Rising obesity and shifting disease patterns in Saudi Arabia: a nine-year population-based analysis of chronic disease burden and multimorbidity profiles

Abstract

Background

Obesity represents a major and growing public health challenge in Saudi Arabia and is commonly accompanied by a broad range of chronic conditions. Understanding population-level changes in body mass index (BMI) distribution and associated disease burden over time is important for surveillance and health system planning. This study aimed to describe temporal trends in BMI categories and examine their associations with chronic disease burden among adults in southern Saudi Arabia over a nine-year period.

Methods

We conducted a retrospective cohort study using electronic health records from the Armed Forces Hospitals Southern Region between January 2017 and April 2025. The dataset included 74,881 adult patients with a total of 956,547 visit entries. Body mass index (BMI) was calculated and categorized based on World Health Organization criteria. Chronic conditions such as type 2 diabetes, hypertension, heart failure, asthma, and others were identified from diagnosis records. Multivariable logistic and negative binomial regressions were used to evaluate the association between BMI and disease risk. Latent class analysis was performed to identify distinct multimorbidity profiles.

Results

Nearly half of the study population fell into obesity categories, with 26.6% in class 1, 13.3% in class 2, and 6.5% in class 3. Obesity rates rose consistently over the study period. Higher BMI was strongly associated with increased odds of chronic diseases. Compared to individuals with normal BMI, those in obesity class 3 had 2.47 times the odds of type 2 diabetes and 2.60 times the odds of hypertension. Multimorbidity also rose with BMI, and age was a significant independent predictor. Latent class analysis revealed three distinct disease profiles, with the most burdened group being older and having the highest average BMI.

Conclusion

In this large healthcare-based population, obesity prevalence increased over time and was associated with a graded burden of chronic disease and multimorbidity. Associations varied by age and sex, highlighting heterogeneity in obesity-related disease patterns. Although causal inferences cannot be drawn, these findings provide important population-level evidence to support ongoing surveillance and the development of age- and sex-responsive prevention and clinical management strategies in Saudi Arabia and similar settings.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-026-26923-2.

Background

Obesity, defined by the World Health Organization (WHO) as a body mass index (BMI) of 30 kg/m² or higher, has emerged as one of the most pressing public health challenges in Saudi Arabia in recent decades [1]. The Kingdom has witnessed a dramatic epidemiological transition, with obesity rates soaring from historically low levels to among the highest in the world within just a few generations. Recent national studies estimate that approximately 50–60% of Saudi adults now meet the criteria for obesity, with particularly high prevalence observed among women and urban populations [2]. This rapid increase has been attributed to profound lifestyle changes accompanying the nation’s economic development, including decreased physical activity, increased sedentary behaviors, and the widespread adoption of energy-dense, nutrient-poor diets.

The clinical implications of this obesity epidemic are particularly concerning in the Saudi context. Emerging evidence suggests that the Saudi population may experience obesity-related comorbidities at younger ages and lower BMI thresholds compared to Western populations [3]. For instance, studies indicate that type 2 diabetes mellitus (T2DM) and metabolic syndrome frequently manifest in Saudi patients with BMIs considered only moderately obese (30–35 kg/m²) by international standards [4]. This phenomenon may reflect unique genetic predispositions, distinct patterns of fat distribution (particularly increased visceral adiposity), or other population-specific metabolic characteristics. Furthermore, the prevalence of severe obesity (BMI ≥ 40 kg/m²) has risen alarmingly, with recent estimates suggesting that nearly 10% of Saudi adults now fall into this highest-risk category [5].

The management of obesity in Saudi Arabia faces numerous systemic challenges. While bariatric surgery rates have increased in recent years, access remains limited primarily to major urban centers [6]. The introduction of novel anti-obesity medications, such as GLP-1 receptor agonists, has been slowed by high costs and limited insurance coverage [7]. Primary care systems may lack the infrastructure for comprehensive obesity management, and some healthcare providers remain inadequately trained in evidence-based obesity treatment approaches [8]. These limitations occur against a backdrop of rapidly increasing obesity prevalence among children and adolescents, portending even greater public health challenges in the coming decades [9].

From a health economics perspective, obesity imposes staggering costs on the Saudi healthcare system. Conservative estimates suggest that obesity-related conditions account for US$116.85 billion from a societal perspective and US$109.67 billion from a healthcare system perspective, with additional substantial losses in workforce productivity [10]. These economic impacts are particularly concerning as Saudi Arabia pursues ambitious health sector goals under Vision 2030, which emphasizes preventive care and non-communicable disease management [11]. The growing recognition of obesity as a gateway to multiple chronic diseases has prompted calls for more aggressive national strategies to address this epidemic.

This study aims to examine population-level temporal trends in body mass index (BMI) categories and the associated burden of chronic diseases among adults in southern Saudi Arabia using electronic health records collected over a nine-year period. Specifically, we assess changes in the distribution of BMI categories over time and evaluate cross-sectional associations between BMI and a range of chronic conditions, with attention to variation by age and sex. We further characterize patterns of multimorbidity using latent class analysis to identify subgroups with distinct disease profiles. By providing a detailed description of obesity-related disease patterns within a large healthcare system, this study seeks to inform surveillance efforts and support the development of targeted prevention and clinical strategies.

Methods

Study design and setting

This study employed a retrospective, population-based analysis of electronic health records (EHRs) from the Armed Forces Hospitals Southern Region (AFHSR), a tertiary healthcare system serving military personnel and their dependents in southern Saudi Arabia. AFHSR delivers comprehensive medical services across all specialties and functions as a referral center for patients from Sharurah, Najran, Jizan, and surrounding areas. The analysis covered the period from January 2017 to April 2025 and reflects a dynamic healthcare population, with individuals entering and exiting care over time, rather than a fixed longitudinal cohort. The study was designed to describe population-level distributions and temporal trends in body mass index categories and associated chronic disease burden within this healthcare setting.

Study population and data source

The dataset included 74,881 unique adult patients aged 18 years and above. Due to longitudinal follow-up visits, the dataset consisted of 956,547 individual visit entries. The study reflects a dynamic healthcare population, with individuals entering and exiting care over time, rather than a fixed longitudinal cohort. Each record included the following variables: medical record number (MRN), gender, age, height, weight, BMI, diagnosis (disease name), visit date, and visit number. Only patients with complete data for height, weight, and diagnosis were included in the analysis. The data were extracted and anonymized before analysis to protect patient confidentiality.

Variables and definitions

BMI was calculated using the standard formula: weight in kilograms divided by the square of height in meters (kg/m²). Patients were categorized based on World Health Organization (WHO) BMI classifications: normal weight (18.5–24.9), overweight (25.0–29.9), and obesity classes 1 (30.0–34.9), 2 (35.0–39.9), and 3 (≥ 40.0) [12]. Chronic conditions were identified based on the “disease name” field in the EHR, with a focus on type 2 diabetes mellitus, hypertension, heart failure, asthma, stroke, back pain, infertility, depression, and anxiety disorders.

Chronic disease identification

In this study, chronic conditions were identified from the electronic health records using the diagnostic labels recorded in the “disease name” field. These diagnoses were mapped to standardized categories based on clinical judgment and consistent terminology. The conditions assessed included type 2 diabetes mellitus (T2DM), hypertension (HTN), dyslipidemia or hyperlipidemia (DLD/HLD), heart failure (HF), atrial fibrillation (AF), chronic kidney disease (CKD), pulmonary embolism (PE), deep vein thrombosis (DVT), gout, metabolic dysfunction-associated steatotic liver disease (MASLD), biliary calculus, obstructive sleep apnea (OSA), asthma, gastroesophageal reflux disease (GERD), and osteoarthritis (OA).

Cases of MASLD were identified through documented diagnoses of nonalcoholic fatty liver disease in the EHR, reflecting terminology used during the data collection period. Atherosclerotic cardiovascular disease (ASCVD) was defined as a composite outcome including either myocardial infarction or cerebrovascular accident (stroke).

Statistical analysis

All data analyses were conducted using R software (version 4.4.3). Data cleaning and preprocessing were performed using the dplyr, tidyr, and data.table packages. Records missing essential variables such as height, weight, or diagnosis were excluded. Descriptive statistics were used to summarize demographic and clinical characteristics across BMI categories. Comparisons between categorical variables were conducted using chi-square tests (stats package), and differences in continuous variables were analyzed using one-way ANOVA. To investigate the association between BMI categories and chronic diseases, multivariable logistic regression models were constructed using the glm() function from the stats package with a binomial family. Models included age and gender as covariates, and adjusted odds ratios (ORs) with 95% confidence intervals (CIs) were reported. To analyze the relationship between BMI category and the number of chronic conditions per individual, negative binomial regression was performed using the glm.nb() function from the MASS package. This model accounted for overdispersion in count data. Incidence rate ratios (IRRs) were calculated, and models were adjusted for age and gender. Age was treated as a continuous variable, and gender as a binary factor.

Temporal trends in BMI categories and chronic disease prevalence from 2017 to 2025 were evaluated by calculating annual distributions. Data visualization was performed using the ggplot2 and plotly packages. Line plots were generated to illustrate yearly changes in the proportions of individuals by BMI classification and in the prevalence of key chronic conditions: type 2 diabetes mellitus, hypertension, heart failure, back pain, and asthma. Latent class analysis (LCA) was conducted to identify underlying patterns of multimorbidity. The poLCA package was used to fit LCA models based on binary indicators for the presence or absence of chronic conditions. Models with two to six classes were estimated, and the Bayesian Information Criterion (BIC) was used to determine the optimal number of classes. The solution with the lowest BIC and meaningful clinical profiles was selected. Each latent class was then profiled in terms of average age, BMI, and disease burden.

To assess the consistency of associations between body mass index (BMI) categories and chronic disease outcomes across key demographic subgroups, additional stratified and interaction analyses were performed. Multivariable logistic regression models were stratified by age group (< 50 vs. ≥ 50 years) and by sex to estimate adjusted odds ratios within each subgroup (Supplementary File 1). Models were adjusted for age (in sex-stratified analyses) or sex (in age-stratified analyses), as appropriate. Formal effect modification by age group and sex was evaluated by including interaction terms between BMI category and age group, and between BMI category and sex, with statistical significance assessed using likelihood ratio tests comparing models with and without interaction terms (Supplementary File 1). Model assumptions were assessed, including checks for multicollinearity and overdispersion, and no major violations were identified.

Results

Demographic and clinical characteristics

Table 1 presents the baseline demographic and clinical characteristics of the study population, stratified by obesity class. Among the 74,881 individuals included in the analysis, 20.9% were classified as having normal weight, 32.7% as overweight, and 46.4% were classified within obesity categories, including 1st degree (26.6%), 2nd degree (13.3%), and 3rd degree (6.5%). A clear pattern emerged in the distribution of gender across BMI categories (p < 0.001). Males predominated in the normal and overweight groups (61% and 59%, respectively), while females became increasingly overrepresented in the higher obesity classes, comprising 55% in Obese 1st degree, 68% in Obese 2nd degree, and 73% in Obese 3rd degree. The average age also varied significantly by BMI category (p < 0.001), increasing with higher BMI. Individuals in the normal weight group had a mean age of 46.2 years, while those in the obesity categories had consistently higher mean ages of approximately 53 years.

Table 1.

Baseline demographic and clinical characteristics by obesity class

Characteristic	Overall N = 74,881	Normal N = 15,664	Overweight N = 24,473	Obese 1st degree N = 19,892	Obese 2nd degree N = 9,966	Obese 3rd degree N = 4,886	p-value*
Gender							< 0.001*
Female	37,374 (50%)	6,177 (39%)	10,029 (41%)	10,865 (55%)	6,729 (68%)	3,574 (73%)
Male	37,507 (50%)	9,487 (61%)	14,444 (59%)	9,027 (45%)	3,237 (32%)	1,312 (27%)
Age	51.1 ± 17.5	46.2 ± 20.7	51.5 ± 17.2	53.1 ± 15.8	53.1 ± 15.4	53.1 ± 15.4	< 0.001*
Type 2 diabetes mellitus	34,992 (47%)	5,374 (34%)	11,116 (45%)	10,296 (52%)	5,453 (55%)	2,753 (56%)	< 0.001*
Hypertension	24,335 (32%)	3,529 (23%)	7,553 (31%)	7,211 (36%)	3,939 (40%)	2,103 (43%)	< 0.001*
Back pain	23,294 (31%)	4,984 (32%)	7,957 (33%)	6,080 (31%)	2,983 (30%)	1,290 (26%)	< 0.001*
Heart failure	12,360 (17%)	1,836 (12%)	3,857 (16%)	3,627 (18%)	2,006 (20%)	1,034 (21%)	< 0.001*
Asthma	8,750 (12%)	2,102 (13%)	2,584 (11%)	2,150 (11%)	1,200 (12%)	714 (15%)	< 0.001*
Infertility	6,704 (9.0%)	1,552 (9.9%)	2,229 (9.1%)	1,726 (8.7%)	862 (8.6%)	335 (6.9%)	< 0.001*
CKD	5,653 (7.5%)	1,501 (9.6%)	1,825 (7.5%)	1,323 (6.7%)	631 (6.3%)	373 (7.6%)	< 0.001*
Anxiety disorder	5,521 (7.4%)	1,376 (8.8%)	1,844 (7.5%)	1,360 (6.8%)	635 (6.4%)	306 (6.3%)	< 0.001*
Gout	2,074 (2.8%)	277 (1.8%)	712 (2.9%)	657 (3.3%)	284 (2.8%)	144 (2.9%)	< 0.001*
Stroke	1,712 (2.3%)	474 (3.0%)	583 (2.4%)	383 (1.9%)	187 (1.9%)	85 (1.7%)	< 0.001*
Depression	79 (0.1%)	13 (< 0.1%)	23 (< 0.1%)	23 (0.1%)	9 (< 0.1%)	11 (0.2%)	0.088
Others	960 (1.3%)	203 (1.3%)	327 (1.3%)	206 (1.0%)	125 (1.3%)	99 (2.0%)	< 0.001*

* Significant values when p-value < 0.05. The overall column represents the full study population using the most recent BMI recorded during the study period. Values are presented as mean ± standard deviation or n/N (%). BMI categories are based on the most recent BMI recorded for each individual during the study period

The burden of chronic diseases increased progressively across obesity classes. The prevalence of type 2 diabetes mellitus increased from 34% in the normal weight group to 56% in the Obese 3rd degree group (p < 0.001). Similar gradients were observed for hypertension (23% to 43%) and heart failure (12% to 21%). Back pain was more evenly distributed across BMI classes, with a slight decline in the highest obesity group. Some conditions demonstrated non-linear associations with BMI. For instance, asthma prevalence was relatively stable across most BMI categories but slightly higher in the Obese 3rd degree group (15%). Infertility and anxiety disorders were more prevalent among individuals with normal weight or overweight status and declined with increasing BMI. Stroke prevalence showed a modest decrease across BMI categories (from 3.0% in normal weight to 1.7% in Obese 3rd degree), and depression was uncommon across all groups, although slightly more frequent in individuals with Obese 3rd degree.

Temporal trends in obesity categories and chronic disease prevalence

Figure 1 presents annual prevalence proportions of body mass index (BMI) categories among adults receiving care between 2017 and 2024. Across the study period, the proportion of individuals classified within higher BMI categories increased overall, although with notable fluctuations over time. All obesity classes and overweight status demonstrated a pronounced peak in 2018, followed by a decline in 2019–2020 and a gradual increase thereafter. By 2024, the prevalence of overweight and obesity categories had increased relative to earlier years, indicating a shift in the population-level BMI distribution within this dynamic healthcare population. These estimates reflect annual prevalence proportions rather than incidence, calculated within each calendar year.

Fig. 1.

Temporal trends in obesity classification in our data from 2017 to 2024

Values represent annual prevalence proportions (%), calculated as the number of individuals classified within each BMI category divided by the total number of individuals observed in that calendar year. Estimates reflect a dynamic healthcare population and represent prevalence rather than incidence

Figure 2 presents population-level temporal trends for the five most frequently recorded chronic conditions: type 2 diabetes mellitus, hypertension, back pain, heart failure, and asthma. A sharp increase in recorded disease counts was observed in 2018, likely reflecting changes in healthcare utilization or electronic health record documentation practices rather than true changes in disease occurrence. Type 2 diabetes mellitus and hypertension were the most prevalent conditions early in the study period and showed a gradual decline after 2019. Back pain demonstrated a consistent upward trend and became the most frequently recorded condition by 2023 and 2024. Heart failure showed a gradual decline over time, while asthma prevalence increased modestly in later years, indicating heterogeneous temporal patterns across conditions.

Fig. 2.

Temporal Trends in the prevalence of the top five chronic diseases from 2017 to 2024

Values represent absolute numbers of recorded diagnoses per calendar year within a dynamic healthcare population and do not represent incidence rates

To assess the robustness of observed temporal trends and account for potential artifacts related to electronic health record implementation, a sensitivity analysis excluding the early years (2017–2018) was conducted. The temporal patterns of BMI categories from 2019 to 2024 remained consistent with the main analysis, with sustained increases across obesity classes and stable relative distributions over time (Supplementary File 2). Excluding early years did not materially change the direction or overall magnitude of the observed trends.

Association of BMI categories with chronic disease risk

Adjusted odds ratios indicated that higher BMI categories were associated with greater prevalence of multiple chronic conditions (Table 2). Compared with individuals with normal BMI, those in the Obese 3rd degree category had higher odds of type 2 diabetes mellitus (adjusted OR 2.47, 95% CI: 2.31–2.64) and hypertension (adjusted OR 2.60, 95% CI: 2.43–2.78). The odds of heart failure also increased across BMI categories, with an adjusted OR of 2.02 (95% CI: 1.86–2.20) in the Obese 3rd degree group. In contrast, the odds of back pain decreased in higher obesity categories (adjusted OR 0.77, 95% CI: 0.72–0.83 in Obese 3rd degree), suggesting a non-linear or age-modified association. Similar inverse patterns were observed for anxiety disorders and stroke. Depression showed higher odds only in the Obese 3rd degree category (adjusted OR 2.72, 95% CI: 1.19–6.08), although estimates were imprecise due to low prevalence.

Table 2.

Adjusted odds ratios (95% CI) for chronic diseases by BMI category

Outcome	Normal	Overweight	Obese 1st degree	Obese 2nd degree	Obese 3rd degree
Type 2 diabetes mellitus	Ref	1.59 (1.53–1.66)	2.05 (1.97–2.15)	2.31 (2.20–2.44)	2.47 (2.31–2.64)
Hypertension	Ref	1.53 (1.47–1.61)	1.96 (1.87–2.05)	2.25 (2.13–2.37)	2.60 (2.43–2.78)
Heart failure	Ref	1.41 (1.33–1.50)	1.68 (1.58–1.78)	1.90 (1.77–2.03)	2.02 (1.86–2.20)
Gout	Ref	1.66 (1.45–1.92)	1.90 (1.65–2.19)	1.63 (1.38–1.93)	1.69 (1.37–2.06)
Depression	Ref	1.13 (0.58–2.30)	1.39 (0.72–2.83)	1.09 (0.45–2.52)	2.72 (1.19–6.08)
Back pain	Ref	1.03 (0.99–1.08)	0.94 (0.90–0.99)	0.92 (0.87–0.97)	0.77 (0.72–0.83)
Anxiety disorder	Ref	0.85 (0.79–0.91)	0.76 (0.70–0.82)	0.71 (0.64–0.78)	0.69 (0.61–0.79)
Asthma	Ref	0.76 (0.72–0.81)	0.78 (0.73–0.83)	0.88 (0.82–0.95)	1.10 (1.01–1.21)
CKD	Ref	0.76 (0.71–0.82)	0.67 (0.62–0.73)	0.64 (0.58–0.70)	0.78 (0.69–0.88)
Infertility	Ref	0.91 (0.85–0.98)	0.86 (0.80–0.93)	0.86 (0.79–0.94)	0.67 (0.59–0.76)
Stroke	Ref	0.78 (0.69–0.88)	0.63 (0.55–0.72)	0.61 (0.52–0.73)	0.57 (0.45–0.71)
Others	Ref	1.03 (0.87–1.23)	0.80 (0.66–0.97)	0.97 (0.77–1.21)	1.58 (1.23–2.00)

Odds ratios were estimated using separate multivariable logistic regression models for each outcome. All models were adjusted for age (continuous) and gender. Normal BMI (18.5–24.9 kg/m²) was used as the reference category. Values are presented as adjusted odds ratios with 95% confidence intervals

Multimorbidity and body mass index

To evaluate the association between BMI and overall disease burden, negative binomial regression models were used to examine the relationship between BMI category, age, sex, and the count of chronic conditions per individual (Table 3). A clear graded association was observed between increasing BMI and a higher burden of multimorbidity. Compared with individuals with normal BMI, those classified as overweight had a 6% higher rate of comorbid conditions (IRR = 1.06, 95% CI: 1.04–1.07), while individuals in Obese 1st, 2nd, and 3rd degree categories had progressively higher rates of multimorbidity of 11% (IRR = 1.11), 17% (IRR = 1.17), and 21% (IRR = 1.21), respectively (all p < 0.001). Age was a strong independent predictor of multimorbidity, with each additional year of age associated with a 1% increase in the number of chronic conditions (IRR = 1.01, 95% CI: 1.01–1.01, p < 0.001). Sex also showed a modest association, with females having a slightly lower multimorbidity burden compared with males (IRR = 0.94, 95% CI: 0.93–0.95, p < 0.001).

Table 3.

Association between BMI, Age, and Gender with multimorbidity count using negative binomial regression

Characteristic	IRR	95% CI	p-value
BMI category
Normal (ref group)	—	—
Overweight	1.06	1.04–1.07	< 0.001*
Obese 1st degree	1.11	1.09–1.13	< 0.001*
Obese 2nd degree	1.17	1.15–1.19	< 0.001*
Obese 3rd degree	1.21	1.18–1.24	< 0.001*
Age (continuous)	1.01	1.01–1.01	< 0.001*
Gender
Male (ref group)	—	—
Female	0.94	0.93–0.95	< 0.001*

* Significant values when p-value < 0.05. Multimorbidity was defined as the total count of chronic conditions recorded per individual during the study period. Incidence rate ratios (IRRs) were estimated using negative binomial regression to account for overdispersion. Models were adjusted for age (continuous) and gender. Normal BMI and male gender were used as reference categories

Stratified and interaction analyses of body mass index associations with cardiometabolic outcomes

In age-stratified analyses, the associations between BMI categories and cardiometabolic outcomes were generally stronger among individuals younger than 50 years compared with those aged 50 years and older (Supplementary file 1). For example, higher BMI categories were associated with substantially increased odds of type 2 diabetes and hypertension among individuals aged < 50 years, whereas the magnitude of associations was attenuated but remained statistically significant in the ≥ 50-year age group. These findings suggest earlier and more pronounced cardiometabolic risk associated with obesity at younger ages.

Sex-stratified analyses demonstrated broadly consistent associations between increasing BMI and cardiometabolic outcomes in both males and females (Supplementary file 1). However, the magnitude of associations differed by sex for selected conditions. In particular, obesity was more strongly associated with hypertension and gout among males, whereas associations with type 2 diabetes were more pronounced among females across most BMI categories. Despite these differences in effect size, the overall direction of associations was consistent across sexes.

Latent class analysis of multimorbidity patterns

To explore underlying patterns of disease burden among individuals with varying body mass index (BMI) levels, a latent class analysis (LCA) was conducted (Table 4). This analysis revealed three distinct subgroups (or “latent classes”) within the study population, each characterized by unique demographic and clinical profiles. Class 1, comprising 24,704 individuals (33% of the cohort), represented the oldest and most clinically burdened group. Individuals in this class had a mean age of 62 years and the highest average BMI (31.3 kg/m²). This class likely reflects individuals with advanced age and a greater accumulation of multiple chronic conditions, consistent with a phenotype of high multimorbidity. Class 2 included 19,086 individuals (25.5%) and displayed an intermediate profile. The mean age in this group was 55 years, with a mean BMI of 30.6 kg/m². While not as burdened as Class 1, this subgroup may represent individuals in midlife with early or established chronic disease, particularly cardiometabolic conditions such as type 2 diabetes and hypertension. Class 3, the largest group (N = 31,091; 41.5%), was notably younger and leaner than the other two. Individuals in this class had a mean age of 40 years and the lowest average BMI (29.0 kg/m²). While this group appeared to have the lowest burden of overt chronic diseases, their age and BMI profiles suggest they may be at risk for future disease development, potentially representing an early-stage or preclinical group.

Table 4.

Baseline characteristics of participants by latent class

Characteristic	Class 1 (N = 24,704)	Class 2 (N = 19,086)	Class 3 (N = 31,091)	p-value
Age (years)	62 ± 15	55 ± 15	40 ± 14	< 0.001*
BMI (kg/m²)	31.3 ± 6.2	30.6 ± 6.0	29.0 ± 5.9	< 0.001*
Gender				< 0.001*
Male	12,506 (51%)	9,776 (51%)	15,221 (49%)
Female	12,198 (49%)	9,310 (49%)	15,870 (51%)

* Significant values when p-value < 0.05. Values are presented as mean ± standard deviation or n/N (%). Latent classes were derived using latent class analysis based on binary indicators of chronic conditions. p-values were obtained using Kruskal–Wallis rank-sum tests for continuous variables and Pearson’s chi-squared tests for categorical variables

Effect modification of body mass index associations by age group and sex

Formal interaction testing provided evidence of effect modification by both age group and sex for key cardiometabolic outcomes (Supplementary file 1). Significant interactions were observed between BMI category and age group, as well as between BMI category and sex, for type 2 diabetes mellitus, hypertension, and heart failure (all global p for interaction < 0.001), indicating that the strength of associations varied across demographic subgroups.

Discussion

This large-scale retrospective analysis of electronic health records, spanning nine years and nearly one million clinical encounters, provides a detailed description of temporal patterns in body mass index (BMI) categories and their associations with chronic disease burden in southern Saudi Arabia. The findings demonstrate a sustained increase in the proportion of individuals classified within higher BMI categories over time and reveal consistent, graded associations between higher BMI and greater prevalence of multimorbidity, particularly for cardiometabolic and cardiovascular conditions. Together, these results contribute to a clearer understanding of the evolving population-level burden of obesity within this healthcare setting.

The observed increase in the proportion of individuals classified within higher BMI categories between 2017 and 2025 reflects a substantial shift in population-level BMI distribution within this healthcare setting. By 2024, each obesity class accounted for approximately 16% of the study population. These patterns are broadly consistent with national estimates from Saudi Arabia, where adult obesity prevalence has been reported to range between 20% and 39% over the past decade [13]. Prior syntheses have similarly documented increasing obesity prevalence across the Kingdom alongside a high burden of associated chronic conditions, including hypertension, type 2 diabetes mellitus, and dyslipidemia [13]. Although the present study was not designed to assess determinants of obesity, these trends likely reflect the combined influence of lifestyle, environmental, and sociocultural factors that have been described in previous research, particularly among women and older adults [14, 15].

Across BMI categories, higher BMI was consistently associated with greater prevalence of major cardiometabolic conditions, including type 2 diabetes mellitus, hypertension, and heart failure. Individuals classified within the highest obesity category had more than twice the odds of these conditions compared with those with normal BMI, even after adjustment for age and sex. These findings align with existing evidence from Saudi Arabia and other settings demonstrating strong associations between excess adiposity and cardiometabolic disease burden [13]. This reinforces the biological mechanisms linking excess adiposity to chronic disease, including insulin resistance, vascular dysfunction, and systemic inflammation [16, 17]. Notably, stratified and interaction analyses indicated that the strength of these associations varied by age and sex, with relatively stronger associations observed among younger individuals, suggesting that obesity-related cardiometabolic burden may manifest earlier in the life course [18].

Not all conditions exhibited monotonic increases across BMI categories. For example, back pain showed a relative decline in prevalence among individuals in the highest obesity category. This pattern may be partially explained by age-related confounding, as individuals in higher obesity classes were generally older and may have other competing health priorities. However, it is also possible that diagnostic overshadowing plays a role, where musculoskeletal complaints in individuals with severe obesity are underrecognized or attributed to their weight without formal diagnosis. Despite this observed decline, the established literature supports a strong, positive association between obesity and back pain [19]. Excess body weight increases mechanical stress on the spine, particularly in the lower back, and accelerates disc degeneration, which can lead to chronic pain, stiffness, and nerve compression [19, 20].

Similarly, the lower observed prevalence of stroke in higher BMI categories should be interpreted cautiously. This finding may reflect survivor bias, differential healthcare utilization, or misclassification rather than a protective effect of obesity. Previous studies have described an “obesity paradox” in certain cardiovascular outcomes [21]; however, more recent evidence suggests a complex, non-linear relationship between BMI and stroke risk [21, 22]. The cross-sectional nature of the present analysis precludes further exploration of these patterns, underscoring the need for prospective investigation.

The latent class analysis identified three distinct multimorbidity profiles that differed in age, BMI, and disease burden. The most clinically burdened class was characterized by older age and higher BMI, whereas a younger, less burdened group was also identified. These findings highlight the heterogeneity of obesity-related disease burden and suggest that individuals with similar BMI values may experience markedly different health profiles. Such heterogeneity supports the value of risk stratification approaches in public health planning and clinical care rather than uniform intervention strategies [23].

The temporal trends in disease prevalence revealed dynamic shifts over the study period. While type 2 diabetes and hypertension showed initial increases followed by a modest decline post-2019, back pain emerged as a leading complaint in recent years. This trend may reflect increased healthcare-seeking behavior for musculoskeletal disorders or greater public awareness [24]. The modest rise in asthma prevalence and the steady decline in heart failure also suggest evolving patterns in disease recognition, healthcare access, or population aging. These fluctuations highlight the importance of continuous surveillance and adaptive healthcare planning [25].

Our stratified and interaction analyses provide additional insight into how the health burden associated with obesity varies across demographic subgroups. We observed stronger associations between higher BMI categories and cardiometabolic outcomes among individuals younger than 50 years compared with older adults, suggesting that obesity may confer substantial cardiometabolic risk earlier in the life course. This finding aligns with growing evidence that excess adiposity in younger adulthood accelerates the development of metabolic dysfunction and may lead to longer cumulative exposure to cardiometabolic risk factors over time [26–28].

Sex-specific analyses further indicated differences in the magnitude of associations between BMI and selected conditions, with stronger associations observed among males for hypertension and gout and among females for type 2 diabetes [29]. Formal interaction testing confirmed statistically significant effect modification by both age group and sex for key cardiometabolic outcomes, reinforcing the importance of considering demographic context when evaluating obesity-related health risks [30]. Collectively, these findings highlight the need for age- and sex-responsive prevention strategies and support targeted early interventions to mitigate the long-term burden of obesity-related chronic disease.

Our results have several important implications. First, the increasing prevalence of obesity across all classes calls for urgent public health initiatives targeting both primary prevention and early intervention. Second, the strong link between BMI and multimorbidity suggests that routine BMI screening should be integrated with risk prediction tools in clinical practice. Third, the identification of multimorbidity patterns through LCA offers an innovative framework for patient stratification and personalized care pathways, particularly in resource-constrained settings.

Limitations

This study has several limitations that should be considered when interpreting the findings. First, this was a retrospective analysis of electronic health records from a single tertiary healthcare system serving a predominantly military-affiliated population. As such, the results may not be fully generalizable to the broader Saudi population or to individuals receiving care outside similar institutional settings. Second, the use of routinely collected clinical data relies on the accuracy and completeness of diagnostic coding. Variability in documentation practices across clinicians and over time may have resulted in misclassification or under-ascertainment of certain conditions, particularly for diagnoses that are less consistently recorded [31].

Third, although models were adjusted for age and sex, other important confounding factors, including physical activity, dietary patterns, smoking status, medication use, and socioeconomic characteristics, were not available in the dataset. The absence of these variables may have resulted in residual confounding, and the direction and magnitude of this bias could not be formally assessed. Fourth, the temporal relationship between obesity and comorbid conditions could not be definitively established, as BMI and diagnoses were assessed within an observational healthcare context rather than through prospective follow-up. Accordingly, the reported associations should be interpreted as descriptive rather than causal.

Additionally, changes in healthcare utilization, electronic health record implementation, and coding practices during the early years of data collection may have influenced observed temporal patterns [32]. To mitigate this concern, sensitivity analyses excluding early years were conducted and demonstrated that overall trends were robust to these exclusions. Finally, although age- and sex-stratified analyses and formal interaction testing were performed, more granular subgroup analyses (e.g., by socioeconomic status or lifestyle factors) were not possible due to data limitations.

Conclusion

This large-scale retrospective analysis demonstrates a sustained increase in obesity prevalence across all BMI categories among adults in southern Saudi Arabia over a nine-year period, alongside a substantial and graded burden of chronic disease. Higher BMI categories were consistently associated with greater prevalence of cardiometabolic conditions, including type 2 diabetes, hypertension, heart failure, and multimorbidity, at the population level. These associations varied by age and sex, with stratified and interaction analyses indicating meaningful heterogeneity in the strength of relationships across demographic subgroups. The identification of distinct multimorbidity profiles further illustrates the complex and heterogeneous patterns of obesity-related disease burden within this population. Although causal inferences cannot be drawn from this observational study, the findings provide important evidence for ongoing surveillance and highlight priority groups for prevention and clinical attention. Collectively, these results support the need for age- and sex-responsive public health strategies, early preventive efforts, and risk-based clinical management approaches to mitigate the long-term health and healthcare system impacts of obesity in Saudi Arabia and comparable settings.

Authors’ contributions

JA, AMA, AMA, MQAA, MMMA, ASA, FAA, MB, MT, NA, BA, and NZA jointly conceptualized the study, contributed to the study design, and developed the methodology. NZA, MB, and MT made major contributions to data analysis, coding, and data visualization. All authors participated in data acquisition, data verification, and interpretation of the findings. JA and NZA performed the primary statistical analysis with input from all authors. The initial manuscript draft was prepared by JA and NZA , and refined through substantial contributions from all authors, who reviewed, edited, and approved the final version. All authors had full access to the study data and took final responsibility for the decision to submit the manuscript for publication.

Funding

None.

Data availability

The datasets generated and/or analyzed during the current study are not publicly available due to institutional regulations but are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The study protocol was reviewed and approved by the Research Ethics Committee of the Armed Forces Hospitals Southern Region (AFHSR), Khamis Mushait, Saudi Arabia (Registration No. H-06-KM-001; Approval Code: AFHSRMREC/2025/ACADEMIC AFFAIRS/811) on 20 July 2025. The approval covered the retrospective secondary analysis of routinely collected electronic health records generated at AFHSR between January 2017 and April 2025 and remains valid for five years. No data were prospectively collected for research purposes prior to ethics approval. The Research Ethics Committee waived the requirement for individual informed consent because the study involved retrospective analysis of anonymized clinical data, posed minimal risk to participants, and did not involve any direct patient contact. All data were extracted from the hospital electronic health record system and fully anonymized prior to analysis, with no access to personal identifiers. The study was conducted in accordance with the principles of the Declaration of Helsinki and complied with applicable institutional policies and national regulations governing the secondary use of health data in Saudi Arabia, including relevant guidance from the Saudi Ministry of Health on informed consent and data protection.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.