Nationwide trends in pediatric obesity in Thailand, 2015–2023: prevalence, morbidity, mortality, and COVID-19–related disparities

Abstract

Background

Childhood obesity has increased markedly in Thailand, yet nationwide data on clinical outcomes, and the impact of the COVID-19 pandemic remain limited, particularly form Southeast Asia. We examined nationwide trends in hospital-based obesity prevalence, associated comorbidities, and mortality among hospitalized Thai children from 2015 to 2023, with specific focus on pandemic-related changes.

Methods

We conducted a retrospective analysis of all pediatric hospital admissions using National Health Security Office data. The study period covered from 2015 to 2023, including hospitalized children aged 1 month to < 18 years. The study period was stratified into pre-pandemic, pandemic, and post-pandemic phases. Obesity was defined using ICD-10-TM diagnosis codes. We used logistic regression to quantify associations between obesity and comorbidities and Cox proportional hazards models to examine mortality risk factors. Standard errors were adjusted for clustering at the hospital level.

Results

Among 14,483,566 hospitalized children, 42,168 included an obesity diagnosis, yielding an overall prevalence of 29.1 (95% CI: 28.8–29.4) per 10,000 hospitalized children. Obesity prevalence more than doubled during the pandemic, rising from 20.1 per 10,000 pre-pandemic to 45.3 per 10,000 during and remaining elevated at 45.5 per 10,000 post-pandemic. Children with obesity exhibited significantly increased odds of serious comorbidities, with the highest associations observed for nonalcoholic fatty liver disease (NAFLD) (adjusted OR 223.67, 95% CI 121.22–412.71) and orthopedic conditions (adjusted OR 117.59; 95% CI 66.82–206.93). The pandemic period was associated with divergent trends in comorbidity patterns: metabolic complications increased, with type 2 diabetes rising 28% (combined OR 1.28, 95% CI 1.09–1.52) and NAFLD increasing 64% post-pandemic (OR 1.64, 95% CI 1.54–1.75), while conditions requiring routine screening showed apparent decreases during pandemic, including documented obstructive sleep apnea (OR 0.55, 95% CI 0.42–0.71) and metabolic syndrome (post-pandemic OR 0.37, 95% CI 0.31–0.45). In-hospital mortality among children with obesity was 4.7 per 1,000 children. Cardiovascular comorbidities conferred the highest mortality risk (adjusted HR 5.86, 95% CI 3.57–9.61).

Conclusion

Childhood obesity among hospitalized children in Thailand increased sharply during the COVID-19 pandemic and remained elevated through 2023. Obesity was strongly associated with metabolic and orthopedic complications, and indicators of glycemic dysregulation worsen during the pandemic. These findings support strengthening pediatric obesity prevention and implementing systematic screening programs and risk stratification to inform service planning and clinical guidance in Thailand and potentially across Southeast Asia.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-026-26570-7.

Introduction

Childhood obesity has increased sharply globally, with the prevalence among children and adolescents tripling since 1990 [1]. Southeast Asia faces some of the highest projected increases, in Thailand, the prevalence of overweight and obesity among children aged 6–14 years has doubled over the past 25 years, rising from 6% to 13%, and approximately 14% of adolescents aged 15–18 years are now affected. If current trends persist, the World Obesity Federation estimates that more than 60% of Thai children will be overweight or obese by 2035 [2].

COVID-19 lockdowns and school closures further disrupted children’s daily routines, with experts warning that these measures would accelerate weight gain. Global evidence confirms that pandemic restrictions were associated with reduced physical activity, increased sedentary behavior, and poorer dietary patterns among youth [3, 4]. Consequently, studies have documented significant increases in children’s body mass index (BMI) and the worsening of obesity-related comorbidities, particularly among children from lower socioeconomic backgrounds [5–7].

Despite these trends, data on obesity among hospitalized children remain limited, especially in low- and middle-income countries. In Thailand, nationwide evidence on temporal trends in obesity prevalence, complications, and inpatient outcomes is lacking. This gap is important because hospital-based data capture more severe cases and complications that may be missed in population surveys, and the COVID-19 pandemic may have altered detection, coding, and severity patterns in pediatric inpatients.

This study therefore aimed (1) to assess the prevalence, comorbidities, metabolic complications, and mortality among Thai children hospitalized with obesity between 2015 and 2023, and (2) to evaluate the impact of the COVID-19 pandemic on trends in pediatric obesity and on disparities in associated comorbidities and metabolic complications across the pre-, during-, and post-pandemic periods.

Methods

Study design and participants

This retrospective nationwide study utilized data from the National Health Security Office (NHSO) database, encompassing data on pediatric hospital admissions across Thailand from January 2015 to December 2023. The NHSO administers the Universal Health Coverage (UHC) scheme, Thailand’s largest public health insurance program, which covers approximately 72% of the national population who are not insured under other government health insurance schemes [8]. The Thai healthcare system comprises public and private hospitals stratified into primary, secondary, and tertiary levels. Primary hospitals provide initial contact and essential medical care. Secondary hospitals offer broader diagnostic and therapeutic services, although pediatric subspecialty care and intensive care may be limited. Tertiary hospitals serve as referral centers equipped with pediatric intensive care units (PICUs), subspecialty, and advanced surgical facilities. Private hospitals vary in their service capacity, with some operating at a tertiary level.

Children aged 1 month to < 18 years who were hospitalized, regardless of obesity status, were included in the study. Obesity was identified according to the International Classification of Diseases, 10th Revision, Thai Modification (ICD-10-TM) code E66.x, whether recorded as a primary or secondary diagnosis. ICD-10-TM codes were obtained from discharge diagnosis fields in the claims database. We searched all available diagnosis positions (primary and secondary) for the relevant codes and applied the same code-based definitions across hospitals and study years. Obesity codes recorded as secondary diagnoses were treated equivalently to those recorded as the primary diagnosis.

To ensure accurate estimates of prevalence and comorbidity, each patient was counted only once. A child was considered to have obesity of at least one inpatient admission with an ICD-10-TM E66.x code occurred during the study period, and only the first qualifying admission (the index event) was used for analysis. Subsequent admissions with an obesity diagnosis were not double counted for that individual. Children without any E66.x code during the study period were included as the comparison group and contributed one record per child, defined using their first observed hospitalization during the study period. The study period was divided into three phases in accordance with the COVID-19 pandemic timeline: pre-pandemic period (before March 2020), pandemic (March 2020 to April 2023), and post-pandemic (from May 2023 onward) [9, 10].

Data collection

Demographic and clinical data were extracted from the NHSO database using ICD-10-TM diagnostic codes, including age, sex, geographic region, and hospital service level. Comorbidities were recorded at the patient level, such that each diagnosis was counted only once per individual regardless of the number or timing of admissions.

Conditions (comorbidities and complications) were identified using prespecified ICD-10-TM code lists recorded in any diagnosis position at discharge. For each child, a condition was coded as present if it appeared at least once in any admission record. The following comorbidities were identified: hypertension (I10), asthma (J45), obstructive sleep apnea (G47.33), polycystic ovary syndrome (E28.2), gastroesophageal reflux disease (K21), depression and/or anxiety disorders (F32, F33, F41.9), attention-deficit/hyperactivity disorder (F90), acanthosis nigricans (L83), hypothyroidism (E03), and orthopedic conditions, including Blount disease (M92.5), slipped capital femoral epiphysis (M93.0). Obesity-related complications included type 2 diabetes mellitus (E11), hyperglycemia or prediabetes (R73), nonalcoholic fatty liver disease (K76.0), nonalcoholic steatohepatitis (K75.81), lipid disorders (E78), and metabolic syndrome (E88). Common causes of in-hospital mortality were categorized as cardiovascular diseases (I51), respiratory diseases (J00–J99), and infectious diseases (A00–B99).

Statistical analysis

Statistical analyses were performed using Stata version 18 (StataCorp LLC, College Station, TX, USA) and visualized using R version 4.5.1 (R Foundation for Statistical Computing, Vienna, Austria). Categorical variables were summarized as frequencies and percentages, while continuous variables were presented using both means with standard deviations (SDs) and medians with interquartile ranges (IQRs) to provide a comprehensive description of their distribution.

Each record in the NHSO database represents a hospitalization (admission-level claim). For this analysis, we converted admission-level records to a patient-level cohort by linking admissions using a unique patient identifier and selecting a single index hospitalization per child. Accordingly, the unit of analysis was individual children (one record per child), not admissions or child-years. The annual prevalence of pediatric obesity was defined as the number of unique children (not admissions) with at least one obesity diagnosis per 10,000 children in the NHSO database for that year. Similarly, the prevalence of comorbidities and complications was estimated at the patient level, with each individual counted only once per condition. In-hospital mortality prevalence for children with obesity was defined as all-cause deaths per 1,000 unique pediatric patients with obesity.

Annual trend in obesity prevalence from 2015 to 2023 was assessed using logistic regression, with results expressed as the odds ratios (ORs) per calendar year. Differences in proportions across study phases were evaluated with chi-square tests. Associations between obesity and comorbidities or complications were examined using multiple logistic regression and reported as adjusted odds ratios (aORs) with 95% confidence intervals (CIs). Cox proportional hazards regression model was applied to identify factors associated with all-cause mortality, with age as the underlying timescale and adjustment for covariates including obesity status, sex, geographic region, and major causes of death (cardiovascular, respiratory, infectious). Results are presented as crude and adjusted hazard ratios (HRs) with 95% CIs. A P-value < 0.05 was considered statistically significant. Missing data were handled using a complete-case approach, excluding records with missing values in model covariates. We accounted for within-hospital correlation by clustering on hospital in all regression models and used the resulting cluster-adjusted standard errors to compute 95% confidence intervals and P-values.

Ethics approval

The study protocol was reviewed and approved by the Human Research Ethics Committee of Khon Kaen University (approval number: HE681378). As the study utilized anonymized data, the requirement for obtaining informed consent was waived by the committee.

Results

Characteristics of participants and temporal trends in obesity, 2015–2023

Over the nine-year period from 2015 to 2023, we analyzed 14,483,566 unique hospitalized Thai children aged 1 month to < 18 years. Of these, 42,168 children had a diagnosis of obesity, yielding an overall period prevalence of 29.1 per 10,000 hospitalized children (95% confidence interval [CI]: 28.8–29.4).

The demographic profile of hospitalized children with obesity differed markedly from that of children admitted for other reasons (Table 1). The obesity cohort was significantly older, with a median age of 10 years (IQR: 6–13), compared to a median age of 2 years (IQR: 0–8) in the non-obesity (children without obesity) cohort. Over 82% of children with an obesity diagnosis were in school-aged children and adolescents (5–18 years), whereas infants (< 1 year) constituted the largest single group of children without obesity (38.9%).

Table 1.

General characteristics of hospital admissions among Thai children aged 1 month to < 18 years, 2015–2023

Characteristics	Total (n = 14,483,566)	Obesity (n = 42,168)	Non-obesity (n = 14,441,398)
Age group (years)
< 1, n (%)	5,616,117 (38.8)	372 (0.9)	5,615,745 (38.9)
1–<5, n (%)	3,575,316 (24.7)	6,880 (16.3)	3,568,436 (24.7)
5–<13, n (%)	3,164,889 (21.9)	22,036 (52.3)	3,142,853 (21.8)
13–<18, n (%)	2,127,244 (14.7)	12,880 (30.5)	2,114,364 (14.6)
Age at diagnosis (years)
Mean ± SD	4.6 ± 5.5	9.6 ± 4.5	4.6 ± 5.5
Median (IQR)	2 (0, 8)	10 (6, 13)	2 (0, 8)
Sex, n (%)
Male	7,712,670 (53.3)	26,811 (63.6)	7,685,859 (53.2)
Female	6,770,896 (46.8)	15,357 (36.4)	6,755,539 (46.8)
Region, n (%)
Bangkok	928,526 (6.4)	8,093 (19.2)	920,433 (6.4)
Central	3,405,846 (23.5)	11,454 (27.2)	3,394,392 (23.5)
Northeast	4,984,491 (34.4)	11,220 (26.6)	4,973,271 (34.4)
North	2,384,733 (16.5)	5,749 (13.6)	2,378,984 (16.5)
South	2,779,896 (19.2)	5,652 (13.4)	2,774,244 (19.2)
Hospital level, n (%)
Primary	1,124,039 (7.8)	2,138 (5.1)	1,121,901 (7.8)
Secondary	8,867,083 (61.2)	15,062 (35.7)	8,852,021 (61.3)
Tertiary	4,182,753 (28.9)	24,545 (58.2)	4,158,208 (28.8)
Private	309,691 (2.1)	423 (1.0)	309,268 (2.1)

Disparities by sex and geography were evident. Males accounted for nearly two-thirds (63.6%) of children with obesity, a greater proportion than in the non-obesity group (53.2%). Geographically, admissions with obesity were concentrated in Bangkok, which represented 19.2% of the obesity cohort but only 6.4% of the non-obesity cohort. Children with obesity were also disproportionately managed in tertiary care centers (58.2% of obesity group vs. 28.9% of non-obesity group).

The prevalence of obesity among hospitalized children increased substantially over the nine-year study period, but this trend was not linear (Fig. 1). In the pre-pandemic period (2015–2019), the prevalence of obesity rose modestly, increasing from 2.5 (95% CI 2.4–2.6) per 10,000 children in 2015 to 3.3 (95% CI 3.2–3.4) per 10,000 children in 2019. The onset of the COVID-19 Pandemic marked a significant inflection point. Following a transient dip in 2020, likely attributable to changes in hospital utilization, the prevalence of obesity accelerated dramatically. The rate surged from 3.3 per 10,000 children in 2020 to a peak of 6.1 (95% CI 5.9–6.2) per 10,000 children in 2022. Overall, the prevalence nearly doubled from the pre-pandemic era to its peak in 2022. Logistic regression analysis indicated a statistically significant increase in obesity prevalence by an average of 13.1% each year (odds ratio [OR] per year 1.131, 95% CI 1.127–1.135; P < 0.001).

Fig. 1.

Annual Prevalence of Obesity Among Hospitalized Thai Children, 2015–2023. The figure displays the annual prevalence of obesity (per 10,000 hospitalized children) among children aged 1month to < 18 years. Circles represent the observed annual prevalence. The solid line shows the predicted trend from a logistic regression model, with the shaded area representing the 95% confidence interval (CI)

Markedly increased burden of cardiometabolic and other comorbidities

Hospitalized children with obesity had a significantly higher prevalence of a wide range of comorbidities compared to those without obesity (Table 2). The associations were strongest for conditions directly linked to excess adiposity. The odds of having a nonalcoholic fatty liver disease (NAFLD) diagnosis were over 200-fold higher among children with obesity (aOR 223.67, 95% CI 121.22–412.71). Similarly, the odds of obstructive sleep apnea (OSA) and weight-related orthopedic conditions (including Blount disease and slipped capital femoral epiphysis [SCFE]) were markedly elevated (aOR 80.89, 95% CI 51.82–126.26; and aOR 117.59, 95% CI 66.82–206.93, respectively). These very large effect estimates coincided with low absolute prevalence in children without obesity, as reflected in the absolute risks reported in Table 2.

Table 2.

Common morbidities and metabolic complications among Thai children with and without obesity, 2015–2023

Condition	Obesity (n = 42,168), n	Non-obesity (n = 14,441,398), n	Absolute risk (%)	aOR (95% CI)	P-value
T2DM					< 0.001
Yes	965	6,203	13.46	13.48 (4.85, 37.47)
No	41,203	14,435,195	0.28	Ref
Hypertension					< 0.001
Yes	3,444	21,072	14.05	38.42 (24.80, 59.50)
No	38,724	14,420,326	0.27	Ref
Dyslipidemia					< 0.001
Yes	262	590	30.75	28.08 (9.45, 83.50)
No	41,906	14,440,808	0.29	Ref
NAFLD					< 0.001
Yes	723	531	57.66	223.67 (121.22, 412.71)
No	41,445	14,440,867	0.29	Ref
OSA					< 0.001
Yes	5,824	22,178	20.80	80.89 (51.82, 126.26)
No	36,344	14,419,220	0.25	Ref
Orthopedic: Blount, SCFE					< 0.001
Yes	272	715	27.56	117.59 (66.82, 206.93)
No	41,896	14,440,683	0.29	Ref
Depression/Anxiety					0.051
Yes	247	45,644	0.54	1.71 (0.99, 2.93)
No	41,921	14,395,754	0.29	Ref
Asthma					< 0.001
Yes	3,511	256,770	1.35	4.35 (4.23, 4.48)
No	38,657	14,184,628	0.27	Ref

T2DM type 2 diabetes mellitus, NAFLD Non-alcoholic fatty liver disease, OSA Obstructive sleep apnea, SCFE slipped capital femoral epiphysis, aOR adjusted odds ratio (accounted for clustering at hospital levels), Ref reference category

Strong associations were also observed for classic cardiometabolic risk factors. Hypertension was over 38 times more likely in the obesity group (aOR 38.42, 95% CI 24.80–59.50), and dyslipidemia was 28 times more likely (aOR 28.08, 95% CI 9.45–83.50). Children with obesity also had 13.5-fold higher odds of having type 2 diabetes mellitus (T2DM) (aOR 13.48, 95% CI 4.85–37.47). The odds of asthma and depression/anxiety were also significantly increased (aOR 4.35 and 1.71, respectively).

Association between obesity and in-hospital mortality

The overall in-hospital mortality rate among 42,168 children with obesity was 4.7 per 1,000 children. The association between obesity and the hazard of in-hospital death was explored using Cox proportional hazard models (Table 3). In the univariable model, an obesity diagnosis was not significantly associated with the risk of in-hospital death (Hazard Ratio [HR] 1.06, 95% CI 0.78–1.44). After adjusting for sex, region, and major comorbid disease categories, the point estimate suggested a potential 13% lower hazard of death for children with obesity, but this association did not reach statistical significance (aHR 0.87, 95% CI 0.65–1.16; P = 0.344).

Table 3.

Factors associated with mortality among hospitalized children in Thailand, 2015–2023

	Death (n = 39,615)	Alive (n = 14,443,951)	cHR (95% CI)	aHR (95% CI)	P-value
Sex					0.001
Female	16,242 (41.0)	6,754,654 (46.8)	Ref	Ref
Male	23,373 (59.0)	7,689,297 (53.2)	1.14 (1.07, 1.23)	1.13 (1.05, 1.21)
Region					< 0.001
Bangkok	4,594 (11.6)	923,932 (6.4)	Ref	Ref
Central	10,513 (26.5)	3,395,333 (23.5)	1.03 (0.76, 1.38)	1.03 (0.75, 1.42)
Northeast	11,217 (28.3)	4,973,274 (34.4)	0.98 (0.46, 2.08)	0.94 (0.45, 1.96)
North	5,449 (13.8)	2,379,284 (16.5)	0.93 (0.68, 1.28)	0.89 (0.67, 1.19)
South	7,842 (19.8)	2,772,054 (19.2)	1.11 (0.64, 1.90)	1.07 (0.63, 1.81)
Obesity	198 (0.5)	41,970 (0.3)	1.06 (0.78, 1.44)	0.87 (0.65, 1.16)	0.344
Cardiovascular disease	413 (1.0)	5,122 (0.04)	7.17 (5.00, 10.27)	5.86 (3.57, 9.61)	< 0.001
Respiratory disease	17,663 (44.6)	3,502,134 (24.3)	1.96 (1.38, 2.79)	1.93 (1.40, 2.66)	< 0.001
Infection	8,385 (21.2)	2,659,313 (18.4)	1.18 (0.89, 1.56)	1.12 (0.96, 1.30)	0.139

cHR crude hazard ratio, aHR adjusted hazard ratio, 95% CI 95% confidence interval, Ref reference category

The analysis accounted for clustering at the hospital levels

Other factors were strong predictors of mortality. Male sex was associated with a 13% higher hazard of death (aHR 1.13, 95% CI 1.05–1.21). Compared to admission in Bangkok, those in the North and Northeast regions had a lower mortality risk. The presence of a cardiovascular disease diagnosis conferred the highest risk of death (aHR 5.86, 95% CI 3.57–9.61).

Impact of the COVID-19 pandemic on the prevalence and demographics of pediatric obesity

The COVID-19 pandemic was associated with a dramatic acceleration in the prevalence of obesity among hospitalized children. The rate more than doubled from a pre-pandemic average of 20.10 per 10,000 children to 45.25 per 10,000 during the pandemic and remained elevated at 45.54 per 10,000 in the post-pandemic period. The surge was accompanied by a demographic shift. While the 5–to-<13-year age group remained the largest proportion, the contribution from adolescents (13–<18 years) increased substantially, rising from 25.3% of the obesity cohort pre-pandemic to over 30–35% during- and post-pandemic. A geographic shift was also noted, with the Northeast region surpassing the Central region and Bangkok in representing the largest share of obesity admissions post-pandemic (Supplementary Material 1).

Shifting patterns of comorbidity during the pandemic

The pandemic period was associated with significant, and at times paradoxical, changes in the profile of documented comorbidities among children with obesity (Table 4). We identified three distinct patterns.

Table 4.

Comparison of obesity-related comorbidities and metabolic complications among children with obesity by pandemic period

Condition	Total (n = 42,168), n (%)	Pandemic periods			OR (95% CI) (2) vs. (1)	OR (95% CI) (3) vs. (1)	OR (95% CI) (2) & (3) vs. (1)	P-value
		Pre (1) (n = 18,706), n (%)	During (2) (n = 19,209), n (%)	Post (3) (n = 4,253), n (%)
Comorbidities
Hypertension	3,444 (8.2)	1,644 (8.8)	1,458 (7.6)	342 (8.0)	0.85 (0.63, 1.16)	0.91 (0.80, 1.03)	0.86 (0.66, 1.13)	0.169
Asthma	3,511 (8.3)	1,473 (7.9)	1,546 (8.1)	492 (11.6)	1.02 (0.91, 1.15)	1.53 (1.33, 1.76)	1.11 (1.03, 1.20)	< 0.001
Obstructive sleep apnea	5,824 (13.8)	3,248 (17.4)	1,982 (10.3)	594 (14.0)	0.55 (0.42, 0.71)	0.77 (0.74, 0.81)	0.59 (0.47, 0.73)	< 0.001
PCOS	98 (0.2)	39 (0.2)	51 (0.3)	8 (0.2)	1.27 (0.91, 1.78)	0.90 (0.61, 1.33)	1.21 (0.86, 1.69)	< 0.001
GERD	191 (0.5)	78 (0.4)	89 (0.5)	24 (0.6)	1.11 (0.97, 1.28)	1.36 (0.90, 2.03)	1.16 (1.08, 1.24)	< 0.001
Depression/Anxiety	247 (0.6)	72 (0.4)	135 (0.7)	40 (0.9)	1.83 (1.74, 1.93)	2.46 (1.91, 3.16)	1.94 (1.77, 2.13)	< 0.001
ADHD	798 (1.9)	320 (1.7)	369 (1.9)	109 (2.6)	1.13 (0.92, 1.37)	1.51 (1.11, 2.06)	1.19 (0.96, 1.49)	< 0.001
Acanthosis nigricans	64 (0.2)	40 (0.2)	16 (0.1)	8 (0.2)	0.39 (0.30, 0.50)	0.88 (0.46, 1.69)	0.48 (0.34, 0.67)	< 0.001
Hypothyroidism	266 (0.6)	117 (0.6)	125 (0.7)	24 (0.6)	1.04 (0.56, 1.95)	0.90 (0.46, 1.77)	1.02 (0.54, 1.91)	< 0.001
Orthopedic: Blount, SCFE	272 (0.7)	158 (0.8)	89 (0.5)	25 (0.6)	0.55 (0.45, 0.67)	0.69 (0.65, 0.74)	0.57 (0.48, 0.68)	< 0.001
Metabolic complications
T2DM	965 (2.3)	371 (2.0)	484 (2.5)	110 (2.6)	1.28 (1.04, 1.57)	1.31 (1.29, 1.34)	1.28 (1.09, 1.52)	< 0.001
Hyperglycemia/Prediabetes	620 (1.5)	302 (1.6)	265 (1.4)	53 (1.3)	0.85 (0.66, 1.10)	0.77 (0.67, 0.88)	0.84 (0.66, 1.06)	< 0.001
NAFLD	723 (1.7)	284 (1.5)	334 (1.7)	105 (2.5)	1.15 (0.90, 1.47)	1.64 (1.54, 1.75)	1.24 (1.00, 1.52)	< 0.001
NASH	136 (0.3)	58 (0.3)	61 (0.3)	17 (0.4)	1.02 (0.84, 1.25)	1.29 (1.15, 1.45)	1.07 (0.89, 1.29)	< 0.001
Lipid disorders	2,164 (5.1)	1,067 (5.7)	894 (4.7)	203 (4.8)	0.81 (0.51, 1.27)	0.83 (0.57, 1.21)	0.81 (0.52, 1.26)	0.621
Metabolic syndrome	524 (1.2)	314 (1.7)	183 (1.0)	27 (0.6)	0.56 (0.37, 0.86)	0.37 (0.31, 0.45)	0.53 (0.36, 0.78)	< 0.001

GERD gastroesophageal reflux disease, NAFLD nonalcoholic fatty liver disease, NASH nonalcoholic steatohepatitis, OR odds ratio (accounted for clustering at hospital levels), PCOS polycystic ovary syndrome, SCFE slipped capital femoral epiphysis, T2DM type 2 diabetes mellitus, 95% CI 95% confidence interval

First, increased prevalence of conditions reflecting metabolic decompensation and mental distress. The odds of T2DM were significantly higher both during and after the pandemic compared to the pre-pandemic period (combined aOR 1.28, 95% CI 1.09–1.52). Similarly, the prevalence of NAFLD increased, particularly in the post-pandemic period (OR 1.64, 95% CI 1.54–1.75). The pandemic also coincided with a sharp rise in mental health diagnoses, with the odds of depression or anxiety nearly doubling (combined aOR 1.94, 95% CI 1.77–2.13).

Second, a decreased prevalence of conditions reliant on screening or non-urgent evaluation. Several key comorbidities showed a marked decrease in documented prevalence. The odds of an OSA diagnosis fell by 41% (combined aOR 0.59, 95% CI 0.47–0.73), and weight-related orthopedic conditions fell by 43% (combined aOR = 0.57, 95% CI 0.48–0.68). Diagnoses of metabolic syndrome also significantly declined. These reductions likely reflect disruptions in non-urgent healthcare, reduced elective surgeries, and decreased routine screening during the pandemic rather than a true improvement in these conditions.

Third, an increase in post-pandemic asthma was observed. Asthma diagnoses were stable during the pandemic but increased significantly in the post-pandemic period (OR 1.53, 95% CI 1.33–1.76), suggesting a potential delayed impact or change in healthcare-seeking behavior for respiratory conditions. No significant temporal differences were observed for hypertension, and lipid disorders, suggesting relative stability across pandemic phases.

Discussion

This nationwide analysis reveals a persistent increase in pediatric obesity among hospitalized Thai children from 2015 to 2023, with the sharpest acceleration observed during the COVID-19 pandemic. Importantly, this trend was accompanied by shifts in demographic patterns and a marked rise in metabolic and psychological comorbidities, reflecting global reports of the pandemic’s impact on child health.

The hospital-based prevalence of obesity peaked in 2022, a period that coincided with the COVID-19 pandemic, supporting the hypothesis that pandemic-related factors exacerbated risk for obesity [11]. Multiple mechanisms likely contributed to the rise. Lockdown, school closures, and public health restrictions disrupted daily routines and reduced opportunities for structured activity and play, while sedentary behaviors increased, especially screen time [12]. Social isolation and home confinement increased sedentary behavior, and heightened stress, which contributed to more frequent snacking and stress-related overeating [12, 13]. Many families also faced financial strain and food insecurity, which limited access to healthy foods and further worsened diet quality [12, 14]. Routine medical visits and weight-management follow-ups were often postponed or canceled [15]. Collectively, these factors likely created an “obesogenic” environment during the pandemic that strongly favored weight gain [12, 16].

The upward pattern is broadly consistent with global evidence of accelerated pediatric weight gain during the pandemic. In the United States, a large CDC analysis reported an increase in childhood obesity prevalence from about 19% pre-pandemic to more than 22% by August 2020, alongside a near doubling in the rate of BMI increase [12]. In Canada, children enrolled in a weight-control program in Ontario experienced greater than expected increases in BMI and weight during the pandemic period [17]. Early data from China reported increases in overweight prevalence among high school students from 26.7% before lockdown to 30.4% after lockdown, and obesity from 16.1% to 19.3% [18]. In South Korea, reduced physical activity during school closures was associated with increases in BMI and metabolic markers during remote learning [6].

The persistence of elevated obesity prevalence after the pandemic suggests that these changes are not transient. Rather, they may reflect lasting behavioral and environmental adaptations. Researchers have expressed concern that prolonged sedentary lifestyles and disrupted routines may have become a “new normal,” making obesity-promoting behaviors more ingrained even after the resumption of daily activities [13].

We observed a shift toward older age groups and a regional redistribution from Central to Northeast Thailand. The age shift aligns with global observations that adolescents experienced the greatest increase in sedentary time and sleep disruptions during the pandemic [19–21]. These sleep disturbances, often associated with stress and high screen exposure, may worsen metabolic and appetite regulation and thereby increase the risk of obesity in adolescents [11].

Regional differences in pediatric obesity hospitalizations, shifting from the Central to the Northeast region, likely reflect disparities in pandemic response, socioeconomic burden, and healthcare capacity. In the Northeast, while local volunteers and community stakeholders mobilized resources for COVID-19 prevention [22], prolonged service disruptions and reduced physical activity likely spurred pediatric weight gain. Northeastern Thailand is the poorest and least-developed region of the country [23], and food insecurity was highest there during the pandemic [24], increasing reliance on inexpensive, calorie-dense foods. Furthermore, lower physician density in the rural Northeast [25], restricted access to preventive and early obesity care. These combined socioeconomic and healthcare disparities may have contributed to higher hospitalization rates, whereas the Central region’s stronger health infrastructure supported a faster recovery and earlier resumption of preventive care and healthier routines.

In our analysis of all hospitalized children, obesity was not associated with increased in-hospital mortality. After adjustment for confounders and accounting for clustering at the hospital level, the hazard of death was lower among children with obesity than among those without obesity, but this association was not statistically significant (aHR 0.87, 95% CI 0.65–1.16; P = 0.344). The point estimate below 1.0 should be interpreted cautiously. In administrative hospital data, residual confounding by indication, differences in admission thresholds, and variation in care intensity for children coded as having obesity may all influence the observed association. Similar patterns have been reported in pediatric inpatient and critical care studies, where obesity was not significantly associated with short-term mortality after risk adjustment [26, 27].

Several comorbidities showed notable temporal shifts, including increases in type 2 diabetes mellitus (T2DM), nonalcoholic fatty liver disease (NAFLD), asthma, and mental health disorders.

A growing body of evidence indicates a marked rise in pediatric T2DM during the COVID-19 pandemic. In the United States, the incidence of new-onset T2DM increased by 77% during the first pandemic year compared with preceding years [28]. Investigators attribute this surge to pandemic-related behavioral changes such as reduced physical activity, increased sedentary time, irregular sleep, and higher consumption of processed foods [29], which collectively promote insulin resistance and weight gain. Consequently, children with obesity were disproportionately affected, exhibiting significantly greater risk of developing T2DM during and after the pandemic.

Similar trends were observed for NAFLD, a common obesity-related liver condition. A study at Yale found NAFLD prevalence rose from 36.2% pre-pandemic to 60.9% during the pandemic among youth with obesity [30]. Moreover, children evaluated during the pandemic had greater hepatic fat content and visceral adiposity than those assessed pre-pandemic [30]. These findings suggest that pandemic-related lifestyle disruptions exacerbated metabolic dysfunction and hepatic steatosis in pediatric populations. However, the very large odds ratios observed in our hospital-based, code-ascertained analyses should be interpreted cautiously because they may be amplified by sparse-data bias, differential coding practices, and surveillance effects if children with obesity undergo more frequent evaluation and documentation for liver disease.

Pandemic-related stress and uncertainty also worsened youth mental health. Rates of depression and anxiety rose sharply, particularly among children with obesity. A 2023 systematic review of longitudinal studies found significant increases in depressive symptoms and modest increases in anxiety among children and adolescents during the pandemic [31]. Many adopted maladaptive coping behaviors, further compounding weight gain and metabolic risk. However, administrative data likely underestimate the true burden because they capture only cases that are recognized in care and coded. In one study that compared survey estimates with linked administrative data, prevalence based on claims was much lower. For anxiety, the claims-based estimate was more than four times lower than the survey estimate, which suggests under-detection and under coding [32]. In Thailand, stigma and shame around mental health may reduce help-seeking. Child and adolescent specialist services are also limited. This may limit assessment and diagnosis, which could lower the number of recorded cases in administrative data and affect trends over time [33]. During COVID-19, service disruptions and the shift to telemedicine may have further limited assessment and follow-up, which could affect recorded diagnoses [34]. As a result, trends in coded depression and anxiety likely reflect both underlying need and changes in access, service delivery, and coding practices.

Asthma likewise showed a post-pandemic uptick, particularly among children with obesity. One study reported an increase in status-asthmaticus admissions from 5.7 to 7.3 per 100,000 children in the year following the lifting of COVID-19 restrictions [35]. In a U.S. cohort of 267 children with asthma, monthly body mass index (BMI) gain was significantly greater during the pandemic year, with the steepest increases observed in those already overweight or obese [36]. This trend effectively expanded the population of children with concurrent obesity and asthma.

In contrast, diagnoses of obstructive sleep apnea (OSA), and metabolic syndrome declined during the pandemic. The widespread suspension of sleep laboratory operations significantly reduced OSA detection. A global survey of 297 sleep centers in 2020 found that 93.6% halted nearly all testing during the pandemic’s onset [37]. At one pediatric hospital, the median time from referral to definitive OSA diagnosis nearly doubled, from about 54 days pre-pandemic to 103 days during the pandemic (P < 0.001) [38]. Reduced access to diagnostic services and deferred otolaryngologic procedures likely delayed evaluation and treatment. Similarly, pediatric metabolic conditions were underdiagnosed during this period. Screening for metabolic syndrome declined substantially due to fewer clinic visits and postponed preventive care. One study reported that 41.3% of parents indicated that their youngest child missed a routine medical visit during this time [39]. Collectively, these findings illustrate how the COVID-19 pandemic not only intensified existing pediatric metabolic and mental health burdens but also hindered timely diagnosis and management of chronic conditions.

This study’s primary strength is its use of a comprehensive, nationally representative hospital database over nine years, enabling robust long-term trend analysis. Standardized ICD-10-TM coding and stratification by pandemic phases enhance internal validity and temporal comparison. In addition, the broad inclusion of comorbidities and mortality outcomes offers a multifaceted view of the health burden posed by pediatric obesity in hospital setting in Thailand. However, several limitations must be acknowledged. First, the hospital-based nature of the data may overrepresent children with more severe diseases and does not capture the population prevalence of obesity or mild undiagnosed cases. Reliance on administrative coding may result in underreporting or misclassification of obesity and comorbidities, especially during period of healthcare disruption. In particular, COVID-19–related reductions in in-person visits, screening, and diagnostic workups may have lowered case detection and coding completeness for chronic conditions. Therefore, apparent declines in prevalence during the pandemic should be interpreted as decreases in recorded diagnoses rather than true reductions in underlying disease burden. Because the NHSO claims database primarily reflects Universal Health Coverage (UHC) scheme beneficiaries, children insured under other schemes were not represented. As the UHC scheme covers the majority of the Thai population, our findings are likely broadly relevant at the national level, but generalizability to children covered through other schemes may be limited due to potential differences in socioeconomic profile, care-seeking, coding practices, and provider networks. Our data lack individual-level anthropometric measurements and risk stratification by obesity severity, precluding more granular risk analysis. Furthermore, socioeconomic factors, lifestyle behaviors, and outpatient follow-up information were unavailable. We also prespecified pandemic-phase cutoffs based on WHO and did not conduct sensitivity analyses using alternative transition dates. As a result, estimates around the transition between phases may be sensitive to phase misclassification, and we cannot quantify the robustness of findings to different cutoff definitions. Finally, this observational design precludes causal inferences, observed associations may be influenced by unmeasured confounders and temporal trends in healthcare utilization.

Conclusion

Pediatric obesity among hospitalized Thai children increased substantially between 2015 and 2023, with the COVID-19 pandemic markedly accelerating this trend and exacerbation of associated metabolic and mental health complications. Regional variation was evident, with an increase in the Northeast. However, our dataset did not include food insecurity or other contextual socioeconomic determinants, so the drivers of these regional patterns cannot be determined. While short-term mortality was not increased, the heightened burden of comorbidity underscores an urgent need for integrated, equity-focused prevention and management. These findings highlight the necessity of restoring and strengthening school, community, and healthcare-based intervention, particularly those address the behavioral, metabolic, and psychosocial risk factors to mitigate the enduring impact of the pandemic on child and adolescent health in Thailand.

Authors’ contributions

Conceptualization and Design: SS, RU, LT, PS, PK. Data Acquisition: TL, SS, RU. Data Analysis and Interpretation: TL, SS, PS, LT, RU, PK. Initial Drafting: TL, SS. Manuscript Revision: RU, PS, LT, PK. Accountability: all authors attest to their review of the final manuscript and their accountability for all aspects of the research.

Funding

This study was supported by the Fundamental Fund, Khon Kaen University, Thailand (FF2569). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data availability

The datasets generated and/or analyzed during the current study are not publicly available but are available from the corresponding author (SS) upon request.

Declarations

Ethics approval and consent to participate

Ethical approval for this research was obtained from the Khon Kaen University Ethics Committee for Human Research (Approval No. HE681378, date Jun 15, 2025). This study was conducted in accordance with the Declaration of Helsinki and the principles of Good Clinical Practice. The need for informed consent was waived by the Khon Kaen University Ethics Committee for Human Research as the study involved the secondary analysis of de-identified administrative data from the National Health Security Office, posing no more than minimal risk to the participants.

Consent for publication

Not applicable. The manuscript does not contain any individual person’s data in any form (including individual details, images, or videos). The research was conducted using an anonymized administrative database provided by the NHSO; therefore, informed consent for publication was not required.

Competing interests

The authors declare no competing interests.