Genetic Polymorphisms Associated with Obesity in Southeast Asian Populations: A Systematic Review without Meta-Analysis Authors: Ubashini Vijakumaran, Nor Azian Abdul Murad, Muhammad Irfan Abdul Jalal, Chin Siok Fong, A Rahman A Jamal, Noraidatulakma Abdullah Published: 2025 ABSTRACT AbstractObesity is a growing global public health challenge, with genetic factors playing a crucial role in its development. This review synthesises findings from Southeast Asian studies to investigate the association between gene polymorphisms and obesity risk across various ethnic populations. A comprehensive search of three databases, PubMed, Scopus, and Web of Science, initially retrieved 2,021 articles, from which 25 studies were meticulously selected based on stringent inclusion and exclusion criteria. The quality of the studies was assessed through the Newcastle-Ottawa Scale (NOS), a risk bias tool. These studies encompass 8,312 participants and examined 33 single nucleotide polymorphisms (SNPs). UCP polymorphism demonstrated a significant association with overall adiposity (OR = 2.02, P = 0.01) in Malaysian women, while the rs659366 UCP2 was linked to weight gain in an Indonesian cohort. LEP variants were not significantly associated with obesity in Malaysians, and FTO variants showed mixed results, with rs9939609 (OR = 3.72, P = 0.009) and rs1421085 (OR = 3.22, P < 0.001) variants being associated with obesity and higher body mass index (BMI) in Indonesians, but no significant findings in Malaysians. These results emphasise the genetic diversity within Southeast Asia and the challenges in replicating genetic associations across populations. To address these inconsistencies and improve our understanding of obesity in Southeast Asia, there is a pressing need for more extensive and diverse cohort studies, complemented by comprehensive genome-wide association studies (GWAS), to identify robust obesity biomarkers in Southeast Asia.Keywords: obesity, overweight, gene polymorphism, single nucleotide polymorphisms, Southeast Asia Introduction Obesity is a complex metabolic disorder that has become a primary global health concern, with its prevalence rising steadily across diverse populations. As of 2022, approximately one in eight people worldwide is affected by obesity. The World Health Organization (2024) (1) reports that 2.5 billion adults are overweight, and 890 million individuals are classified as obese. While environmental factors undoubtedly play a significant role in this increase, obesity primarily results from the interplay of genetic predispositions and environmental influences. Genetic variants among individuals can affect their susceptibility and response to obesogenic environments (2). Typically, obesity is categorised into two types. Monogenic obesity is a rare condition caused solely by the impairment of genes inherited in a Mendelian pattern (3). Polygenic obesity, the more common type, results from the interaction between a myriad of genetic polymorphisms, epigenetic regulation, and environmental factors (4). The central nervous system and neuronal pathways that modulate food intake and subsequent energy metabolism are major regulators of body fat and weight in monogenic and polygenic obesity (2). The involvement of multiple pathways in obesity (5, 6) presents significant challenges in understanding the complex mechanisms underlying the condition. One of the most studied and critical among these is the leptin-melanocortin pathway, which regulates energy metabolism through approximately 60 genes expressed in the hypothalamic region (7). A key component of this pathway is the leptin (LEP) which plays a pivotal role in energy balance and appetite regulation. LEP is expressed in adipocytes and encodes the leptin hormone, which acts on the hypothalamus to convert proopiomelanocortin (POMC) into alpha-melanocyte stimulating hormone (α-MSH), which suppresses appetite and food intake (8). Individuals with leptin resistance have impaired signalling in the leptin-melanocortin pathway, leading to increased food intake and reduced energy expenditure (9). Additionally, children with impaired leptin-melanocortin signalling pathways also exhibit disruptive eating patterns from early childhood, leading to severe early-onset obesity (10). Leptin regulates intracellular pathways such as PI3K/ AKT/mTOR, JAK2/STAT3, and ERK/MAPK to control cell proliferation, differentiation, survival, migration, and invasion (11, 12). The MAPK pathway, including the extracellular signal-regulated kinase (ERK) 1/2, c-Jun N-terminal kinase (JNK), and p38 MAPK, also plays a crucial role in the pathogenesis of obesity by regulating adipogenesis, glucose homeostasis, and thermogenesis (13, 14). Notably, elevated expression of p38 MAPK in human adipose tissue has been linked to hypertrophic adipocytes and associated with elevated levels of insulin, glucose, and triglyceride (15). While these signalling pathways provide important insights into the biological mechanisms of obesity, understanding the genetic underpinnings is equally critical. Advances in molecular and population-based approaches have facilitated the identification of genetic variations linked to obesity risk. Initially, gene polymorphism was explored through candidate gene studies, which later expanded into more comprehensive genome-wide association studies (GWAS) (16). The first GWAS in 2007 reported the presence of a genetic polymorphism in the first intron of rs9930506 FTO that is positively associated with body mass index (BMI), (P = 8.6 × 10−7), hip circumference (in cm) (P = 3.4 × 10−8), and weight (in kg) (P = 9.1 × 10−7) (17). Another GWAS comprising 38,759 participants revealed that adults with homozygous A allele at rs9939609 FTO weight 3 kg more than those who do not have it, with 1.67 increased odds of obesity (18). To date, GWAS studies have discovered more than 300 loci associated with obesity traits (19). However, few genes, such as ADRB3 (20), MCR4 (21), BDNF (22), LEPR (23), FTO (24), PCSK1 (25), CNR1 (26), and PPARG (27), have replicable associations with obesity (2). A large meta-analysis reported 463 signals associated with body fat in 694,649 European individuals (27). Similarly, studies in Asian populations have identified several key loci; for instance, a GWAS in Taiwan reported 13 significant SNPs, including FTO and RALGAPA1, in 6,546,460 individuals (28), while six SNPs were reported in the Han Chinese population (29). In addition to these common variants identified through population-level studies, rare mutations have also been implicated in severe obesity cases. One such example is the ADCY3, which regulates MC4R trafficking. A homozygous frameshift mutation (c.3315del; p. Ile1106Serfs) in exon 21 of ADCY3 was recently reported in severely obese Pakistani children (30). Besides, a different ADCY3 variant, homozygous c.1102G>A (p. Asp368Asn), was identified through whole-exome sequencing in 10-year-old Turkish children (31), which has been associated with early-onset obesity, insulin resistance, and neurodevelopmental issues (31). Additional gene variants from the same pathway, such as semaphorin 3A-G (SEMA3A-G), plexinA1–4 (PLXNA1–4), and neuropilin 1–2 (NRP1–2), were identified in the adult obese population (32). In 2020, the FDA approved the first MC4R agonist, setmelanotide (RM-493), for the treatment of severe obesity (33, 34), marking a major milestone more than two decades after MC4R was first identified as a therapeutic target. This highlights the translational potential of genetic discoveries into clinical practice, as seen with the approval of setmelanotide for MC4R-related obesity. However, such pharmacogenomic advancements have largely been derived from studies conducted in Western populations, raising concerns about their relevance and applicability in other regions. The Southeast Asian region comprises Brunei, Myanmar, Cambodia, Timor-Leste, Indonesia, Laos, Malaysia, Philippines, Singapore, Thailand, and Vietnam (35). It has experienced a significant rise in obesity prevalence, projected to double from 2010 to 2030, affecting up to 45 million individuals (36). Given the region’s rich ethnic diversity, unique dietary patterns, and distinct genetic backgrounds, it remains unclear whether genetic risk factors and treatment responses observed in Western populations are applicable to Southeast Asians. This raises an important question: do population-specific genetic markers exist that better explain obesity risk and potentially guide personalised treatment strategies in this region? Building on this need for region-specific genetic data, a GWAS meta-analysis performed on Singaporeans, Malays, and Asian Indians confirmed the association between the FTO and obesity (37). The study successfully replicated 13 loci previously identified in European cohorts and reported three novel SNPs (rs2287019, rs2241423, and rs516175) linked to increased BMI in these Asian populations. Interestingly, 16 loci found in European GWAS were absent in Southeast Asians, which could be attributed to genetic heterogeneity. These findings underscore the importance of tailoring genetic research to local populations. Further supporting the relevance of FTO in Asian contexts, a meta-analysis of 26 studies concluded a significant association between FTO SNPs and obesity risk in the Chinese population (OR = 1.30; 95% CI: 1.19, 1.42; P < 0.001) (38). Similarly, in Asian Indians, a minor A allele of the FTO, rs9939609, was associated with increased obesity risk (OR = 1.15; 95% CI: 1.08, 1.21, P = 2.14 × 10−5 (39). In contrast to findings in other Asian populations, no association was reported between 31 variants of FTO SNPs and obesity traits in Malaysians (40). Similarly, although a meta-analysis from 39 studies revealed a significant association between rs1137101 LEPR and obesity under the homozygous model (GG vs AA: OR = 1.39; 95% CI: 1.12, 1.73, P = 0.003 in Caucasians and Asians (23), this association was not observed in the multi-ethnic Malaysian populations (41, 42). These inconsistencies across ethnic groups highlight the complexity of genetic predisposition to obesity, which may differ widely due to ancestry diversity, environmental interactions, and population sampling differences in Southeast Asia. Hence, this review aims to map the current landscape of genetic polymorphisms associated with obesity in the Southeast Asian population. By identifying established findings and highlighting knowledge gaps, we hope to support future research tailored to the region’s unique genetic and environmental context. Methods Search Strategy This systematic review followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 and a systematic review without meta-analysis (SWiM) guidelines. The review has been registered under the International Prospective Register of Systematic Reviews (PROSPERO) (ID: CRD42024575327). The protocol can be assessed through https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=575327. A combination of Medical Subject Headings (MeSH) terms was used for the search, including: 1) Adiposity; 2) Body Mass Index; 3) Obesity; 4) Overweight; 5) Gene Polymorphism; 6) Single Nucleotide Polymorphism; 7) SNPs; 8) Genome-Wide Association Studies; 9) GWAS; and 10) Southeast Asia. These terms were combined using Boolean operators (AND/OR) as follows: [(1 OR 2 OR 3 OR 4 OR 5 OR 6 OR 7 OR 8 OR 9) AND 10]. A total of 2,021 articles were retrieved from three electronic databases: Web of Science, PubMed, and Scopus. These articles were downloaded into EndNote 20.6 software, and 469 duplicate articles were removed. The titles and abstracts of the remaining studies were screened to select those relevant to gene polymorphism in the Southeast Asian population. The full texts of 55 studies were assessed for eligibility, and 25 were selected for review. The study retrieval and selection summary are presented in Figure 1 as a PRISMA flow chart. The PRISMA flow chart summarises the study retrieval, screening, and selection procedure Selection Criteria Full-text articles published in English within the past 15 years (2010–2024) were selected. The PECOS criteria for study inclusion and exclusion are detailed in Table 1. Only case-control, cross-sectional, and cohort studies were included; animal studies and clinical trials were excluded. The target population was Southeast Asians aged 18 to 70 years. Newborns, children, and individuals over 70 years old were excluded. Full texts of the studies were meticulously screened, and only those focusing on gene polymorphism related to obesity parameters were selected for data extraction. Table 1PECOS criteria of the studiesCriteriaInclusionExclusionPopulationSoutheast Asia participants19 to 69 years oldAll genders and ethnicitiesIn vitro cell modelAnimals< 18 years old> 70 years oldExposureCombination of genetic susceptibility (e.g. candidate genes, genome-wide association studies, polygenic risk scores, genetic risk scores, single nucleotide polymorphisms) and obesity risk parametersStudies that did not include both exposuresComparisonNon-obese participants-OutcomeObesity risk factor (weight-related anthropometric measurements [BMI, weight, waist circumference, waist-to-hip ratio, body fat])Studies that failed to report the desired outcomeStudy designCohort, case-control, cross-sectionalClinical trials, including randomised and non-randomised, animal studies PECOS criteria of the studies Data Extraction and Management Data extraction was performed from 25 studies that passed the eligibility screening criteria. The following information was extracted and tabulated in Table 2. The following information was extracted: Table 2Summary of study characteristics and main findings of the selected articlesGene and chromosome locationSNPs and allelesCountry (reference)Study design and sample sizeGenotyping methodObese definitionSample characteristicsAge (years)EthnicityFindingsHighest OR95% CIP-valueG: FTOL: 16q12.2rs1077128rs11643744rs7186521rs13334933rs16952517rs6499643rs4784323rs7206790rs9939973rs1421085rs1558902rs10852521rs16952522rs17817288rs1121980rs16945088rs17817449rs8050136rs9935401rs3751812rs9939609rs7190492rs7204609rs17218700rs11642841rs1861867rs11075994rs1421090rs17818902rs7191513Malaysia (40)Cross-sectionalN = 587Obesen = 158Non-obesen = 429Sequenom MassARRAY®iPLEX Platform(MALDI-TOF)≥ 30.0 kg m248.29 ± 9.89Malayrs17817288 was significantly associated with LDL-CNo significant difference in genotype or allelic frequency for SNPs between obese and normal groupsNRG: LEP A19GL: 7q32.1rs2167270Malaysia (41)Cross-sectionalN = 408Obesen = 190Non-obesen = 218PCR-RFLP> 27 kg/m252.4 ± 13.7Malay = 148Chinese = 177Indian = 83Not all SNPs are associated with obesity among MalaysiansOR = NRG: LEP G2548AL: 7q32.1rs7799039G: LEPR K109RL: 1p31.3rs1137100G: LEPR Q223RL:1p31.3rs1137101G: LEPG2548AL: 7q32.1rs7799039Malaysia (42)Cross-sectionalN = 185Obesen = 95Non-obesen = 87PCR-TaqMan> 29.9 kg/m220.84 ± 2.2Malay = 61Chinese = 45Indian = 56Others = 23No significant association was found between ethnic groups with LEP, G2548A and LEPR, Q223ALEPR polymorphism showed no association with plasma leptinOverweight and obese participants were in stage I of leptin resistanceLEPR, Q223R of AA genotype association with obesityOR = 2.50CI = 0.11, 59.9P = 0.50G: LEPRQ223RL: 1p31.3rs1137101G: ADRA2AL: 10q25.2rs553668Malaysia (44)Cross-sectionalN = 214Obesen = 142Non-obesen = 72PCR-RFLP≥ 25 kg/m226.27 ± 11.93Malay = 45Chinese = 116Indian = 53People with both ADRA2A, rs553668 GG and ACE I/D II significantly has lowest WHR, could be due to gene-gene interactionACE associated with central adiposityOR = 2.02CI = 0.87, 4.70P = 0.10G: ACEL: 17q23.3rs4646994G: ADRB2L: 5q32rs1042714Malaysia (45)Cross-sectionalN = 672Obesen = 17Non-obesen = 500PCR-Taqman SNP> 90 cm for men> 80 cm for women48.22 ± 10.05Malayrs1042714 was significantly associated with log-transformed HDL-C, DBP, and TGNo association found towards BMI and obesityOR = NRG: DRD2Taq1AL: 11q23.2rs1800497Malaysia (46)Case-controlN = 394Obesen = 67Non-obesen = 327PCR-RFLP≥ 25 kg/m252.4 ± 13.7Chinese = 308Indian = 86DRD2, Taq1 SNPs were not associated with adiposity and obesity, but have influenced eating behaviourDRD2, Taq1 A genotype association with WHROR = 1.19CI = 0.57, 2.47P = 0.64G: DRD2L: 11q23.2rs1079597G: DRD2Taq1DL: 11q23.2rs1800498G: ADIPOQL: 3q27.3rs17366568Malaysia (47)Cross sectionalN = 574Obesen = 150Non-obesen = 424Sequenom MassARRAYTaqMan PCR≥ 30 kg/m245.89 ± 6.34MalayADIPOQ rs17366568 significantly associated with obesity, while rs3774261 not significantly associatedGenotypes AG and AA of rs17366568 significantly higher in obese groupA allele in obese groupOR = 2.15CI = 1.13, 4.09P = 0.024rs3774261G: ACEL: 17q23.3Insertion/ deletion (I/D)Malaysia (48)Cross sectionalN = 219Obesen = 94Non-obesen = 123PCR-RFLP≥ 30 kg/m235.22 ± 10.10MalayACE gene I/D polymorphism is not associated with obesity and obesity-related phenotypesACE ID genotype association with obesityOR = 1.32CI = 0.75, 2.34P = 0.335G: FTOL: 16q12.2Rs9930506Malaysia (49)Cross-sectionalN =178Obesen = 79Non-obesen = 99PCR-RFLP≥ 27.5 kg m241.7 ± 0.9Chinese = 42Malay = 86Indian = 50rs9930506 (G) not associated with obesity in dominant model but associated with codominant modelrs9930506, GG, associated with obesity in codominant modelOR = 2.87CI = 1.14, 7.19P = 0.02G: IL1RAintron 2L: 2q14.1NRMalaysia (50)Case-controlN = 315Obesen = 54Non-obesen = 261PCR-RFLP≥ 25 kg/m220.6Malay = 23Chinese = 251Indian = 41Individuals with IL1RA I/II genotype or allele II showed higher risk of having adiposityIL1RA, I/IIOR = 12.21CI = 2.54, 58.79P = 0.002G: IL4L: 5q31.1NRG: LEP G2548AL: 7q32.1rs7799039Malaysia (51)Case-controlN = 249Obesen = 148Non-obesen = 101PCR-RFLP≥ 25 kg/m2Age: NRMalayNo significant single association between SNPs and BMI and obesityHaplotype AAG of G2548A, H1328080, and A19G reported to have significant association with obesityAAG haplotype association with obesityOR = 8.89CI = 1.59, 49.78P = 0.002G: LEP A19GL: 7q32.1rs2167270G: LeptinH1328080L: chr7128238282rs12535747G: UCP1L: 4q31.1rs1800592 (A/G)Malaysia (52)Case-controlN = 447Obesen = 111Non-obesen = 336PCR-RFLP≥ 25 kg/m224.66 ± 6.90Malay = 46Chinese = 339Indian = 62Lack of association between single UCP1/3 and obesityCombination of UCP1 A allele and UCP3 C allele associated with WHRCombination of UCP1 AA and UCP3 CC genotypes showed elevated BMIT allele had significantly less risk of obesity in ChineseOR = 5.69CI = 5.48, 1.00P = 5.04G: UCP3L: 11q13.4rs1800849 (T/C)G: UCP2L: 11q13.4Insertion/ deletion (I/D)Malaysia (53)Case-controlN = 926Obesen = 265Non-obesen = 661PCR-RFLP≥ 25 kg/m233.96 ± 8.51Malay = 102Chinese = 672Indian = 152UCP2 45-bp I/D polymorphism is associated with overall adiposity in Malaysian womenID genotype associated with overall adiposityOR = 2.02CI = 1.18, 3.45P = 0.01G: PPARαL162VL: 22q13.31NRMalaysia (54)Cross-sectionalN = 307Obesen = 18Non-obesen = 127PCR-RFLP≥ 25 kg/m253.3 ± 14.2Malay = 97Chinese = 85Indian = 55No significant association of SNPs with obesity and metabolic syndromePPARα, V162 allele carriers associate with high plasma IL-6 levelPPARα, L162V association with obesity and metabolic syndromeOR = 1.732CI = 1.34, 8.77P = 0.50G: PPARγ2C161TL: 6p21.31NRG: PPARδT294CL: 3p25.2NRG: FTOL:16q12.2rs9939609Indonesia (55)Case-controlN = 80Obesen = 38Non-obesen = 40ARMS PCR≥ 25 kg/m232IndonesianParticipants with the AT/AA genotypes reported to have 3.72 greater risk of obesity and 5.98 times higher risk for dietary fat intakeAT/AA genotypes associated with obesityOR = 3.72CI = 1.19, 11.64P = 0.009G: FTOL:16q12.2rs1421085Indonesia (56)Case-controlN = 71Obesen = 35Non-obesen = 36ARMS PCR≥ 25 kg/m232 (27.5 ± 36.8)IndonesianCC genotype reported to show higher BMITC/CC genotypes had higher monounsaturated and saturated fatty acid intakes than TT genotypeCC genotype has high BMI in recessive modelCoef = 12.38CI = 5.3, 19.46P ≤ 0.001G: UCP2L: 11q13.4rs659366Indonesia, Yogyakarta (57)Prospective cohort studyN = 323PCR-RFLPNR42·8 ± 9·7IndonesianGG genotype showed more weight gain and influenced energy intake UCP2 does not associated with adiposity in 2 2-year follow-upGG genotype positively correlated with body weightβ = 0·232P = 0·016G: GHRLL: 3p25.3Leu72MetIndonesia (58)Case-control studyN = 198Obesen = 100Non-obesen = 98PCR-RFLP≥ 25 kg/m222.06 ± 4.08JavaneseCA mutant genotype was found in 42.2% of obese groupLeu72Met polymorphism increases the risk of obesityGenotype AA+CA in obeseOR = 4.96CI = 2.43, 10.12P ≤ 0.001G: FTOL:16q12.2rs9939609Indonesia (59)Cross-sectional studyN = 612Normaln = 270Overweightn = 218Obesen = 121ARMS PCR≥ 25 kg/m246.6 ± 14.6BalineseThe minor AA of rs9939609 and CC of rs1421085 increased BMI and were associated with obesityrs1421085 is associated with high BMIOR = 3.22CI = NRP ≤ 0.001rs1421085G: LEPRL: 1p31.3rs1137100Indonesia (60)Cross-sectional studyN = 110Obesen = 55Non-obesen = 55PCR-RFLPNR20.7 ± 3.7Indonesianrs1137100 (K109R) and rs1137101 (Q223R) correlated with obesity and leptin levelLEPR, rs1137100 (K109R) genotype association with body weightOR = NRP = 0.000rs1137101G: RETNL: 19p13.2rs3745368Indonesia (61)Case-controlN = 122Obesen = 61Non-obesen = 61PCR-RFLP≥ 27 kg/m222JavaneseNo association found between genotypes and alleles and obesity parameters and resistin levelsRETN, GG phenotype association with obesityOR = 2.07CI = 0.36, 11.74P = 0.680G: IL-6L: 7p15.3-174 G>CIndonesia-western ethnic (62)Case-controlN = 178Obesen = 89Non-obesen = 89PCR-RFLP≥ 25 kg/m222.1 ± 4.1IndonesianCC genotype had higher plasma CRP and lower IL-6 levels than the GC and GG genotypes in obese groupObese population has more CC genotype than controlCC genotype associated with obese and control subjectsOR = 7.39CI = 2.26, 25.71P = 0.0005G: UCP2L: 11q13.4rs660339Indonesia (63)Case-controlN = 200Obesen = 100Non-obesen = 100PCR-RFLP≥ 25 kg/m222.35 ± 4.81JavaneseI/D genotypes were associated with obesity after gender stratificationFrequency of 45 bp II genotype in obese and control groupsOR = 2.33CI = 0.58, 9.40P = 0.22G: UCP2G(−866)AL: 11q13.4rs659366Indonesia (64)Cross-sectionalN = 603Obesen = 217Non-obesen = 386PCR-RFLP≥ 25 kg/m249 ± 13BalineseNo association between both UCP polymorphism and obesity traitsRural people with A/A genotype of G(−866)A has high BMIOR = NRG: UCP2Ala55Val (C>T)L:11q13.4rs660339G: UCP1L: 4q31.1rs1800592Vietnam, Hai Duong (65)Case-controlN = 140PCR-RFLP≥ 25 kg/m255.6 ± 3.8,Postmenopausal womenVietnameseVDR polymorphisms, Basim and ApaI associated with overweight and obesityVDR and ESR1 interaction reported to have effect on adiposityVDR ApaI association with overweight and obese populationOR = 3.00CI = 1.08, 8.36P = 0.03G: UCP2L:11q13.4rs659366G: ADRA2BL: 2q11.2rs4994G: LEPRL: 1p31.3rs1137101G: ESR1L: 6q25.1 q25.2rs2234693rs9340799G: VDRL: 12q13.11rs2228570rs1544410rs7975232rs731236I/D = insertion/deletion; OR = odds ratio; CI = confidence interval; NR = not reported; PCR = polymerase chain reaction; RFLP = restriction fragment length polymorphism; SNP = single nucleotide polymorphism; WHR = waist-to-hip ratio; HDL-C = high-density lipoprotein cholesterol; DBP = diastolic blood pressure; TG = triglycerides; BMI = body mass index; ARMS = amplification refractory mutation system Summary of study characteristics and main findings of the selected articles I/D = insertion/deletion; OR = odds ratio; CI = confidence interval; NR = not reported; PCR = polymerase chain reaction; RFLP = restriction fragment length polymorphism; SNP = single nucleotide polymorphism; WHR = waist-to-hip ratio; HDL-C = high-density lipoprotein cholesterol; DBP = diastolic blood pressure; TG = triglycerides; BMI = body mass index; ARMS = amplification refractory mutation system Gene, chromosome location, SNPs, and allele Gene, chromosome location, SNPs, and allele Country (population), study design, sample size, and sample characteristics (age and ethnicity) Country (population), study design, sample size, and sample characteristics (age and ethnicity) Genotyping method, definitions of obesity based on different BMI thresholds, and overall findings Genotyping method, definitions of obesity based on different BMI thresholds, and overall findings Odds ratio (95% CI), regression (β) coefficient (95% CI), and P-value Odds ratio (95% CI), regression (β) coefficient (95% CI), and P-value Significant heterogeneity was identified during data extraction, including the study design, type of SNPs, BMI thresholds for obesity, and population types. Hence, a SWiM was carried out per the published guideline (43). As per SWiM protocol, the populations, study design, and outcome were tabulated and illustrated via graphical presentation. Risk of Bias Grading Classification The quality of the selected articles was assessed using the Newcastle-Ottawa Scale (NOS), a tool specifically designed for evaluating non-randomised studies. According to Wells et al. (66), the NOS has four main criteria for assessing the quality of studies screened for inclusion in meta-analyses: the selection of study groups, the comparability of study groups, the methods used and non-response rates of the participants for exposure assessment. The scoring system is as follows: High quality: Studies that score between 7 and 9 stars High quality: Studies that score between 7 and 9 stars Fair quality: Studies that score between 4 and 6 stars Fair quality: Studies that score between 4 and 6 stars Low quality: Studies that score between 1 and 3 stars Low quality: Studies that score between 1 and 3 stars Two investigators were independently assigned to evaluate the quality of the identified studies based on the NOS scale. Results General Characteristics of the Included Studies Twenty-five studies published between 2010 and 2024 were selected through a systematic search. No studies were published in 2010, 2013, 2023, or 2024. The highest number of publications was in 2018, with four articles (16% of the total). Regarding the geographical distribution of research, 14 studies focused on Malaysians (40–42, 44–54), followed by 10 studies on Indonesians (55–64) and one study on Vietnamese (65). The combined population across the 25 studies consisted of 8,312 participants representing five ethnic groups: Malay, Chinese, Indian, Javanese, and Balinese. Malay participants comprised the largest group, comprising 35% of the total, with 2,909 individuals. This was followed by Chinese participants, who comprised approximately 24.4% of the total with 2,035 individuals, and Balinese participants, who represented 14.6% with 1,215 individuals. Notably, about 12% of participants across seven studies did not have their ethnicity disclosed, with these studies primarily conducted in Indonesia and Vietnam. A summary of the distribution of studies by year, country, and ethnicity is illustrated in Figure 2. This review included 25 studies, comprising 12 case-control studies, 12 cross-sectional studies, and one cohort study. Regarding genotyping methods, Polymerase Chain Reaction-Restriction Fragment Length Polymorphism (PCR-RFLP) was the most used, featured in 18 studies, accounting for 72% of the total. Amplification refractory mutation system (ARMS) PCR was used in three studies, quantitative PCR (qPCR) was used in two studies, and Sequenom MassArray® was used in two studies. Figure 3 illustrates the summary of study designs and genotyping methods. The distribution of the 25 selected studies by years, their respective populations, and ethnicities The distribution of genotyping methods and study designs among the 25 studies Risk of Bias Evaluation We evaluated the quality of the selected studies using the NOS. Of the 25 studies assessed, 18 were categorised as high quality, six as fair quality, and one as low quality. Sample size calculations were reported in the minority of studies (n = 9; 36%). One study (64) was rated as low quality, receiving only two stars due to the absence of data on the frequency of obesity among the obese and non-obese groups. The authors accumulated medical history through questionnaires from the participants, which showed ascertainment of exposure. The participants were postmenopausal women, who were not an accurate representation of the obese Vietnamese population. Additionally, six studies (44, 46, 50, 52, 53, 63) recruited participants from university students and staff, limiting the generalisability of the results to the broader population. Students’ unique lifestyles and environments do not reflect those of the general working adult population. Furthermore, due to the pervasive use of the convenience sampling method, several studies oversampled Chinese participants (44, 46, 50, 52, 53), which does not reflect the actual population’s demographic composition. Similarly, other studies focused exclusively on the Malay ethnicity (40, 45, 47, 48, 51), limiting the generalizability of their findings to the overall Malaysian population. Four studies (49, 55, 56, 65) relied on self-reported medical histories, which may introduce recall bias. One study (62) utilised BMI values on a continuous scale as the outcome measure, which is insufficient for predicting the associations between the investigated SNPs and the obesity category. Meanwhile, another study (51) did not report the anthropometric measurements of the participants, further limiting the study’s conclusions. The summary of the risk of bias assessment is presented in Tables 3, 4 and 5. Table 3The Newcastle-Ottawa Scale for the quality assessment of the selected case-control studiesAuthorSelectionComparabilityExposureTotal scoreIs the case definition adequate?Representativeness of the casesSelection of controlsDefinition of controlsComparability of cases and controls on the basis of the design or analysisAscertainment of exposureSame method of ascertainment for cases and controlsNon-response rateApalasamy et al. (45)✵✵✵✵✵✵✵7Wan Rohani et al. (51)✵✵✵✵✵✵✵7Lee et al. (52)✵✵✵✵✵5Say et al. (53)✵✵✵✵✵✵6Chia et al. (54)✵✵✵✵✵✵✵7Al-Jawadi et al. (56)✵✵✵✵✵✵✵8Muhammad et al. (57)✵✵✵✵✵✵✵7Priliani et al. (59)✵✵✵✵✵✵✵✵8Pramudji et al. (62)✵✵✵✵✵✵✵✵8Surniyantoro et al. (63)✵✵✵✵✵✵✵✵8Oktavianthi et al. (64)✵✵✵✵✵✵✵7Binh et al. (65)✵✵2 The Newcastle-Ottawa Scale for the quality assessment of the selected case-control studies Table 4The Newcastle-Ottawa Scale for the quality assessment of the selected cross-sectional studiesAuthorSelectionComparabilityOutcomeTotalRepresentatives of the sampleSample sizeNon-respondentsAscertainment of the exposure (risk factor)Comparability of studies on the basis of the design or analysisAssessment of the outcomeStatistical testFan et al. (41)✵✵✵✵✵✵✵7Mohanraj et al. (42)✵✵✵✵✵✵✵7Campbell et al. (43)✵✵✵✵✵✵✵7Apalasamy et al. (45)✵✵✵✵✵✵6Lek et al. (46)✵✵✵✵✵✵✵7Apidi et al. (48)✵✵✵✵✵✵6Mitra et al. (49)✵✵✵✵✵✵✵7Kok et al. (50)✵✵✵✵✵5Daya et al. (55)✵✵✵✵✵✵✵7Hastuti et al. (60)✵✵✵✵✵✵✵7Utami et al. (61)✵✵✵✵✵✵✵7 The Newcastle-Ottawa Scale for the quality assessment of the selected cross-sectional studies Table 5The Newcastle-Ottawa Scale for the quality assessment of the selected cohort studyAuthorSelectionComparabilityOutcomeTotalRepresentatives of the exposed cohortSelection of the non-exposed cohortAscertainment of exposureDemonstration that outcome of interest was not present at start of studyComparability of the cohort on the basis of the design or analysisAssessment of the outcomeWas follow-up long enough for outcomes to occurAdequacy of follow-up of cohortsMartantiningtyas et al. (58)✵✵✵✵✵✵✵✵8 The Newcastle-Ottawa Scale for the quality assessment of the selected cohort study Due to the profound heterogeneity in the inter-study methodological quality, we did not proceed with meta-analysing the study results. Therefore, we opted for a narrative systematic review to synthesise the results from different studies. Discussion The current review highlights several Southeast Asian studies investigating the association between obesity-related gene polymorphisms and obesity risk. This region offers a unique opportunity to understand how genetic variation, interacting with diverse environmental exposures and ethnic backgrounds, contributes to the complexity of obesity. Among the genes most extensively studied in the selected literature are LEP and LEPR (41, 42, 51, 60). Leptin, encoded by the LEP, plays a key role in regulating appetite, body weight, and fat metabolism (67). It is secreted primarily by white adipose tissue and acts by binding to its receptor, LEPR, in hypothalamic neurons, where it suppresses food intake and promotes energy expenditure (68, 69). Several studies in Malaysia and Indonesia have evaluated the common LEP and LEPR variants, though findings have been inconsistent. For example, in a study of 408 Malaysians, Fan and Say (41) reported no significant association between LEP SNPs rs2167270 and rs7799039, as well as LEPR SNPs rs1137100 and rs1137101 with obesity risk. However, these variants were significantly associated with obesity in the Chinese populations. The lack of significant findings in the Indian subgroup was likely influenced by a smaller sample size, where the statistical power ranged from 57% to 63% (41). Similarly, Wan Rohani et al. (51) found no significant association between the LEP variants rs2167270 and rs7799039 and obesity in a study involving 249 Malays in Malaysia. Nevertheless, the same study reported that the AAG haplotype formed by three LEP SNPs, rs7799039, rs2167270, and rs4731426, was significantly associated with obesity, with an odds ratio (OR) of 8.89 (95% CI: 1.59, 49.78) (51). In contrast, another study involving 185 multi-ethnic Malaysians found no association between BMI and LEP rs7799039 (P = 0.117) and LEPR rs1137101 (P = 0.469), nor between leptin levels and these SNPs (LEP rs7799039: P = 0.196; LEPR Q223A: P = 0.453) (42). Conversely, a cross-sectional study from Indonesia found a positive association between rs1137100 and rs1137101 LEPR with both obesity and leptin levels (60). Specifically, obese individuals with the RR genotype of rs1137100 in the LEPR had leptin levels that were 99% higher, with an average BMI of 36.10, compared to controls with a BMI of 22.04. Additionally, obese individuals with the KK+KR genotype had leptin levels that were 28.3% higher, with an average BMI of 33.86, compared to controls with a BMI of 22.17 (60). Unlike these positive associations observed in Indonesia and other populations, most studies in Malaysia have reported no significant links between LEP variants and obesity. This contradicts the findings from Finnish (70), Caucasians (71, 72), Saudi Arabian (73), and European populations (74), where LEP polymorphisms were associated with obesity traits. These inconsistencies may be attributed to small sample sizes, varying study designs, and the use of non-random purposive sampling, which limits the generalizability of findings (41, 51). Additionally, genetic factors such as differences in linkage disequilibrium patterns, allele frequencies, and population structure across ethnic groups may contribute to the divergent associations observed in obesity-related SNPs (75, 76). Supporting this complexity, a meta-analysis of nine studies involving 2,988 participants concluded that the rs7799039 LEP polymorphism is not generally associated with the development of obesity. However, a significant association was observed in specific populations, including a small South American (n = 788) (77) and in the Tunisian (n = 329) population (78), where rs7799039 was associated with increased obesity risk, suggesting the influence of population-specific genetic factors. Similarly, LEPR polymorphisms have shown ethnic variation. In Caucasian populations, rs1137100 showed a significant association in both BMI and fat mass (P = 0.02 and P = 0.05), while rs1137101 was also significantly linked to BMI (P = 0.005 and P = 0.03) (79). In addition, FTO variants, first identified as being positively associated with obesity in European populations (80), have been studied across four Asian ethnicities: Balinese Indonesians, Malays, Chinese, and Indians in Malaysia (40, 49, 55, 56, 59). A study conducted in Jakarta, although not specifying participants’ ethnicity, found that those with the AT/AA genotypes of rs9939609 FTO had a 3.72-fold higher risk of obesity (55). Similarly, the minor AA genotype of rs9939609 FTO increased BMI by 1.25 kg/m2 (P = 0.012), while the CC genotype of rs1421085 increased by 1.12 kg/ m2 (P = 0.022), particularly in females (59). The minor allele frequency (MAF) of this study was 0.19, consistent with findings from studies conducted among multi-ethnic Malaysians (81) and Minangkabau Indonesians (82). The study involving 275 Minangkabau Indonesians reported that the AA genotype of rs9939609 FTO has significantly higher body fat, weight, waist-hip ratio, and BMI of 27.39 (SD 4.69) compared to TT, which has a BMI of 23.96 (SD 4.51) and TA of about 24.56 (SD 4.8) (82). Additionally, the GG genotype of another rs9930506 FTO was found to be associated with obesity in the Malaysian population under a codominant model (49). The CC genotype of the rs1421085 variant in Indonesians was linked to a higher BMI of 12.58 kg/m2 (P = 0.001) in codominant and 12.38 kg/m2 (P < 0.001) in the recessive model compared with those without this variant, indicating a recessive trait (56). The MAF was 22%, lower than that observed in the Balinese population (59). However, a study involving 587 Malays failed to find an association between 31 FTO SNPs genotypes and allelic frequencies with obesity, except for rs17817288, which was significantly associated with LDL-C (40). Moreover, a GWAS study conducted among Singaporean Chinese also found no association between FTO variants and BMI (83). This highlights the possibility that ethnic-specific gene-environment interactions or other modifying loci may modulate the effect of FTO variants. UCP, primarily located in mitochondria, plays a key role in regulating thermogenesis and energy expenditure while also offering protection against oxidative stress (84). The A/G genotype of rs1800592 UCP was the initial SNP established by Canadian scientists in 1994 to be associated with high BMI in individuals with a family history of obesity. In Southeast Asia, UCP polymorphisms were studied in three different populations, including Malaysia, Indonesia, and Vietnam (52, 53, 57, 63, 65). The combination of the rs1800592 UCP1 AA genotype and the rs1800849 UCP3 CC genotype has been notably associated with a higher waist-hip ratio (WHR) of 0.85 (SD 0.10) and a BMI of 26.17 (SD 5.27) compared to other genotype combinations such as AA/TT, GG/CC, and GG/TT, which are associated with lower BMIs of 22.11 (SD 3.01), 22.05 (SD 3.77), and 22.38 (SD 4.47), respectively, in the Malaysian Chinese population (52). Interestingly, the presence of the T allele of UCP3 in the Chinese population is linked to a reduced risk of obesity, as carriers of the T allele exhibit a 30% lower risk of central obesity and a 2.5% lower WHR compared to those with the C allele (52). This suggests a potential protective effect of the T allele against obesity. Moreover, while UCP1 alone is not significantly associated with WHR, its combination with UCP3 appears to have a synergistic effect, significantly influencing obesity and adiposity. Meanwhile, unlike Malaysians, the G allele of rs1800592 UCP1 was susceptible to weight gain compared to the A allele in the Finnish population (85). A similarly significant association was found for the minor G allele with BMI (OR: 1.52, CI: 1.10, 2.08, P = 0.009) in the Saudi Arabian population (86) and associated to fat mass (P = 0.002) and muscle mass (P = 0.019) in Mexican adults (87). Furthermore, Say et al. (53) reported a significant association between the UCP2 gene polymorphism and adiposity in Malaysian women. In the Chinese population, the MAF of UCP2 was 0.12, consistent with findings in other Chinese cohorts (88). Among Indians, the MAF was higher, at 0.21, similar to data from southern Indians of Tamil Nadu, India (89). This difference may be attributed to the historical migration patterns and ancestral connections between Malaysian Indians and Chinese and their mainland counterparts. In contrast, the TT genotype and T allele of UCP2 carry less risk of obesity compared to the C allele, as the OR CI of TT was 0.4 (95% CI: 0.13, 1.20), while the T allele was 0.55 (95% CI: 0.32, 0.95) in male Indonesian Javanese (63). Besides, an Indonesian Nutrigenetic Cohort reported that individuals with the GG genotype for rs659366 UCP2 are associated with high energy intake, body fat, and weight (57), and the AA genotype has been positively associated with higher BMI in the rural Balinese population (64). Besides, a pairwise gene-gene interaction analysis of vitamin D receptor (VDR ApaI) and UCP2 demonstrated a significant interaction (P = 0.003) and 69% predictive accuracy for overweight and obesity in postmenopausal Vietnamese women (65). The ADR2A was first noted to have a crucial role in regulating insulin secretion and lipolysis (OR = 1.62; 95% CI: 1.06, 2.49; P = 0.026) in Swedish obese individuals without T2DM (90), with a stronger association with obesity in Swedish women with T2DM (OR = 7.61; 95% CI: 1.70, 34.17; P = 0.008) (91). In Malaysians, a gene-gene interaction between rs553668 ADR2A and the angiotensin-converting enzyme rs4646994 ACE polymorphisms was associated with lower central adiposity (44, 53). However, in a separate study in the Malay population, no significant association was found between the ACE variant and obesity parameters (48). Based on the findings from these studies, several limitations and challenges were found that prevent the definitive ascertainment of SNPs associated with obesity and the generalisability of the results to the wider Southeast Asian population. First, there is a significant gap in replicating the results consistently across different populations and ethnic groups due to population-specific genetic factors. Second, the small sample sizes employed in many studies have also limited the study power to detect significant associations between SNPs of smaller effect sizes (e.g. smaller ORs) and obesity. Thus, inter-country collaboration within the Southeast Asian region may be leveraged to increase the number of study participants and improve study power. Besides, Southeast Asia has experienced rapid urbanisation and economic growth in recent decades, which has led to substantial changes in diet and physical activity patterns (36). These environmental shifts could exacerbate the impacts of genetic factors on obesity through gene-environment interaction, particularly among individuals carrying high-risk alleles (92). Furthermore, as reflected in Table 2, only three countries, Malaysia, Indonesia, and Vietnam, have conducted genetic association studies on obesity, highlighting a significant geographical research gap within the Southeast Asian region. This limited representation restricts the ability to generalise findings across the diverse ethnic and genetic backgrounds present in Southeast Asia. Conclusion In conclusion, the reviews highlight the complex and multifaceted nature of obesity, particularly in the context of gene polymorphisms within Southeast Asian populations. The LEP and UCP polymorphisms are among the most studied in Southeast Asian populations. The UCP I/D polymorphism has been significantly associated with overall adiposity in Malaysian women, while in the Indonesian cohort, the GG genotype of UCP2, particularly the rs659366 variant, was linked to weight gain. On the other hand, LEP variants did not show a significant association with obesity in Malaysians. Similarly, the FTO shows population-specific effects. The rs9939609 variant was associated with obesity in Indonesians in two studies, and the rs1421085 variant was linked to a higher BMI. However, FTO variants were not significant determinants of obesity in the Malaysian population. These findings emphasise the potential of genetic polymorphisms to serve as population-specific biomarkers for obesity, underlining the importance of precision-based approaches in prevention and management. Moreover, the findings underscore the importance of considering the genetic diversity and environmental factors contributing to obesity and the challenges in replicating genetic associations across different ethnic groups. Future research should prioritise large-scale, multi-ethnic genomic studies to identify robust and reproducible biomarkers. Such efforts can inform targeted public health strategies and early interventions. Cross-border collaboration, particularly among developing nations, could enhance the exploration of population-specific genetic factors related to obesity. Ultimately, these initiatives could facilitate the development of biomarker-driven risk assessments, screening tools, and personalised interventions tailored to the unique genetic and environmental context of Southeast Asia. Acknowledgements We thank the UKM Medical Molecular Biology Institute (UMBI), the Malaysian Cohort staff members and research assistants for their technical assistance in data collection. We also gratefully acknowledge the voluntary participation of all study participants. Footnotes Ethics of Study: This study was reviewed and approved by the Institutional Review and Ethics Committee of Universiti Kebangsaan Malaysia (JEP-2024-113). All procedures involving human participants were conducted in accordance with the ethical standards of the institutional and/or national research committees and with the Declaration of Helsinki (as revised in 2013). Written informed consent was obtained from all participants prior to study participation. Conflict of Interest: None. Funds: We would like to acknowledge Universiti Kebangsaan Malaysia for their research grant (Dana Impak Perdana 2.0 [DIP-2023-025]. The funder did not participate in the preparation of the manuscript or influence its content or publication decision. Authors’ Contributions: Conception and design: NA, UV, MIAJ, ARAJ Analysis and interpretation of the data: NA, UV, MIAJ Drafting of the article: NA, UV Critical revision of the article for important intellectual content: NA, UV, MIAJ, ARAJ, CSF, NAAM Final approval of the article: NA, UV, MIAJ, ARAJ, CSF, NAAM Provision of study materials or patients: NA, ARAJ Statistical expertise: NA, MIAJ Obtaining of funding: NA References