ABSTRACT
Background
Obesity is a major global health issue. While tirzepatide demonstrates robust weight reduction in clinical trials, data comparing low maintenance doses (2.5 vs. 5 mg) in non‐diabetic adults with obesity remain limited. This study evaluated the comparative efficacy and safety of 2.5 versus 5 mg tirzepatide combined with lifestyle intervention.
Methods
This multicentre, prospective, non‐randomised cohort study enrolled non‐diabetic Japanese adults (18–65 years) with BMI ≥ 30 kg/m2 or ≥ 27 kg/m2 with comorbidities. Patients initiated tirzepatide at 2.5 mg weekly; after 4 weeks, the dose was maintained at 2.5 mg or increased to 5 mg based on shared decision‐making. All participants received standardised dietary and exercise counselling. The primary outcome was the proportion achieving BMI < 25 kg/m2 or > 15% body weight reduction at 6 months.
Results
Of 112 participants analysed (58 on 2.5 mg, 54 on 5 mg), baseline characteristics were well balanced (mean BMI 30.8 kg/m2). At 6 months, the primary outcome was achieved in 62.1% of the 2.5 mg group and 63.0% of the 5 mg group (HR 0.93, 95% CI 0.58–1.49; p = 0.772). Mean percent weight loss was −15.3% and −16.1%, respectively (p = 0.563), with similar reductions in BMI, skeletal muscle mass, and bone mass. Adherence to dietary therapy was excellent in both groups (96.4%), while exercise adherence was 37.5%. Adverse events were more frequent with 5 mg (50% vs. 35%), primarily gastrointestinal symptoms.
Conclusion
In non‐diabetic Japanese adults with obesity, 2.5 and 5 mg tirzepatide combined with lifestyle modification yielded similar weight loss at 6 months, with better tolerability at the lower dose. These findings support the potential utility of low‐dose tirzepatide strategies in real‐world practice.
Keywords: cohort study, GLP‐1 analogue, obesity therapy, real‐world evidence
1. Introduction
Obesity represents a major global public health challenge, contributing to heightened risks of cardiometabolic diseases, reduced quality of life, and escalating healthcare costs [1]. In non‐diabetic individuals with obesity, lifestyle interventions alone typically achieve only modest and poorly sustained weight loss. As a result, pharmacotherapies targeting the incretin pathway have gained significant attention as effective adjunctive treatments.
Tirzepatide is a dual glucose‐dependent insulinotropic polypeptide (GIP) and glucagon‐like peptide‐1 (GLP‐1) receptor agonist. In large‐scale clinical trials, it has demonstrated substantial weight reduction, particularly at maintenance doses of 5–15 mg once weekly [2]. Although higher doses yield greater dose‐dependent weight loss, they are frequently associated with gastrointestinal side effects such as nausea, which can limit tolerability and dose escalation. Consequently, lower doses of tirzepatide (2.5 or 5 mg) may provide an optimal balance between efficacy and safety, especially in real‐world clinical practice or resource‐limited settings [3]. However, direct comparative data on these lower starting and maintenance doses in non‐diabetic obese populations remain limited.
The purpose of this study was to evaluate the comparative effectiveness of 2.5 versus 5 mg tirzepatide, combined with lifestyle modification, in non‐diabetic obese adults, with a primary focus on 6‐month weight reduction and changes in body composition.
2. Methods
2.1. Study Design and Participants
This was a multicentre, prospective cohort study involving patients who initiated tirzepatide treatment between November 2024 and September 2025 at Hiroshima Cardiovascular Clinic and Onomichi General Hospital. The study population included non‐diabetic Japanese adults (aged 18–65 years) with obesity disease, defined as BMI ≥ 30 kg/m2 or BMI ≥ 27 kg/m2 with at least one weight‐related comorbidity. This definition follows the diagnostic criteria for obesity disease in Japan and aligns with the approved insurance indication for tirzepatide in this population (BMI ≥ 27 kg/m2 with comorbidities), which is higher than the general obesity threshold of BMI ≥ 25 kg/m2 used in Japanese practice [4]. Patients started tirzepatide at 2.5 mg subcutaneously once weekly [2]. Individuals with type 2 diabetes (HbA1c ≥ 6.5%), poorly controlled hypertension, severe renal or hepatic impairment, or any contraindications to tirzepatide were excluded. All participants received standardised lifestyle counselling focused on calorie‐restricted diets and increased physical activity. The study protocol was approved by the Institutional Ethics Board of Onomichi General Hospital (approval number: OJH‐202435), and written informed consent was obtained from all participants.
2.2. Dietary and Exercise Therapy
All patients received individualised dietary and exercise counselling in accordance with established obesity management guidelines [4]. Physicians and clinical staff provided personalised nutritional guidance, with target energy intake calculated based on BMI and meal plans developed to achieve appropriate protein, fat, and carbohydrate (PFC) balance. Target energy intake was set at a 500 kcal daily deficit from estimated requirements, with a macronutrient balance of approximately 20%–25% protein, 25%–30% fat, and 45%–55% carbohydrates (PFC ratio), consistent with Japanese obesity guidelines. Patients were encouraged to maintain food records to promote self‐management. Adherence to dietary therapy was defined as ≥ 80% compliance with the prescribed calorie and macronutrient targets, assessed by self‐reported food records and physician interviews at each visit.
Exercise prescriptions followed the FITT principle (Frequency, Intensity, Time, Type), targeting at least 150 min per week of moderate‐intensity aerobic exercise combined with resistance training [5]. Progress was monitored using pedometers or mobile applications, with gradual adjustments to intensity as tolerated [6]. Counselling sessions were provided monthly by physicians and clinical staff to assess progress and facilitate behavioural modifications. Adherence to exercise therapy was defined as ≥ 80% achievement of the prescribed weekly exercise volume (≥ 150 min of moderate‐intensity activity), assessed by pedometer data, smartphone applications, and self‐reports. Individualisation of programmes, accounting for each participant's age, comorbidities, and lifestyle habits, was prioritised throughout.
2.3. Treatment and Follow‐Up
Tirzepatide was initiated at 2.5 mg once weekly in all patients [7]. After 4 weeks, dose adjustment was determined through shared decision‐making (SDM) at the 1‐month visit. This involved discussion between the physician and patient regarding tolerability, initial weight loss response (targeting at least 2%–3% body weight reduction), patient preference, and lifestyle factors, with patients maintained on either 2.5 mg or increased to 5 mg. The assigned dose remained fixed for the duration of the 6‐month observation period, with no further escalation permitted. Adherence to diet and exercise therapy was assessed via self‐reports and medical records. Body weight, BMI, skeletal muscle mass, and bone mass were measured using bioelectrical impedance analysis (BIA) (InBody 580 or 770, InBody Co. Ltd., Seoul, Korea) every month (at baseline and at 1, 2, 3, 4, 5, and 6 months) [8] Laboratory tests and vital signs were monitored according to routine clinical practice.
Efficacy analyses were performed in 112 participants who completed the 6‐month follow‐up (58 in the 2.5 mg group and 54 in the 5 mg group). Safety analyses included all participants who received at least one dose of tirzepatide (n = 60 in the 2.5 mg group and n = 58 in the 5 mg group).
2.4. Outcomes
The primary outcome was the proportion of participants achieving either a BMI < 25 kg/m2 or a greater than 15% reduction in body weight (BW) from baseline at 6 months [2, 6]. Secondary outcomes included changes from baseline in body weight, BMI, skeletal muscle mass, and bone mass at 6 months in both treatment groups, as well as adherence rates to dietary and exercise therapy and overall safety.
2.5. Statistical Analysis
Continuous variables are expressed as mean ± standard deviation and were compared between groups using Student's t‐test. Categorical variables are presented as n (%) and compared using the χ 2 test or Fisher's exact test, as appropriate. No formal sample size calculation was performed because this was an exploratory real‐world prospective cohort study.
The primary outcome was a binary endpoint assessed at 6 months. To estimate the effect between dose groups, we used a Cox proportional hazards model treating time to achievement of the primary outcome (or censoring at 6 months) as the event time. This model yielded hazard ratios (HR) with 95% confidence intervals (CI). As sensitivity analyses, risk ratios and odds ratios were also calculated and produced highly consistent results. All models were adjusted for baseline age and sex.
To minimise the risk of overfitting, a parsimonious multivariable model was constructed including only clinically important variables and/or those with univariate associations (p < 0.20). The full univariate and multivariate model with all candidate variables is presented in Table S1.
A two‐sided p < 0.05 was considered statistically significant. All analyses were performed using available data without imputation for missing values. Statistical analyses were conducted using JMP software, version 18.2 (SAS Institute, Cary, NC, USA).
3. Results
3.1. Baseline Characteristics
During the study period, 129 patients met the inclusion criteria. Of these, 6 patients in the 2.5 mg group and 5 patients in the 5 mg group were lost to follow‐up. Additionally, 2 patients in the 2.5 mg group and 4 patients in the 5 mg group discontinued treatment for various reasons, including adverse effects. Ultimately, 112 patients were included in the final analysis (58 in the 2.5 mg group and 54 in the 5 mg group).
Baseline demographic and clinical characteristics were well balanced between the two groups (Table 1). The mean age was approximately 41 years, and the mean BMI was 30.8 kg/m2. Males constituted approximately 24% of the participants. With respect to lifestyle factors, the proportion of smokers tended to be higher in the 5 mg group (though the difference was not statistically significant), while alcohol consumption rates were similar between groups. At baseline, 14 patients (24.1%) in the 2.5 mg group and 18 patients (33.3%) in the 5 mg group were receiving dietary therapy (p = 0.516). Exercise therapy was being performed by 15 patients (25.9%) in the 2.5 mg group and 10 patients (18.5%) in the 5 mg group (p = 0.240). The prevalence of comorbidities, including hypertension (approximately 30%) and hyperlipidaemia (approximately 46%), was comparable between the groups.
TABLE 1.
Baseline characteristics between the groups.
| 2.5 mg | 5 mg | p | |
|---|---|---|---|
| N | 58 | 54 | |
| Variable | |||
| Age, years | 41.9 ± 6.5 | 41.3 ± 6.4 | 0.741 |
| Male, n (%) | 14 (24.1%) | 13 (24.1%) | 0.837 |
| Body weight, kg | 79.1 ± 8.9 | 80.9 ± 9.0 | 0.54 |
| Body mass index, kg/m2 | 30.7 ± 5.5 | 30.9 ± 5.6 | 0.81 |
| Skeletal muscle mass, kg | 41.8 ± 6.5 | 42.4 ± 6.7 | 0.686 |
| Bone mass, kg | 2.63 ± 1.6 | 2.65 ± 1.7 | 0.772 |
| Lifestyle factors | |||
| Smoking | 12 (20.7%) | 17 (31.5%) | 0.27 |
| Alcohol | 20 (34.5%) | 19 (35.2%) | 0.939 |
| Diet | 14 (24.1%) | 18 (30%) | 0.516 |
| Exercise | 15 (25.9%) | 10 (16.7%) | 0.24 |
| Physical examination | |||
| Systolic blood pressure, mmHg | 120.2 ± 11 | 122.5 ± 11.1 | 0.464 |
| Diastolic blood pressure, mmHg | 77.3 ± 8.8 | 79.2 ± 9.1 | 0.238 |
| Comorbidities | |||
| Hypertension, n (%) | 16 (27.6%) | 17 (31.5%) | 0.655 |
| Hyperlipidaemia, n (%) | 27 (46.6%) | 25 (46.3%) | 0.978 |
| Liver dysfunction, n (%) | 20 (34.5%) | 20 (37%) | 0.78 |
| Psychiatric disorders, n (%) | 2 (3.4%) | 3 (5.6%) | 0.593 |
| Laboratory data | |||
| Haemoglobin, g/dL | 13.7 ± 3.7 | 13.9 ± 3.7 | 0.391 |
| AST, U/L | 23.2 ± 4.8 | 25.9 ± 5.1 | 0.381 |
| ALT, U/L | 28.2 ± 5.3 | 32.4 ± 5.7 | 0.247 |
| GGT, U/L | 29.2 ± 5.4 | 45.3 ± 6.7 | 0.125 |
| LDL‐C, mg/dL | 133 ± 11.5 | 130.5 ± 11.4 | 0.751 |
| HDL‐C, mg/dL | 57.5 ± 7.6 | 60.7 ± 7.8 | 0.29 |
| Triglycerides, mg/dL | 131.3 ± 11.5 | 133.7 ± 11.6 | 0.887 |
| FPG, mg/dL | 93.9 ± 9.7 | 98.3 ± 9.9 | 0.161 |
| HbA1c, % | 5.5 ± 2.3 | 5.5 ± 2.4 | 0.713 |
| eGFR, mL/min/1.73 m2 | 79.8 ± 8.9 | 74.6 ± 8.6 | 0.128 |
Note: Data are presented as n (%), or mean ± SD.
Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; FPG, fasting plasma glucose; GFR, estimated glomerular filtration rate; GGT, gamma‐glutamyl transferase; HbA1c, Glycated haemoglobin; HDL‐C, high‐density lipoprotein cholesterol; LDL‐C, low‐density lipoprotein cholesterol.
3.2. Outcomes
The primary outcome was the proportion of participants achieving a BMI < 25 kg/m2 or > 15% body weight reduction at 6 months. This outcome was achieved by 62.1% (36/58) of patients in the 2.5 mg group and 63.0% (34/54) of patients in the 5 mg group (HR 0.93, 95% CI 0.58–1.49; p = 0.772; Table 2). Hazard ratios were estimated using a Cox proportional hazards model. Risk ratios and odds ratios showed similar results (Risk ratio 0.99, 95% CI 0.74–1.32; Odds ratio 0.96, 95% CI 0.45–2.06). Components of the primary outcome were also analysed separately. Achievement of BMI < 25 kg/m2 was observed in 39.7% (23/58) of the 2.5 mg group and 35.2% (19/54) of the 5 mg group (HR 0.82, 95% CI 0.45–1.50; p = 0.518). Achievement of > 15% BW reduction was observed in 48.3% (28/58) and 51.9% (28/54), respectively (HR 0.99, 95% CI 0.58–1.67; p = 0.958).
TABLE 2.
A summary of primary and secondary outcomes.
| Variables | No. (%) of patients | HR (95% CI) | p | |
|---|---|---|---|---|
| 2.5 mg | 5 mg | |||
| N = 58 | N = 54 | |||
| Primary outcome | ||||
| Target achievement rate of BMI < 25 or > 15% reduction of BW, N (%) | 36 (62.1%) | 34 (63%) | 0.93 (0.58–1.49) | 0.772 |
| Components of primary outcome | ||||
| BMI < 25 kg/m2, N (%) | 23 (39.7%) | 19 (35.2%) | 0.82 (0.45–1.5) | 0.518 |
| > 15% BW reduction, N (%) | 28 (48.3%) | 28 (51.9%) | 0.99 (0.58–1.67) | 0.958 |
| Secondary outcomes | ||||
| Adherence to dietary therapy, N (%) | 56 (96.6%) | 52 (96.3%) | 0.943 | |
| Adherence to exercise therapy, N (%) | 22 (37.9%) | 20 (37%) | 0.923 | |
Note: Data are presented as n (%), or mean ± SD.
Abbreviations: BMI, body mass index; BW, body weight; CI, confidence interval; HR, hazard ratio.
Adherence to the study interventions was generally high. Dietary adherence was excellent and comparable between the groups (96.6% vs. 96.3%; p = 0.943). In contrast, exercise adherence was more modest but similar between the groups (37.9% vs. 37.0%; p = 0.923; Table 2). Adherence to tirzepatide therapy was excellent in both groups, with > 98% of prescribed doses administered (assessed by self‐report and injection records).
3.3. Changes in Body Composition
Changes in body composition are summarised in Table 3. Both groups demonstrated progressive weight loss throughout the 6‐month period. At 4 weeks, the mean body weight reduction was −3.9 ± 1.4 kg (−4.9%) in the 2.5 mg group and −3.8 ± 1.6 kg (−4.7%) in the 5 mg group (p = 0.811). At 6 months, the mean body weight reduction was −12.3 ± 3.5 kg (−15.3%) in the 2.5 mg group and −13.0 ± 3.6 kg (−16.1%) in the 5 mg group (p = 0.563). BMI decreased by −4.7 ± 2.2 kg/m2 (−15.0%) and −5.0 ± 1.8 kg/m2 (−16.1%) in the 2.5 and 5 mg groups, respectively (p = 0.457). Skeletal muscle mass decreased by −4.1% (2.5 mg group) versus −5.6% (5 mg group) (p = 0.105), and bone mass decreased by −7.5% versus −7.8%, respectively (p = 0.637). These reductions showed no significant between‐group differences at any measured time point.
TABLE 3.
Changes in body weight and body composition parameters from baseline in the groups.
| Variables | Change from baseline (%) | p | |
|---|---|---|---|
| 2.5 mg | 5 mg | ||
| N = 58 | N = 54 | ||
| Body weight (kg) | |||
| 1 months | −3.9 ± 1.4 (−4.9%) | −3.8 ± 1.6 (−4.7%) | 0.811 |
| 3 months | −7.7 ± 2.8 (−9.5%) | −8.6 ± 2.9 (−10.9%) | 0.2 |
| 6 months | −12.3 ± 3.5 (−15.3%) | −13 ± 3.6 (−16.1%) | 0.563 |
| Body mass index (kg/m2) | |||
| 1 months | −1.5 ± 1.2 (−4.7%) | −1.5 ± 1.2 (−4.8%) | 0.823 |
| 3 months | −2.9 ± 1.7 (−9.6%) | −3.4 ± 1.8 (−10.9%) | 0.149 |
| 6 months | −4.7 ± 2.2 (−15%) | −5.0 ± 1.8 (−16.1%) | 0.457 |
| Skeletal muscle mass (kg) | |||
| 1 months | −0.46 ± 0.68 (−1.2%) | −0.6 ± 0.77 (−1.5%) | 0.427 |
| 3 months | −1.1 ± 0.68 (−2.9%) | −2.16 ± 0.77 (−4.6%) | 0.06 |
| 6 months | −1.72 ± 1.3 (−4.1%) | −2.48 ± 1.58 (−5.6%) | 0.105 |
| Bone mass (kg) | |||
| 1 months | −0.05 ± 0.22 (−1.7%) | −0.05 ± 0.21 (−1.6%) | 0.898 |
| 3 months | −0.12 ± 0.34 (−4.5%) | −0.13 ± 0.36 (−4.8%) | 0.559 |
| 6 months | −0.2 ± 0.44 (−7.5%) | −0.21 ± 0.46 (−7.8%) | 0.637 |
Note: Data are presented as n (%), or mean ± SD.
Abbreviations: CI, confidence interval; HR, hazard ratio.
Longitudinal changes in body composition parameters are illustrated in Figure 1. The trajectories were nearly parallel between the two groups, with no clinically meaningful differences.
FIGURE 1.

Mean changes from baseline in (A) body weight, (B) BMI, (C) skeletal muscle mass, and (D) bone mass over 6 months. Data are shown as mean ± standard deviation. p‐Values represent between‐group comparisons at 6 months.
3.4. Multivariate Analyses of Primary Outcome
Univariate and multivariate analyses of predictors for achieving the primary outcome are presented in Table 4. The full model including all candidate variables is presented in Table S1. In the multivariate analysis, a baseline BMI > 30 kg/m2 was identified as an independent negative predictor of achieving the primary outcome.
TABLE 4.
Hazard ratios for the primary outcome in the parsimonious multivariable model.
| Univariate | Multivariate | |||
|---|---|---|---|---|
| HR (95% CI) | p | HR (95% CI) | p | |
| Age > 50 years | 0.97 (0.57–1.66) | 0.921 | ||
| Male | 0.63 (0.35–1.16) | 0.138 | 0.81 (0.43–1.51) | 0.504 |
| BMI > 30 kg/m2 | 0.39 (0.24–0.65) | < 0.001 | 0.45 (0.26–0.77) | 0.003 |
| Dietary adherence | 3.67 (0.51–26.5) | 0.197 | 4.15 (0.57–30) | 0.159 |
| Exercise adherence | 1.56 (0.97–2.5) | 0.066 | 1.46 (0.91–2.35) | 0.114 |
Abbreviations: BMI, body mass index; CI, confidence interval; HR, hazard ratio.
3.5. Adverse Events and Safety
The incidence of adverse events is shown in Table 5. Adverse events were more frequent in the 5 mg group than in the 2.5 mg group (50% vs. 35%). No serious adverse events occurred in either group. Discontinuation due to adverse events was uncommon (6.9% vs. 3.3%). Gastrointestinal symptoms, including nausea, vomiting, dyspepsia, diarrhoea, and constipation, were more commonly reported in the 5 mg group. No cases of pancreatitis or severe hypoglycaemia were observed in either treatment arm.
TABLE 5.
Adverse events and safety between the groups.
| 2.5 mg | 5 mg | |
|---|---|---|
| N | 60 | 58 |
| Patients with ≥ 1 adverse events | 21 (35%) | 29 (50%) |
| Patients with ≥ 1 serious adverse events | 0 | 0 |
| Adverse events leading to treatment discontinuation | 2 (3.3%) | 4 (6.9%) |
| Adverse events | ||
| Nausea | 4 (6.7%) | 7 (12.1%) |
| Vomiting | 2 (3.3%) | 4 (6.9%) |
| Dyspepsia | 2 (3.3%) | 3 (5.2%) |
| Diarrhoea | 2 (3.3%) | 0 |
| Constipation | 10 (16.7%) | 11 (19%) |
| Injection site reaction | 0 | 1 (1.7%) |
| Headache | 0 | 2 (3.4%) |
| Alopecia | 2 (3.3%) | 2 (3.4%) |
4. Discussion
This study provides the first prospective comparative data on the effectiveness of continuous low‐dose tirzepatide (2.5 mg vs. 5 mg once weekly) combined with lifestyle intervention in non‐diabetic adults with obesity. The principal findings demonstrate that both doses produced clinically meaningful weight reduction of approximately 15%–16% at 6 months, with over 60% of participants in each group achieving the prespecified primary outcome (BMI < 25 kg/m2 or > 15% body weight loss). Adherence to dietary and exercise recommendations was comparable between groups, and longitudinal changes in body composition—including body weight, BMI, skeletal muscle mass, and bone mass—showed no clinically or statistically significant differences.
The absence of meaningful differences in efficacy between the 2.5 and 5 mg doses in this cohort suggests that the starting dose of 2.5 mg tirzepatide may confer substantial benefit in non‐diabetic obese individuals, particularly when paired with structured lifestyle support [9]. It is important to note that although no statistically significant difference was observed between the 2.5 and 5 mg groups, this study was not designed as an equivalence or non‐inferiority trial. Therefore, our findings should be interpreted as the absence of evidence for a difference rather than evidence of equivalence between the two doses. This observation aligns with emerging real‐world evidence, in which many patients remain on lower doses due to tolerability, cost, or supply constraints yet still achieve robust weight loss.
These results are broadly consistent with the robust efficacy of tirzepatide observed in the SURMOUNT programme. However, the SURMOUNT‐1 trial demonstrated clear dose‐dependent weight loss at 72 weeks (mean reductions of 16.0%, 21.4%, and 22.5% with 5 mg, 10 mg, and 15 mg, respectively), with longer follow‐up and higher maintenance doses [2]. Several factors likely contributed to the comparable outcomes between the 2.5 and 5 mg groups in the present study. First, the 6‐month observation period may have been too short to fully capture dose‐dependent divergence, as weight loss trajectories with tirzepatide often continue to separate beyond the initial months, particularly with escalation to higher doses. Second, the relatively modest baseline BMI (approximately 30.8 kg/m2) in this population may have attenuated potential differences, as greater absolute weight loss is typically observed in individuals with higher baseline adiposity. Third, the dual GIP/GLP‐1 receptor agonism of tirzepatide exerts potent effects on appetite suppression and energy balance even at low doses, with substantial early weight reduction frequently observed during the initial titration phase.
Notably, these favourable outcomes with low‐dose tirzepatide may be partly attributable to ethnic differences in treatment response. Recent meta‐analyses have shown that Asian patients, including Japanese and Chinese populations, tend to achieve greater body weight reduction with tirzepatide compared to non‐Asian (primarily Caucasian) patients, despite having lower baseline BMI [10]. In dedicated Asian trials such as SURMOUNT‐J (Japanese) and SURMOUNT‐CN (Chinese), substantial weight loss was observed even with higher doses, often comparable to or exceeding results from global trials (e.g., mean weight loss of ~17.8%–22.7% in SURMOUNT‐J with higher doses) [11, 12] This enhanced responsiveness in Asians—potentially related to differences in body composition, visceral adiposity, and incretin physiology—likely contributes to the clinically meaningful weight loss achieved with only 2.5 or 5 mg in the present study. These findings support the possibility that lower maintenance doses may be particularly effective in Asian populations.
Given the greater weight loss responsiveness observed in Asian populations, future studies should evaluate whether low‐dose tirzepatide regimens (2.5 or 5 mg) can also deliver clinically meaningful benefits in non‐Asian populations. Such investigations would help determine if dose minimisation strategies can be more broadly applied across ethnic groups while maintaining efficacy and improving tolerability.
Individual variability in response represents another important consideration. Post hoc analyses of SURMOUNT‐1 identified a subgroup of late responders (approximately 18% of participants with < 5% weight loss at week 12), who were more likely to be male and have higher baseline body weight, BMI, and waist circumference [13]. Although the majority of these late responders ultimately achieved clinically meaningful weight loss by week 72 with continued treatment and dose escalation, their early trajectory was blunted. In the current study, dose assignment after 4 weeks was guided by shared decision‐making; thus, the 5 mg group may have been enriched with patients exhibiting a less robust initial response to 2.5 mg. Baseline characteristics in the 5 mg group trended towards slightly higher body weight, BMI, and hypertension prevalence—factors previously associated with attenuated early response—potentially contributing to the observed similarity in outcomes.
Separate analyses of the individual components of the primary endpoint revealed that achievement of BMI < 25 kg/m2 was numerically higher in the 2.5 mg group (39.7% vs. 35.2%), whereas the proportion of participants achieving > 15% body weight reduction was slightly higher in the 5 mg group (48.3% vs. 51.9%). These modest differences are likely explained, at least in part, by the slightly higher mean baseline body weight in the 5 mg group (approximately 1.8 kg). Individuals with higher baseline body weight require a greater absolute reduction to reach a BMI < 25 kg/m2, which may have placed the 2.5 mg group at a relative advantage for this component. In contrast, the 5 mg group showed a small numerical advantage in meeting the relative weight loss threshold (> 15%). Additional factors, including differences in body composition, behavioural adherence, or unmeasured confounders associated with the shared decision‐making process for dose assignment, may also have contributed. Nevertheless, these small divergences did not translate into a statistically significant difference in the composite primary outcome, supporting the overall similarity in efficacy between the two low‐dose regimens.
The integration of structured lifestyle intervention likely amplified the efficacy of low‐dose tirzepatide. Nearly all participants achieved high adherence to individualised dietary counselling emphasising guideline‐concordant energy intake and macronutrient balance. In contrast, exercise adherence remained lower (~38%), consistent with known barriers such as time constraints, comorbidities, and deconditioning. High dietary adherence, combined with tirzepatide‐mediated appetite suppression (reported in > 90% of participants), likely enhanced caloric deficit and contributed to the robust weight loss observed even at the 2.5 mg dose. This synergistic effect may explain why the magnitude of weight reduction in this 6‐month study approached that seen in the 5 mg arm of longer SURMOUNT trials.
Regarding body composition, the observed reductions in skeletal muscle mass (−4.1% and −5.6%) and bone mass (−7.5% and −7.8%) were relatively modest, particularly given the lower doses and shorter observation period compared with previous trials [14]. The modest changes observed here are consistent with the lower doses and shorter duration used in our study. While the proportion of weight lost as fat versus lean mass appears generally preserved across tirzepatide studies (roughly 75% fat mass), longer‐term data on bone health and fracture risk remain limited and warrant further investigation.
This study has several strengths. It was conducted as a multicentre prospective cohort study in a real‐world clinical setting, with well‐balanced baseline characteristics between groups and comprehensive serial evaluation of body composition using bioelectrical impedance analysis. This study also has several limitations. First, the treatment allocation was not randomised but based on shared decision‐making after 4 weeks of 2.5 mg tirzepatide. Consequently, patients with a less robust initial response or poorer tolerability may have been more likely to escalate to 5 mg. This potential selection bias and confounding by indication could have attenuated possible differences between the two doses. To partially mitigate this concern, we have provided 4‐week weight change data showing similar early responses between the eventual groups. Nevertheless, we cannot exclude the possibility that the observed similarity in 6‐month outcomes partly reflects this allocation process rather than true dose equivalence. Second, the moderate sample size limits statistical power and generalisability. We acknowledge that the multivariable analysis may be subject to overfitting due to the relatively small number of events. Therefore, the predictor findings should be interpreted with caution. Third, the lack of a placebo‐controlled or higher‐dose control group makes it difficult to fully isolate the specific contribution of each tirzepatide dose. Fourth, no formal a priori sample size calculation was performed for this observational cohort study. The confidence intervals around the primary endpoint are relatively wide and do not exclude clinically meaningful differences between the two doses. Therefore, the lack of statistical significance may reflect insufficient statistical power (Type II error) rather than true equivalence between the 2.5 and 5 mg regimens. Fifth, the 6‐month follow‐up period is relatively short and precludes assessment of the long‐term durability of weight loss, body composition changes, and safety. Finally, as an observational study, unmeasured confounding factors cannot be entirely excluded.
5. Conclusions
Low‐dose tirzepatide (2.5 or 5 mg) combined with lifestyle modification was associated with similar weight loss outcomes at 6 months between the two doses in this cohort. These findings, observed in a Japanese population, suggest the potential utility of low‐dose strategies, especially in Asian patients who may derive greater relative benefit. Larger, longer‐term randomised controlled trials are needed to confirm these observations and to establish optimal dosing strategies.
Funding
The authors have nothing to report.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Table S1: Hazard ratio for primary outcome.
Amioka M., Amioka H., Kinoshita H., Sairaku A., and Nakano Y., “Low‐Dose Tirzepatide for Obesity: Comparative Efficacy of 2.5 mg Versus 5 mg in Non‐Diabetic Japanese Adults,” Diabetes, Obesity and Metabolism 28, no. 10 (2026): 9395–9402, 10.1111/dom.71162.
Handling Editor: Richard Donnelly
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request. The data are not publicly available due to privacy or ethical restrictions.
References
- 1. Powell‐Wiley T. M., Poirier P., Burke L. E., et al., “Obesity and Cardiovascular Disease: A Scientific Statement From the American Heart Association,” Circulation 143, no. 21 (2021): e984–e1010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Jastreboff A. M., Aronne L. J., Ahmad N. N., et al., “Tirzepatide Once Weekly for the Treatment of Obesity,” New England Journal of Medicine 387, no. 3 (2022): 205–216. [DOI] [PubMed] [Google Scholar]
- 3. Hunter Gibble T., Grabner E. M., Masi L. M., et al., “Real‐World Use of Tirzepatide Among Individuals Without Type 2 Diabetes,” Diabetes, Obesity and Metabolism 27, no. 6 (2025): 3185–3194. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Japan Society for the Study of Obesity , Guidelines for the Management of Obesity Disease 2022 (Life Science Publishing, 2022). [Google Scholar]
- 5. Garber C. E., Blissmer B., Deschenes M. R., et al., “American College of Sports Medicine Position Stand. Quantity and Quality of Exercise for Developing and Maintaining Cardiorespiratory, Musculoskeletal, and Neuromotor Fitness in Apparently Healthy Adults: Guidance for Prescribing Exercise,” Medicine and Science in Sports and Exercise 43, no. 7 (2011): 1334–1359. [DOI] [PubMed] [Google Scholar]
- 6. American Diabetes Association Professional Practice Committee , “8. Obesity and Weight Management for the Prevention and Treatment of Type 2 Diabetes: Standards of Care in Diabetes‐2025,” Diabetes Care 48, no. 1 Suppl 1 (2025): S167–S180, 10.2337/dc25-S008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Garvey W. T., Frias J. P., Jastreboff A. M., et al., “Tirzepatide Once Weekly for the Treatment of Obesity in People With Type 2 Diabetes (SURMOUNT‐2): A Double‐Blind, Randomised, Phase 3 Trial,” Lancet 402, no. 10402 (2023): 613–626. [DOI] [PubMed] [Google Scholar]
- 8. Looney D. P., Williams L. J. B., Roberts M. D., et al., “Reliability, Biological Variability, and Accuracy of Multi‐Frequency Bioelectrical Impedance Analysis,” Frontiers in Nutrition 11 (2024): 1491931. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Hunter Gibble T., Chinthammit C., Ward J. M., et al., “Real‐World Use of Tirzepatide Among Individuals Without Evidence of Type 2 Diabetes: Results From the Veradigm Database,” Diabetes, Obesity & Metabolism 27, no. 6 (2025): 3185–3194. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Cui Y., Yao J., Qiu X., et al., “Comparative Efficacy and Safety of Tirzepatide in Asians and Non‐Asians With Type 2 Diabetes Mellitus: A Systematic Review and Meta‐Analysis,” Diabetes Therapy 15 (2024): 781–799. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Kadowaki T., Kiyosue A., Shingaki T., Oura T., and Yokote K., “Efficacy and Safety of Once‐Weekly Tirzepatide in Japanese Patients With Obesity Disease (SURMOUNT‐J): A Multicentre, Randomised, Double‐Blind, Placebo‐Controlled Phase 3 Trial,” Lancet Diabetes and Endocrinology 13 (2025): 384–396. [DOI] [PubMed] [Google Scholar]
- 12. Zhao L., Cheng Z., Lu Y., et al., “Tirzepatide for Weight Reduction in Chinese Adults With Obesity: The SURMOUNT‐CN Randomized Clinical Trial,” Journal of the American Medical Association 332, no. 1 (2024): 39–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Ard J., Lee C. J., Gudzune K., et al., “Weight Reduction Over Time in Tirzepatide‐Treated Participants by Early Weight Loss Response: Post Hoc Analysis in SURMOUNT‐1,” Diabetes, Obesity & Metabolism 27 (2025): 5064–5071. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Look M., Dunn J. P., Kushner R. F., et al., “Body Composition Changes During Weight Reduction With Tirzepatide in the SURMOUNT‐1 Study of Adults With Obesity or Overweight,” Diabetes, Obesity and Metabolism 27, no. 5 (2025): 2720–2729. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: Hazard ratio for primary outcome.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request. The data are not publicly available due to privacy or ethical restrictions.
