxiety Disorder 7; KOS, King's Obesity Scale; PHQ‐9, Patient Health Questionnaire 9; T2D, type 2 diabetes.

Among completers, those who attended 7 group sessions (n = 61) lost a mean (SD) 12.8 (7.6) kg body weight, compared to 8.5 (7.6) kg in those who attended 6 sessions (n = 16) and 6.5 (5.9) kg in those who attended ≤5 sessions (n = 15) (p = 0.006 using ANOVA).

Five people were prescribed liraglutide 3.0 mg during their first 12 weeks under TMKT care (generally due to <5% weight loss at 4 weeks of MR‐LED). One person discontinued the medication due to gastrointestinal side effects and one was lost to follow‐up. For the three people who continued liraglutide alongside the MR‐LED, the mean (SD) weight loss at 12 weeks was 23.0 (12.7) kg. No individuals were prescribed naltrexone/bupropion.

4. DISCUSSION

MR‐LED delivered to individuals living with very high body weights in an ethnically diverse population in New Zealand who had been declined bariatric surgery show clinically meaningful weight loss and improvement in obesity‐related comorbidities. This is the first report of using a structured MR‐LED in such a population. Interest in the programme among those contacted to participate was initially high. However, almost 40% of those who accepted an invitation to an information session ultimately did not attend the session. Among those who did attend an information session, 90% proceeded to a medical specialist appointment, which is higher than the 48% reported in a similar UK service. ¹⁷ Overall, 34% of those eligible and contactable ultimately entered the programme.

Thirty‐six of the 128 people (28%) who started an MR‐LED intervention did not complete the first 12 weeks of the programme, with most of these individuals being lost to follow‐up. The 12‐week retention rate of 72% is somewhat lower than that reported by other groups offering an MR‐LED intervention.^{12, 13, 14, 25} The reasons for this lower completion rate are unclear, though non‐completers were more likely to be from the YAD cohort and those of non‐European ethnicity. This aligns with previous evidence that younger adults often have lower engagement with structured programmes due to competing life priorities, work or social commitments ²⁶ and may experience higher psychological distress. ²⁷ Qualitative studies highlight cultural norms and food practices as barriers to adherence in non‐European populations and interventions that completely exclude food posing greater challenges.^{28, 29} In addition, the initial MR‐LED programme was group‐based and highly structured, which may not suit younger adults who often prefer flexible, individualised support. ³⁰

The TMKT cohort treated during this period was more complex than those previously described. The mean baseline BMI of 49.9 kg/m² is almost 45% higher than the average baseline BMI of 23 cohorts described in a recent meta‐analysis of MR‐LED interventions. ³¹ Additionally, there was a high level of ethnic diversity, with historically disadvantaged ethnic groups comprising the majority of those treated by TMKT. The mean nine‐domain KOS score was 7.4, representing significant obesity‐related comorbidity, and in those with T2D, glycaemic control was poor. TMKT participants had very high baseline levels of psychological distress, with over 40% meeting criteria for moderate depression and 30% for moderate anxiety.

Despite high levels of complexity, the mean weight loss of 10.9 kg in TMKT completers is similar to that observed by other groups.^{12, 13, 14, 15, 25} Though the numbers were too small for comparative analysis, completers prescribed liraglutide had more weight loss than those completing MR‐LED without medication support. There was a trend towards higher weight loss with increased group session attendance, which has been previously reported.^{32, 33} However, this observational study could not investigate the directionality of this association. Active evaluation of later components of the 12‐month TMKT programme is ongoing.

As well as reductions in body weight, people who completed the first 12‐weeks of the MR‐LED programme had meaningful improvements in wellbeing and psychological health metrics. Additionally, despite fewer diabetes medications, HbA1c improved by a mean 6.9 mmol/mol and 33% of people achieved an HbA1c <50 mmol/mol on no medication (the New Zealand threshold for diabetes definition and remission ³⁴ ).

There is one previous study randomising 39 New Zealand adults with high BMI and prediabetes or type 2 diabetes to either MR‐LED or usual care. ¹⁶ The 12‐week MR‐LED intervention was delivered in the community by a single dietitian to 20 ethnically diverse adults. Over 12 weeks, 25% of the cohort in the MR‐LED arm discontinued the intervention with a mean weight loss among completers of 6.9 kg (6.1%). The cohort in this study had a similar intensity of contacts to those under TMKT care but was managed in primary care with less multidisciplinary input. It is possible that the additional support of the TMKT multidisciplinary team accounts for the greater weight loss observed in our cohort despite similar baseline characteristics and rates of early discontinuation.

This study had a number of limitations. First, it was observational in nature with only moderate numbers of participants. Second, we report outcomes following the 12‐week MR‐LED intervention and it is not yet known how durable these results will be in our population. Third, there were differences between completers and non‐completers which may have led to attrition bias in the primary outcome. Fourth, this was primarily a service evaluation so no multivariate adjustment was performed which may have affected the results. Fifth, two different MR products were used which may confound the results. Finally, diabetes remission was assessed at the 12‐week mark where the HbA1c may have been affected by the low carbohydrate content of the MR products. Assessment of HbA1c at a later timepoint (e.g., 12 months) is required to establish the proportion with sustained remission. Our findings may not be generalisable to a less complex population without the provision of free MR products or MDT support.

Currently, TMKT is New Zealand's only fully publicly‐funded specialist weight management service, providing all programme components including meal‐replacement products and multidisciplinary clinical support at no cost to participants. The service was founded in South Auckland, where the prevalence of severe obesity is highest nationally and where Māori, Pacific and communities experiencing socio‐economically deprivation bear disproportionate burdens of obesity‐related disease. Regional public funding of this service addresses the substantial financial and structural barriers that typically limit access to evidence‐based obesity treatment. ³⁵ TMKT also functions as a model for developing more efficient and scalable models of publicly funded obesity care with ongoing evaluation intended to inform future service expansion and support more equitable national access.

Overall, the present study shows that a structured MR‐LED programme can be acceptable and effective in a multi‐ethnic New Zealand cohort with very high body weights and high levels of physical and psychological comorbidity. TMKT experienced high rates of early attrition but still delivered comparable weight loss outcomes to those seen in less comorbid populations.

CONFLICT OF INTEREST STATEMENT

J.S. has received speaking honoraria from Novo Nordisk. N.B. is an employee and shareholder at Counterweight. Y.J. has nothing to disclose. R.M. has received speaking and advisory board honoraria from Eli Lilly, Novo Nordisk, Boehringer Ingelheim, and Abbott.

DATA AVAILABILITY STATEMENT

Data will not be made available.