Comparative analysis of basal metabolic rate measurement methods in overweight and obese individuals: A retrospective study

Abstract

The global prevalence of overweight and obesity is on the rise, presenting significant health challenges worldwide. Obesity is associated with various chronic diseases and imposes substantial economic burdens on society. Accurate assessment of basal metabolic rate (BMR) is crucial for effective weight management strategies. This retrospective study, conducted at Baskent University Hospital between October 2019 and October 2023, analyzed data from 133 overweight and obese individuals. Various methods including indirect calorimetry (IC), predictive equations (Harris–Benedict and Mifflin–St Jeor), and bioelectrical impedance analysis (BIA) were used to estimate BMR. Additionally, demographic, clinical, and body composition data were recorded. The mean BMR measured using IC was 1581 ± 322 kcal/day, which was significantly lower than the BMR estimated by other methods such as BIA (1765.8 ± 344.09 kcal/day), Harris–Benedict (1787.64 ± 341.4 kcal/day), and Mifflin–St Jeor equations (1690.08 ± 296.36 kcal/day) (P < .001). Among the predictive equations, the Mifflin–St Jeor method provided BMR estimates closest to the gold standard IC. When BMR measurement methods were compared to IC, 36.8% of measurements with the Harris–Benedict equation method, 50.4% with the Mifflin–St Jeor equation method, and 36.1% with the BIA method were within ± 10% agreement with IC measurements. Significant correlations were found between BMR and body composition parameters such as fat-free mass, muscle mass, and fat mass (R = 0.681, P < .001; R = 0.699, P < .001; R = 0.595, P < .001, respectively). Regression analysis identified that variables such as weight, height, body mass index, and muscle mass significantly predicted BMR measured by IC, accounting for 69.1% of the variance. This study underscores the challenges in assessing BMR in overweight and obese individuals. While IC remains the gold standard, predictive equations and BIA offer alternative methods. The Mifflin–St Jeor equation emerged as a practical option, closely aligning with IC results. However, discrepancies between methods and the influence of body composition highlight the importance of individualized approaches to BMR assessment in weight management strategies.

1. Introduction

The worldwide occurrence of overweight and obesity is steadily increasing, with over 2 billion individuals currently categorized as having excess body weight. This represents roughly 30% of the global population.^[1] Obesity and its associated disorders have emerged as significant health issues on a global scale.^[2] Obesity poses a significant health problem as it greatly raises the likelihood of developing diseases like type 2 diabetes mellitus, fatty liver disease, hypertension, coronary artery disease, dementia, osteoarthritis, obstructive sleep apnea, and various cancers. Consequently, it leads to a decrease in both quality of life and life expectancy.^[3] Obesity not only contributes to morbidity but also imposes significant economic burdens on society due to the chronic diseases associated with it.^[4] The primary contributors to obesity are excessive and unhealthy eating habits, coupled with physical inactivity.^[5] Dietary management is a crucial component of weight control, aiming to achieve a negative energy balance at a fundamental level.^[6] The total daily energy demands consist of 3 components: basal metabolic rate (BMR), energy expended for the thermic effect of meals, and energy due to physical activity.^[7] BMR represents the amount of energy needed on a daily basis to sustain fundamental biological activities.^[8] It represents the primary source of energy expenditure and serves as a crucial factor in estimating daily energy needs.^[9,10] A review of the literature suggests that accurately predicting BMR is crucial for determining energy requirements and designing effective weight management strategies, particularly in individuals with severe obesity.^[11–13] BMR fluctuates based on factors such as age, gender, weight, height, and is intricately linked to an individual’s body composition.^[14] The measurement of BMR can be conducted using either direct or indirect calorimetric techniques. Direct calorimetry directly quantifies heat output, whereas indirect calorimetry (IC) quantifies oxygen intake and carbon dioxide production, which are subsequently translated into their respective energy equivalents. Additionally, predictive equations can also be used to estimate BMR in adults.^[10]

The efficacy of these techniques for assessing BMR has been evaluated across various patient cohorts.^[15,16] While IC being the gold standard, standardized equations like the Harris–Benedict and Mifflin equations offer convenient application but exhibit limited reliability.^[11,17] A systematic analysis of predictive equations in people with overweight and obesity revealed that the Harris–Benedict equation demonstrated the greatest accuracy for individuals with overweight and obesity, while the Lazzer equation showed the lowest bias for individuals with obesity (BMI > 30 kg/m²).^[18] In another systematic review, it was found that no single prediction equation provides accurate and precise resting energy expenditure estimates in all obese adults, and the review suggested using Mifflin equations for overweight and obese adults.^[19] The findings of a recent systematic review and meta-analysis, which synthesized results from multiple studies, demonstrated that the WHO and Harris & Benedict equations were the most accurate and precise in predicting BMR among individuals with severe obesity.^[20]

Despite the studies in the literature, confusion still persists regarding the most accurate methods for measuring BMR across different populations. The aim of this study was to assess the consistency between BMR measured via IC and estimated BMR determined by bioelectrical impedance analysis (BIA), as well as through the Harris–Benedict and Mifflin–St Jeor equations. Furthermore, an investigation into the correlation between BMR measured using these methods and biochemical, demographic, and body composition factors was also conducted.

2. Methods

This retrospective study was conducted at Baskent University Hospital between October 2019 and October 2023. The study retrospectively examined the data of 133 individuals who were either obese or overweight. Patient demographics including age, gender, weight, height, body mass index (BMI), and lifestyle information were recorded. By dividing the weight (in kilograms) by the square of the height (in meters), BMI was calculated. The hemogram and blood biochemistry parameters were assessed at 8 am. The following parameters were included in the evaluation: fasting plasma glucose, alanine aminotransferase, and creatinine. The study was approved by the Institutional Ethics Committee of Baskent University and it was carried out in compliance with the ethical principles outlined in the Declaration of Helsinki (KA24/171).

2.1. Bioelectrical impedance analysis (BIA)

A Tanita (BC-420MA, Tanita Corp., Tokyo, Japan) brand BIA device was used to assess the body composition of the patients. This device was utilized to measure patients’ body fat percentage, fat-free mass, fat mass, muscle mass, body water content, and metabolic rate. Measurements were taken via electrodes placed on the soles of the patients’ feet after they removed their shoes and socks.

2.2. Indirect calorimetry (IC)

In this study, basal metabolic rate (BMR) was measured using IC (The Fitmate™, Cosmed, Rome, Italy) under controlled conditions to ensure accurate and reliable results. Participants were required to fast for at least 12 hours prior to the measurement to eliminate the thermic effect of food. Measurements were conducted in a thermoneutral environment, ensuring a stable temperature to avoid any influence on metabolic rate. Participants were asked to rest in a supine position for at least 30 minutes before the measurement to ensure they were in a true basal state. The calorimetry device measured oxygen consumption (VO₂) and carbon dioxide production (VCO₂) over a period of 20 to 30 minutes. By adhering to these stringent conditions, we ensured that the BMR values obtained accurately reflected the participants’ BMRs, devoid of external influences such as recent physical activity or dietary intake.

2.3. Harris–Benedict equation

BMR was estimated using the Harris–Benedict equation,^[21] which takes into account age, gender, weight, and height of the participants. The equation used for men was:

$${\displaystyle \begin{array}{llll}{\displaystyle BMRmen=}& 88.362+(13.397\times \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{k}\mathrm{g})& +(4.799\times \mathrm{h}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{c}\mathrm{m})& -(5.677\times \mathrm{a}\mathrm{g}\mathrm{e}\mathrm{i}\mathrm{n}\mathrm{y}\mathrm{e}\mathrm{a}\mathrm{r}\mathrm{s})\end{array}}$$

For women, the equation used was:

$${\displaystyle \begin{array}{llll}{\displaystyle BMRwomen=}& 447.593+(9.247\times \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{k}\mathrm{g})& +(3.098\times \mathrm{h}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{c}\mathrm{m})& -(4.330\times \mathrm{a}\mathrm{g}\mathrm{e}\mathrm{i}\mathrm{n}\mathrm{y}\mathrm{e}\mathrm{a}\mathrm{r}\mathrm{s})\end{array}}$$

2.4. Mifflin–St Jeor equation

BMR was estimated using the Mifflin–St Jeor equation,^[22] which takes into account age, gender, weight, and height of the participants. The equations used for men and women are as follows:

For men:

$${\displaystyle \begin{array}{lll}{\displaystyle BMRmen=}& (10\times \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{k}\mathrm{g})+(6.25\times \mathrm{h}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{c}\mathrm{m})& -(5\times \mathrm{a}\mathrm{g}\mathrm{e}\mathrm{i}\mathrm{n}\mathrm{y}\mathrm{e}\mathrm{a}\mathrm{r}\mathrm{s})+5\end{array}}$$

For women:

$${\displaystyle \begin{array}{lll}{\displaystyle BMRwomen=}& (10\times \mathrm{w}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{k}\mathrm{g})+(6.25\times \mathrm{h}\mathrm{e}\mathrm{i}\mathrm{g}\mathrm{h}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{c}\mathrm{m})& -(5\times \mathrm{a}\mathrm{g}\mathrm{e}\mathrm{i}\mathrm{n}\mathrm{y}\mathrm{e}\mathrm{a}\mathrm{r}\mathrm{s})-161\end{array}}$$

2.5. Statistical analysis

The data were analyzed using the SPSS (Statistical Package for Social Sciences) 22.0 package program. In descriptive analyses, frequency data were presented as number (n) and percentage (%), while numerical data were presented as mean ± standard deviation (minimum–maximum). The normal distribution of numerical data was assessed through histogram and Q–Q plot graphs in the sample size and data distribution. The distribution of normally distributed numerical data in dependent 2 groups was examined using the Paired samples T test. The degree of agreement between the gold standard and other BMR measurement methods was evaluated using the Bland–Altman plot. The agreement limits were defined as mean ± 2 SD. The relationship between normally distributed 2 numerical variables was examined using Pearson correlation analysis. Univariate and backward multivariate linear regression analysis were conducted to assess the predictive properties of independent predictors on BMR parameters. Correlation relationships were classified as follows: R = 0.05–0.30 for low or weak correlation, R = 0.30–0.40 for low-moderate correlation, R = 0.40–0.60 for moderate correlation, R = 0.60–0.70 for strong correlation, R = 0.70–0.75 for very strong correlation, and R = 0.75–1.00 for perfect correlation. A statistical significance level of P < .05 was considered for all tests.^[23]

3. Results

The study included 133 overweight and obese participants, with 78.2% (n = 104) being female. The average age was 42.26 ± 12.13 years (ranging from 18–80 years), and the average BMI was 35.54 ± 5.57 kg/m² (ranging from 25.1–47.00 kg/m²). The mean BMR assessed using IC was 1581.00 ± 322.00 kcal/day (range: 658–2705 kcal/day). BMR measured with BIA was higher, averaging 1765.80 ± 344.09 kcal/day (range: 1219–2812 kcal/day). The fat percentage measured by BIA averaged 41.66 ± 6.80%, with fat-free mass at 56.59 ± 11.28 kg, muscle mass at 53.77 ± 10.64 kg, and bone mass at 3.07 ± 2.74 kg (Table 1).

Table 1.

The demographic and clinical characteristics of patients.

Variables	Results (n = 133)
Gender (female); n(%)	104 (78.2)
Age (years)	42.26 ± 12.13 (18–80)
Fasting blood glucose (mg/dL)	94.89 ± 12.61 (75–148)
ALT (IU/L)	26.21 ± 15.24 (8–94)
Creatinine (mg/dL)	0.72 ± 0.14 (0.42–1.69)
Height (cm)	166.20 ± 8.83 (150–188)
Weight (kg)	98.65 ± 19.74 (66–153)
BMI (kg/m2)	35.54 ± 5.57 (25.1–47)
IC BMR (kcal)	1581 ± 322 (658–2705)
Harris–Benedict BMR (kcal)	1787.64 ± 341.4 (1212.6–2881.8)
Mifflin–St Jeor BMR (kcal)	1690.08 ± 296.36 (1066.5–2526.4)
BIA BMR (kcal)	1765.80 ± 344.09 (1219–2812)
Fat percentage (%)	41.66 ± 6.8 (20.8–54.9)
Fat free mass (kg)	56.59 ± 11.28 (31.2–88.1)
Muscle mass (kg)	53.77 ± 10.64 (39.2–83.8)
Fat mass (kg)	41.77 ± 12.59 (10.9–75)
Bone mass (kg)	3.07 ± 2.74 (2.1–34)

Mean ± SD (min–max).

BIA = bioelectrical impedance analysis, BMI = body mass index, BMR = basal metabolic rate, IC = indirect calorimetry.

Comparing BMR measurement methods (Table 2), IC yielded significantly lower BMR values than other methods (P < .001). The Harris–Benedict method calculated higher BMR values than both the Mifflin–St Jeor and BIA methods (P < .001 and P = .008, respectively). BIA also provided higher BMR values compared to Mifflin–St Jeor (P < .001). Figures 1, 2, and 3 illustrate the agreement results between BMR measured with IC and other methods. When evaluated against the gold standard IC, 36.8% of Harris–Benedict, 50.4% of Mifflin–St Jeor, and 36.1% of BIA measurements were within ± 10% of IC measurements (Table 3).

Table 2.

Comparison of BMR measurement methods.

	Mean ± SD	Correlation	P *
IC BMR (kcal/day) Harris–Benedict BMR (kcal/day)	1581 ± 322 1787.64 ± 341.401	0.744	<.001
IC BMR (kcal/day) Mifflin–St Jeor BMR (kcal/day)	1581 ± 322 1690.08 ± 296.36	0.779	<.001
IC BMR (kcal/day) BIA BMR (kcal/day)	1581 ± 322 1765 ± 344.09	0.740	<.001
Harris–Benedict BMR (kcal/day) Mifflin–St Jeor BMR (kcal/day)	1787.64 ± 341.401 1690.08 ± 296.36	0.983	<.001
Harris–Benedict BMR (kcal/day) BIA BMR (kcal/day)	1787.64 ± 341.401 1765 ± 344.09	0.963	.008
Mifflin–St Jeor BMR (kcal/day) BIA BMR (kcal/day)	1690.08 ± 296.36 1765 ± 344.09	0.957	<.001

Bold P values indicate statistical significance.

Mean ± SD (min–max).

BIA = bioelectrical impedance analysis, BMR = basal metabolic rate, IC = indirect calorimetry.

^* Paired samples T test.

Figure 1.

Bland–Altman plot of agreement between predicted BMR by the Harris–Benedict equation and BMR result by IC.

Figure 2.

Bland–Altman plot of agreement between predicted BMR by the Mifflin–St Jeor equation and BMR result by IC.

Figure 3.

Bland–Altman plot of agreement between predicted BMR by BIA and BMR result by IC.

Table 3.

Evaluation of measurement methods according to IC.

	Lower than IC Measurement; n (%)	±%10 IC; n (%)	Higher than IC measurement; n (%)
Harris–Benedict BMR	23 (17.3)	49 (36.8)	110 (82.7)
Mifflin–St Jeor BMR	36 (27.1)	67 (50.4)	97 (72.9)
BIA BMR	23 (17.3)	48 (36.1)	110 (82.7)

BIA = bioelectrical impedance analysis, BMR = basal metabolic rate, IC = indirect calorimetry.

Table 4 presents the relationships between BMR assessment methods and other variables. Significant positive correlations were found between BMR measured with IC and fat-free mass (R = 0.681, P < .001), muscle mass (R = 0.699, P < .001), and fat mass (R = 0.595, P < .001). Similar significant positive correlations were observed for Harris–Benedict, Mifflin–St Jeor, and BIA methods with various body composition metrics.

Table 4.

Relationship between BMR measurement methods and biochemical and clinical data.

		IC BMR (kcal)	Harris–Benedict BMR (kcal)	Mifflin–St Jeor BMR (kcal)	BIA BMR (kcal)
Age (years)	r P	−0.128 .143	−0.332 <.001	−0.350 <.001	−0.242 .005
Fasting blood glucose (mg/dL)	r P	0.196 .024	0.089 .308	0.088 .313	0.146 .093
ALT (IU/L)	r P	0.179 .040	0.272 .002	0.241 .005	0.290 .001
Creatinine (mg/dL)	r P	0.129 .137	0.290 .001	0.266 .002	0.356 <.001
Height (cm)	r P	0.631 <.001	0.761 <.001	0.793 <.001	0.787 <.001
Weight (kg)	r P	0.791 <.001	0.891 <.001	0.923 <.001	0.866 <.001
BMI (kg/m2)	r P	0.560 <.001	0.585 <.001	0.616 <.001	0.549 <.001
Fat percentage (%)	r P	0.237 .006	0.054 .0541	0.132 .131	−0.064 .466
Fat free mass (kg)	r P	0.681 <.001	0.911 <.001	0.895 <.001	0.964 <.001
Muscle mass (kg)	r P	0.699 <.001	0.921 <.001	0.909 <.001	0.984 <.001
Fat mass (kg)	r P	0.595 <.001	0.574 <.001	0.637 <.001	0.485 <.001
Bone mass (kg)	r P	0.172 .047	0.250 .004	0.246 .004	0.301 <.001

r = Pearson correlation coefficient.

Bold P values indicate statistical significance.

BIA = bioelectrical impedance analysis, BMI = body mass index, BMR = basal metabolic rate, IC = indirect calorimetry.

One-way regression analysis showed that age, height, weight, BMI, fat percentage, fat-free mass, muscle mass, fat mass, and bone mass were predictive factors of BMR (P < .005) (Table 5). The backward multivariate regression analysis revealed that the model encompassing all variables had an explanatory power of 69.1%, indicating these data as significant predictive factors of BMR (Table 6).

Table 5.

Univariate regression analysis results for IC BMR.

	B (% 95 CI)	Beta	t	P	Zero-order
Age	−3.387 (−7.937 to 1.162)	−0.128	−1.473	.143	−0.128
Height (cm)	22.996 (18.109–27.883)	0.631	9.308	<.001	0.631
Weight (kg)	12.905 (11.181–14.629)	0.791	14.809	<.001	0.791
BMI (kg/m2)	32.34 (24.068–40.162)	0.560	7.734	<.001	0.560
Fat percentage (%)	11.222 (3.274–19.170)	0.237	2.793	.006	0.237
Fat free mass (kg)	19.445 (15.835–23.055)	0.681	10.656	<.001	0.681
Muscle mass (kg)	21.13 (17.389–24.871)	0.699	11.172	<.001	0.699
Fat mass (kg)	15.216 (11.666–18.767)	0.595	8.478	<.001	0.595
Bone mass (kg)	20.186 (0.244–40.127)	0.172	2.002	.047	0.172

B: unstandardized coefficient, Beta: standardized coefficient.

Bold P values indicate statistical significance.

BMI = body mass index, BMR = basal metabolic rate.

Table 6.

Multivariate backward regression analysis results for IC BMR.

Model	R	R square	Adjusted R Square	Std. error of the estimate	R square change	F change	df1	df2	Sig. F change	Durbin Watson
1	0.831	0.691	0.668	185.513	0.691	30.522	9	123	<0.001
2	0.831	0.691	0.671	184.768	<0.001	0.007	1	123	0.935
3	0.831	0.691	0.673	184.087	<0.001	0.080	1	124	0.777
4	0.830	0.690	0.675	183.623	0.001	0.365	1	125	0.547
5	0.830	0.689	0.676	183.174	0.001	0.380	1	126	0.539
6	0.829	0.687	0.678	182.835	0.001	0.527	1	127	0.469	2.017

1. Predictors: (constant), bone mass (kg), fat mass (kg), age (years), height (cm), fat-free mass (kg), fat percentage (%), BMI (kg/m²), muscle mass (kg), weight (kg).

2. Predictors: (constant), bone mass (kg), fat mass (kg), age (years), height (cm), fat percentage (%), BMI (kg/m²), muscle mass (kg), weight (kg).

3. Predictors: (constant), bone mass (kg), fat mass (kg), age (years), height (cm), fat percentage (%), BMI (kg/m²), weight (kg).

4. Predictors: (constant), bone mass (kg), fat mass (kg), age (years), fat percentage (%), BMI (kg/m²), weight (kg).

5. Predictors: (constant), fat mass (kg), age (years), fat percentage (%), BMI (kg/m²), weight (kg).

6. Predictors: (constant), fat mass (kg), fat percentage (%), BMI (kg/m²), weight (kg).

BMI = body mass index; BMR = basal metabolic rate.

4. Discussion

The aim of this study is to compare different methodologies for estimating BMR with respect to biochemical parameters, body composition, and the gold standard IC method. The findings of this study shed light on the challenges and complexities of assessing BMR in overweight and obese individuals. The remarkable finding of the study was that alternative methods yielded higher BMR values in comparison to IC, recognized as the gold standard method. Among the methodologies discussed in the study, the Mifflin–St Jeor Equation method yielded the most closely aligned BMR estimation with IC.

As the prevalence of overweight and obesity continues to rise globally, it becomes increasingly important to understand the factors influencing BMR measurement in order to develop effective weight management strategies.^[24,25] While IC is considered the gold standard, predictive equations and BIA can be used to estimate BMR.^[24,26,27] IC is considered the gold standard for BMR measurement, boasting high validity. However, its widespread adoption is hindered by factors such as limited availability, high cost, time requirements, and the need for specialized personnel. As a result, weight loss interventions typically rely on prediction equations to estimate BMR.^[11,28,29] The literature extensively evaluates the adequacy and reliability of these methods.^{[11,12,29,30]} Luy et al found that in obese Filipinos with type 2 DM or prediabetes, both the Harris–Benedict equation and BIA have a tendency to overestimate BMR when compared to measurements acquired using IC.^[30] In our study, we observed results consistent with those reported in the study by Luy et al We found that BMR measured using predictive equations (Harris–Benedict and Mifflin equations) and BIA yielded significantly different results. Specifically, BMR measured with IC was consistently lower compared to estimates from predictive equations and BIA. This highlights the limitations of relying solely on predictive equations or BIA for assessing BMR, especially in overweight and obese populations. The discrepancies observed between BMR measurement methods underscore the importance of using a multimodal approach to assess energy expenditure in overweight and obese individuals. Because overestimating BMR could potentially hinder the success of weight loss programs, particularly among obese individuals. Literature indicates that these equations tend to overestimate BMR across various patient groups.^[31] In our study, we observed that the Mifflin–St Jeor method provided the BMR calculation closest to that obtained via IC. A systematic review revealed that the Mifflin–St Jeor equation exhibits a greater likelihood of predicting resting metabolic rate within 10% of the measured value when compared to other equations assessed. According to this systematic review, it is advised to use the Mifflin–St Jeor equation if a prediction formula is used.^[11] The results regarding the Mifflin–St Jeor equation in our study were in line with the findings reported in this systematic review.

BMR is primarily influenced by body composition, specifically fat-free mass and fat mass, as well as factors such as gender, age, physical activity, and nutritional status. Fat-free mass is the primary factor that determines BMR.^[14] Molnar et al find that fat-free mass primarily influences resting metabolic rate, although age, gender, and fat mass also significantly contribute to it.^[32] Similarly, Johnstone et al also found that factors contributing to the variability in BMR include fat-free mass, fat mass, age, and circulating thyroxine.^[33] In our study, we observed a noteworthy correlation between fat-free mass, muscle mass, fat mass, and the calculation methods of BMR. Univariate and multivariate regression analyses revealed that age, height, weight, BMI, fat percentage, fat-free mass, muscle mass, fat mass, and bone mass were predictive factors of BMR. These findings underscore the influence of body composition on BMR and highlight the importance of considering individual characteristics when estimating energy expenditure.

This study has several limitations that should be considered when interpreting the results. First of all, the retrospective design introduces inherent biases and makes it challenging to establish causal relationships between variables. The sample size might not encompass the full diversity within the overweight and obese population, potentially impacting the robustness of the findings. The study solely included overweight and obese participants. The results may not be applicable to individuals with normal weight or underweight conditions. Finally, while body composition variables were included in the analysis, other factors influencing body composition, such as hormonal status and lifestyle factors, were not fully accounted for.

In conclusion, our study highlights the challenges and complexities associated with assessing BMR in overweight and obese individuals. Although IC is considered the gold standard, alternative methods such as predictive equations and BIA offer alternative methods for estimating BMR. The Mifflin–St Jeor equation is notable for its results being closest to the IC and its practical usability. However, discrepancies between measurement methods and the influence of body composition underscore the need for a comprehensive approach to BMR assessment. By considering individual characteristics, clinicians can better tailor weight management strategies to meet the needs of overweight and obese individuals. Further research is warranted to explore additional factors that may influence BMR in this population and to validate the findings of this study in larger and more diverse cohorts.

Acknowledgments

We would like to express our deepest gratitude to “Prof Dr Okan Sefa Bakiner’’ for his invaluable guidance and encouragement throughout the development of this study.

Author contributions

Conceptualization: Baris Karagun.

Data curation: Baris Karagun.

Formal analysis: Baris Karagun.

Investigation: Baris Karagun.

Methodology: Baris Karagun.

Resources: Baris Karagun.

Supervision: Nuh Baklaci.

Validation: Baris Karagun.

Visualization: Baris Karagun, Nuh Baklaci.

Writing – original draft: Baris Karagun.

Writing – review & editing: Baris Karagun, Nuh Baklaci.