Positive childhood experiences and their association with behavioral problems in Thai early adolescents

Abstract

BACKGROUND

Positive childhood experiences (PCEs) may promote resilience and reduce behavioral and emotional problems, but evidence from Asian low- and middle-income countries is limited.

AIM

To describe PCE prevalence and its associations with parent-reported behavioral and emotional problems among early adolescents in Bangkok, Thailand.

METHODS

A cross-sectional school-based survey in Bangkok enrolled adolescents aged 10-13 years and their parents (n = 540). Adolescents completed a Thai-validated 7-domain PCE measure; parents completed the Strengths and Difficulties Questionnaire. Multivariable logistic regression estimated adjusted odds ratios (aORs) with 95% confidence intervals (CIs). Receiver operating characteristic analyses evaluated the discrimination of PCE counts.

RESULTS

Participants’ mean age was 11.4 years; 54.8% were female, and 81.1% reported ≥ 4 PCEs. Parent-reported the Strengths and Difficulties Questionnaire risk/problem prevalence ranged from 13.0% to 22.2%. In multivariable models, mentorship was associated with a lower odds of emotional problems (aOR = 0.14, 95%CI: 0.06-0.32), hyperactivity (aOR = 0.35, 95%CI: 0.16-0.77), and low prosocial behavior (aOR = 0.09, 95%CI: 0.04-0.20). The opportunity to share ideas with caregivers was also protective for emotional problems (aOR = 0.13, 95%CI: 0.05-0.35) and hyperactivity (aOR = 0.07, 95%CI: 0.03-0.17). Having ≥ 4 PCEs was associated with lower odds of emotional problems (aOR = 0.26, 95%CI: 0.08-0.83) and hyperactivity (aOR = 0.32, 95%CI: 0.11-0.98). Receiver operating characteristic analyses showed acceptable-to-good discrimination; for the emotional problems area under the curve was 0.835 (95%CI: 0.790-0.881).

CONCLUSION

PCEs, particularly mentorship and caregiver-child idea sharing, were robustly associated with fewer behavioral and emotional difficulties in Thai early adolescents, supporting culturally relevant resilience-promoting strategies.

Key Words: Positive childhood experiences; Behavioral problems; Early adolescents; Protective factors; Thailand

Core Tip: Positive childhood experiences (PCEs) are increasingly recognized as resilience factors, yet evidence from Southeast Asia is scarce. In a Bangkok school-based sample, we mapped seven PCE domains and linked them to the parent-rated Strengths and Difficulties Questionnaire outcomes. Mentorship and caregiver-child opportunity to share ideas showed the strongest, consistent protection across emotional and hyperactivity domains. A higher cumulative PCE count also meaningfully distinguished adolescents with emotional difficulties in receiver operating characteristic analysis (area under the curve). These findings highlight actionable relational targets for early prevention.

INTRODUCTION

Adverse childhood experiences (ACEs), such as abuse, neglect, or living with caregivers who have mental illness or substance abuse problems, are well-established risk factors for poor developmental and health outcomes. Children exposed to ACEs are more likely to experience academic difficulties, social challenges, and behavioral and emotional problems, including aggression, anxiety, and depression. They also face elevated risks for long-term physical and mental health issues. Evidence suggests that having four or more ACEs increases the likelihood of adolescent behavioral problems by three to four times compared with those without ACEs[1,2].

In recent years, research has shifted towards understanding the protective role of positive childhood experiences (PCEs), which include supportive relationships, safe environments, and opportunities for meaningful engagement. PCEs foster emotional resilience, healthy social connections, and adaptive coping skills, even in the presence of adversity[3,4]. Studies have shown that each additional PCE can reduce adolescent behavioral problems by approximately 18%[5], and that children with six or more PCEs have up to a 72% lower risk of adult depression and anxiety, even when ACEs are present[6]. These findings underscore the importance of not only preventing negative experiences but also actively promoting positive ones.

Early adolescence is a critical developmental stage characterized by rapid physical, psychological, and social transitions. PCEs during this period can mitigate behavioral risks, enhance emotional well-being, and promote lifelong health[7]. Large-scale studies in the United States have demonstrated that factors such as afterschool activities, resilient family environments, and supportive neighborhoods are associated with reduced anxiety, depression, and delinquent behavior in adolescents[8,9]. However, most evidence comes from Western contexts, and the applicability of these findings to Thailand, where cultural norms, family structures, and community support systems differ, remains unclear. This study is the first to examine the prevalence of PCEs among Thai early adolescents and their association with behavioral problems, aiming to inform culturally relevant strategies to foster positive development and reduce behavioral risks.

MATERIALS AND METHODS

Study design and population

This study examined the prevalence of PCEs and their associations with behavioral problems among early adolescents in Bangkok, Thailand. The sample size was calculated using an expected PCE prevalence of 65.2% from Choi et al[5], with 4% precision and a 5% significance level, yielding a minimum of 448 participants; to account for incomplete responses, 20% was added, resulting in a target sample of 540. Participants were adolescent-parent pairs, including adolescents aged 10-13 years, and both adolescents and parents were required to be able to read and understand Thai. To define a school-based non-clinical sample, pairs were excluded if the adolescent had a prior clinician-diagnosed neurodevelopmental or psychiatric disorder (e.g., autism spectrum disorder, attention deficit hyperactivity disorder, or depression).

Data collection and instruments

This study adhered to the principles of the Declaration of Helsinki, and the Institutional Review Board of the Faculty of Medicine, Siriraj Hospital, Mahidol University, approved the study (Approval No. Si 219/2025). Data were collected between April 2025 and November 2025. School administrators across Bangkok were contacted to facilitate participant recruitment. Using convenience sampling, adolescents were recruited from participating schools. Adolescents completed the PCE questionnaire, while parents completed the Strengths and Difficulties Questionnaire (SDQ). Completed forms were returned by mail in pre-addressed, stamped envelopes. All responses were anonymized using participant codes.

The PCE questionnaire[4], adapted from the United States National Survey of Children’s Health and validated in Thai, assesses seven domains of supportive relationships and environments during the first 18 years of life: Mentorship, family resilience, safe neighborhood, supportive neighborhood, afterschool activities, volunteerism, and the opportunity to share ideas with caregivers. Three items use a yes/no format, and four use a 4-point Likert scale. Internal consistency reliability coefficients were 0.87 for Likert-scale items and 0.75 for dichotomous items.

The parent-reported SDQ, validated for use in Thai children aged 4-16 years, comprises 25 items across five domains: Emotional symptoms, hyperactivity/inattention, conduct problems, peer relationship problems, and prosocial behavior. Each item is rated on a 3-point scale (0 = not true, 1 = somewhat true, 2 = definitely true), with domain scores categorized as normal, risk, or problem according to established cut-offs. The Thai SDQ has a sensitivity of 63.3%, specificity of 94.6%, and Cronbach’s alpha of 0.7[10].

Statistical analysis

Data were analyzed using SPSS version 25.0 (IBM Corp., Armonk, NY, United States). Descriptive statistics were used to summarize participant characteristics. Univariate logistic regression was conducted to examine associations between individual sociodemographic factors, PCE domains, and behavioral outcomes measured by the SDQ. Variables with P < 0.2 in univariate analysis were included in multivariable logistic regression models. Adjusted odds ratios (aORs) and 95% confidence intervals (CIs) were reported, with statistical significance set at P < 0.05. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the discriminative ability of the number of PCEs for risk/problematic behavioral outcomes. The area under the curve (AUC) quantified discrimination, and cutoff values were selected based on the sensitivity-specificity balance.

RESULTS

Participant characteristics

A total of 540 adolescents participated (mean age 11.43 ± 1.16 years), with 54.8% female and 51.9% in upper primary school. Just over half (50.4%) attended public schools, and most had a cumulative grade-point average (GPAX) > 3.00 (77.6%; GPAX on a 0-4.00 scale, higher scores indicate better academic performance). The majority reported no physical health problems (91.5%). Regarding caregivers, most were aged 41-50 years (65.7%), female (90.0%), married (95.6%), and the child’s mother (88.1%). Most fathers (90.2%) and mothers (87.2%) held at least a bachelor’s degree. Government employment was reported by 40.4% of fathers and 27.0% of mothers. Most families (89.4%) had a monthly income of > 1544 United States dollars (approximately 55000 Thai Baht). For reference, Thailand’s nationwide average monthly household income in 2023 was 29030 Thai Baht (National Statistical Office). Physical and mental health problems among caregivers were reported by 3.1% and 5.4%, respectively (Table 1).

Table 1 Demographic characteristics of adolescents and their caregivers, n (%).

Demographic characteristics	Results, n = 540
Age in years	11.43 ± 1.16
Sex
Male	244 (45.2)
Female	296 (54.8)
Education: Grade
Upper primary school: 4-6	280 (51.9)
Junior high school: 7-9	260 (48.1)
School type
Public school	272 (50.4)
Private school	268 (49.6)
GPAX
2.01-3.00	121 (22.4)
> 3.00	419 (77.6)
Physical health problems
No	494 (91.5)
Yes	46 (8.5)
Caregiver’s age in years
31-40	156 (28.9)
41-50	355 (65.7)
> 50	29 (5.4)
Caregiver’s sex
Male	54 (10.0)
Female	486 (90.0)
Caregiver’s marital status
Married	516 (95.6)
Divorce	24 (4.4)
Relationship with child
Mother	476 (88.1)
Father	43 (8.0)
Others	21 (3.9)
Father’s educational level
Below a bachelor’s degree	53 (9.8)
Bachelor’s degree and above	487 (90.2)
Mother’s educational level
Below a bachelor’s degree	69 (12.8)
Bachelor’s degree and above	471 (87.2)
Father’s occupation
Government employee	218 (40.4)
Private company employee	322 (59.6)
Mother’s occupation
Government employee	146 (27.0)
Private company employee	389 (72.0)
Unemployed	5 (0.9)
Family monthly income
617-1544 United States dollars1	57 (10.6)
> 1544 United States dollars1	483 (89.4)
Caregiver’s physical health problems	17 (3.1)
Caregiver’s mental health problems	29 (5.4)

¹1 United States dollar = 32.39 Thai Bahts.

GPAX: Cumulative grade-point average.

Prevalence and patterns of PCEs

The most common PCEs were sharing ideas with a caregiver (86.8%), having a mentor (83.3%), and family resilience (82.6%). The least common were volunteerism (29.6%) and having a supportive neighborhood (38.7%). Most adolescents (81.1%) reported four or more PCEs (Figure 1A and B).

Figure 1 Positive childhood experiences and behavioral-emotional problems among Thai early adolescents. A: Prevalence of individual positive childhood experiences among adolescents. Data represents the number of individuals in each domain with percentages labeled (n = 540). An individual could have more than one positive childhood experience domain; B: Number of positive childhood experiences experienced by adolescents. Data represent the number of individuals and the number of positive childhood experiences domains they possess, with percentages labeled (n = 540); C: Proportion of adolescents classified as at-risk/problem across the Strengths and Difficulties Questionnaire domains. PCEs: Positive childhood experiences.

Prevalence of behavioral and emotional difficulties (SDQ)

As shown in Figure 1C, when combining children at risk or problem levels, the highest proportion was found in the emotional problems (22.2%), followed by hyperactivity (20.9%), conduct problems (20.4%), and peer problems (13.7%). For prosocial behavior, 13.0% of children were at risk of low prosocial skills. No participants reached the problem level in the peer and prosocial domains.

Factors associated with behavioral and emotional problems

In univariate logistic regression analysis (Table 2), higher GPAX (> 3.00), higher parental education, and higher household income were significantly associated with reduced odds of risk or problematic levels across emotional, conduct, hyperactive, peer, and prosocial domains. Female sex was associated with lower odds of risk or problematic conduct problems (OR = 0.653, P = 0.047). Caregiver divorce was linked to higher odds across all domains. Several PCEs, especially sharing ideas with caregivers, showed strong protective effects (e.g., emotional problems OR = 0.059, P < 0.001). Having ≥ 4 PCEs was consistently associated with lower odds of risk or problematic behavioral and emotional outcomes.

Table 2 Univariate logistic regression results for the association between demographic characteristics and risk or problematic behavioral and emotional problems¹.

Factors	Risk or problematic behavioral and emotional problem
	Emotional problem	Conduct problem	Hyperactive problem	Peer problem	Prosocial behavior
Child’s factor
Sex
Male	Reference
Female	0.814 (0.542, 1.222); 0.321	0.653 (0.429, 0.994); 0.047a	0.732 (0.483, 1.109); 0.141a	1.028 (0.628, 1.684); 0.912	0.750 (0.453, 1.240); 0.262
Education
Upper primary school	Reference
Junior high school	1.100 (0.733, 1.651); 0.645	1.098 (0.722, 1.669); 0.663	1.074 (0.709, 1.626); 0.736	0.903 (0.552, 1.476); 0.683	0.732 (0.440, 1.217); 0.229
School type
Public school	Reference
Private school	0.896 (0.597, 1.345); 0.597	0.848 (0.558, 1.291); 0.443	0.761 (0.501, 1.154); 0.199a	0.791 (0.483, 1.295); 0.352	1.017 (0.616, 1.681); 0.947
GPAX
2.01-3.00	Reference
> 3.00	0.408 (0.261, 0.639); < 0.001a	0.534 (0.336, 0.853); 0.009a	0.427 (0.271, 0.674); < 0.001a	0.548 (0.321, 0.936); 0.028a	0.886 (0.492, 1.595); 0.686
Physical health problems
No	Reference
Yes	1.261 (0.632, 2.520); 0.511	1.094 (0.525, 2.281); 0.810	1.208 (0.593, 2.461); 0.603	0.751 (0.287, 1.967); 0.560	1.467 (0.654, 3.289); 0.352
Caregiver and family factors
Caregiver’s age in years
31-40	Reference
41-50	1.003 (0.629, 1.599); 0.990	0.955 (0.595, 1.534); 0.850	1.106 (0.689, 1.776); 0.676	1.073 (0.606, 1.900); 0.808	0.698 (0.405, 1.205); 0.197a
> 50	4.152 (1.818, 9.479); 0.001a	2.464 (1.056, 5.748); 0.037a	1.890 (0.783, 4.564); 0.157a	3.245 (1.291, 8.154); 0.012a	1.435 (0.529, 3.893); 0.478
Caregiver’s sex
Male	Reference
Female	0.585 (0.317, 1.082); 0.087	1.140 (0.554, 2.344); 0.722	0.731 (0.383, 1.396); 0.343	0.667 (0.320, 1.392); 0.281	0.474 (0.236, 0.953); 0.036a
Caregiver’s marital status
Married	Reference
Divorce	15.614 (5.693, 42.820); < 0.001a	8.979 (3.733, 21.596); < 0.001a	13.295 (5.140, 34.389) < 0.001a	3.409 (1.404, 8.278); 0.007a	11.500 (4.878, 27.111); < 0.001
Relationship with child
Mother	Reference
Father	1.792 (0.913, 3.519); 0.090a	0.905 (0.407, 2.014); 0.806	1.512 (0.749, 3.054); 0.248	1.526 (0.677, 3.442); 0.308	1.908 (0.871, 4.180); 0.107a
Others	1.160 (0.415, 3.243); 0.777	1.583 (0.598, 4.189); 0.355	0.919 (0.302, 2.795); 0.882	1.571 (0.512, 4.822); 0.430	1.201 (0.343, 4.204); 0.774
Father’s educational level
Below a bachelor’s degree	Reference
Bachelor’s degree and above	0.355 (0.197, 0.640); 0.001a	0.411 (0.224, 0.754); 0.004a	0.226 (0.126, 0.406); < 0.001	0.349 (0.181, 0.673); 0.002a	0.604 (0.288, 1.265); 0.182a
Mother’s educational level
Less than Bachelor’s	Reference
More than a Bachelor’s	0.286 (0.169, 0.484); < 0.001a	0.309 (0.181, 0.528); < 0.001a	0.179 (0.105, 0.305); < 0.001a	0.265 (0.148, 0.475); < 0.001a	0.352 (0.191, 0.647); 0.001a
Father’s occupation
Public sector	Reference
Private sector	0.658 (0.437, 0.990); 0.045a	0.670 (0.439, 1.021); 0.062a	1.029 (0.674, 1.572); 0.894	0.872 (0.531, 1.432); 0.588	0.778 (0.469, 1.289); 0.330
Mother’s occupation
Public sector	Reference
Private sector	0.972 (0.616, 1.533); 0.902	0.930 (0.583, 1.486); 0.763	1.394 (0.852, 2.281); 0.186	0.918 (0.532, 1.587); 0.760	0.838 (0.482, 1.455); 0.530
Unemployed	0.856 (0.092, 7.924); 0.891	0.927 (0.100, 8.598); 0.947	1.210 (0.130, 11.289); 0.867	1.488 (0.158, 13.972); 0.728	1.488 (0.158, 13.972); 0.728
Family monthly income (per month)
$617-$1544	Reference
> $1544	0.189 (0.107, 0.335); < 0.001a	0.212 (0.120, 0.375); < 0.001a	0.132 (0.074, 0.237); < 0.001	0.284 (0.152, 0.530); < 0.001a	0.290 (0.154, 0.547); < 0.001a
Caregiver’s physical health problems	0.744 (0.210, 2.631); 0.646	0.833 (0.235, 2.952); 0.777	0.805 (0.227, 2.849); 0.736	1.364 (0.382, 4.866); 0.632	0.411 (0.054, 3.150); 0.392
Caregiver’s mental health problems	-2	10.395 (4.583, 23.578); < 0.001a	22.760 (8.455, 61.267); < 0.001a	4.346 (1.962, 9.624); < 0.001a	4.682 (2.109, 10.395); < 0.001a
PCEs
Mentorship	0.072 (0.043, 0.121); < 0.001a	0.216 (0.133, 0.352); < 0.001a	0.101 (0.061, 0.166); < 0.001a	0.217 (0.127, 0.371); < 0.001a	0.076 (0.043, 0.133); < 0.001a
Family resilience	0.107 (0.065, 0.175); < 0.001a	0.237 (0.146, 0.383); < 0.001a	0.120 (0.074, 0.196); < 0.001a	0.233 (0.137, 0.398); < 0.001a	0.155 (0.090, 0.266); < 0.001a
Safe neighborhood	0.504 (0.321, 0.792); 0.003	0.385 (0.244, 0.607); < 0.001a	0.476 (0.301, 0.753); 0.001a	0.548 (0.321, 0.936); 0.028a	0.744 (0.420, 1.316); 0.310
Supportive neighborhood	0.358 (0.223, 0.576); < 0.001a	0.470 (0.294, 0.751); 0.002a	0.423 (0.263, 0.678); < 0.001a	0.729 (0.433, 1.227); 0.234	0.694 (0.405, 1.187); 0.182a
Afterschool activity	0.623 (0.413, 0.940); 0.024a	0.613 (0.401, 0.937); 0.024a	0.715 (0.469, 1.089); 0.118a	0.573 (0.350, 0.939); 0.027a	0.474 (0.285, 0.786); 0.004a
Volunteerism	0.490 (0.298, 0.806); < 0.005a	0.726 (0.450, 1.174); 0.192	0.441 (0.261, 0.744); 0.002a	0.862 (0.498, 1.495); 0.598	1.021 (0.590, 1.766); 0.942
Share ideas with the caregiver	0.059 (0.030, 0.116); < 0.001a	0.126 (0.070, 0.227); < 0.001a	0.041 (0.020, 0.084); < 0.001a	0.135 (0.073, 0.248); < 0.001a	0.149 (0.081, 0.276); < 0.001a
Number of PCEs ≥ 4	0.048 (0.028, 0.081); < 0.001a	0.116 (0.071, 0.187); < 0.001a	0.057 (0.034, 0.096); < 0.001a	0.249 (0.147, 0.422); < 0.001a	0.154 (0.090, 0.264); < 0.001a

^aP < 0.2, potential significant factor, which will be included in multivariable logistic regression models.

¹Data are presented as the odds ratio (95% confidence interval).

²Without emotional problems, caregivers with mental health problems. This represents complete separation in the data. Because of this, logistic regression could not produce a valid odds ratio due to infinite estimates.

GPAX: Cumulative grade-point average; PCEs: Positive childhood experiences.

In multivariable logistic regression analysis (Table 3), GPAX > 3.00 remained protective against risk or problematic emotional problems (aOR = 0.441, P = 0.027), whereas caregiver age > 50 years was associated with markedly higher odds (aOR = 17.098, P = 0.010). Mentorship was protective for risk or problematic emotional (aOR = 0.140, P < 0.001), hyperactive (aOR = 0.350, P = 0.009), and prosocial problems (aOR = 0.091, P < 0.001). Family resilience and opportunities to share ideas with caregivers were also significantly protective in multiple domains. Adolescents with ≥ 4 PCEs had lower odds of risk or problematic emotional (aOR = 0.256, P = 0.023) and hyperactive problems (aOR = 0.324, P = 0.046).

Table 3 Multivariable logistic regression results for the association between demographic characteristics and risk or problematic behavioral and emotional problems¹.

Factors	Risk or problematic behavioral and emotional problems
	Emotional problem	Conduct problem	Hyperactive problem	Peer problem	Prosocial behavior
Child’s factor
Sex
Male	-	Reference		-	-
Female	-	0.704 (0.427, 1.160); 0.168	0.884 (0.485, 1.613); 0.688	-	-
School type
Public school	-	-	Reference	-	-
Private school	-	-	1.286 (0.650, 2.541); 0.470	-	-
GPAX
2.01-3.00	Reference				-
> 3.00	0.441 (0.213, 0.913); 0.027a	0.993 (0.543, 1.818); 0.983	0.792 (0.390, 1.609); 0.519	0.862 (0.447, 1.663); 0.658	-
Caregiver and family factors
Caregiver’s age in years
31-40	Reference
41-50	0.956 (0.469, 1.949); 0.902	1.018 (0.581, 1.785); 0.905	1.621 (0.808, 3.251); 0.174	1.223 (0.635, 2.354); 0.548	0.542 (0.268, 1.093); 0.087
> 50	17.098 (1.971, 148.291); 0.010a	1.900 (0.629, 5.744); 0.255	0.702 (0.149, 3.309); 0.655	2.842 (0.957, 8.438); 0.060	1.876 (0.343, 10.264); 0.468
Caregiver’s sex
Male	Reference	-	-		-
Female	3.154 (0.150, 66.287); 0.460	-	-		-
Caregiver’s marital status
Married	Reference				-
Divorce	2.906 (0.363, 23.258); 0.315	2.599 (0.847, 7.971); 0.095	1.871 (0.394, 8.875); 0.430	1.079 (0.338, 3.444); 0.898	-
Relationship with child
Mother	Reference	-	-	-	-
Father	2.657 (0.112, 63.297); 0.546	-	-	-	-
Others	0.303 (0.031, 2.935); 0.303	-	-	-	-
Father’s educational level
Less than Bachelor’s	Reference
More than a Bachelor’s	1.219 (0.219, 6.796); 0.822	1.613 (0.483, 5.384); 0.437	1.003 (0.283, 3.556); 0.997	1.197 (0.378, 3.795); 0.760	0.604 (0.348, 1.278); 0.124
Mother’s educational level
Less than Bachelor’s	Reference
More than a Bachelor’s	1.744 (0.288, 10.547); 0.544	1.156 (0.341, 3.916); 0.816	0.379 (0.106, 1.354); 0.135	0.351 (0.110, 1.122); 0.77	0.354 (0.089, 1.417); 0.142
Father’s occupation
Public sector	Reference		-	-	-
Private sector	0.567 (0.292, 1.099); 0.093	0.634 (0.374, 1.075); 0.091	-	-	-
Mother’s occupation
Public sector	-	-	Reference	-	-
Private sector	-	-	0.968 (0.473, 1.984); 0.930	-	-
Unemployed	-	-	0.178 (0.005, 6.711); 0.351	-	-
Family monthly income
$617-$1544	Reference
> $1544	0.536 (0.096, 2.985); 0.476	0.449 (0.136, 1.484); 0.189	0.450 (0.120, 1.685); 0.236	1.744 (0.487, 6.241); 0.392	1.238 (0.290, 5.280); 0.773
Caregiver’s mental health problems	-	2.319 (0.820, 6.553); 0.113	6.077 (1.394, 26.490); 0.016a	1.468 (0.510, 4.228); 0.477	0.768 (0.232, 2.543); 0.666
PCEs
Mentorship	0.140 (0.062, 0.316); < 0.001a	0.646 (0.318, 1.315); 0.228	0.350 (0.158, 0.772); 0.009a	0.359 (0.172, 0.749); 0.006a	0.091 (0.042, 0.198); < 0.001a
Family resilience	0.390 (0.166, 0.916); 0.031a	0.789 (0.380, 1.639); 0.525	0.375 (0.168, 0.836); 0.016a	0.322 (0.154, 0.672); 0.003a	0.278 (0.122, 0.633); 0.002a
Safe neighborhood	1.839 (0.693, 4.882); 0.221	0.657 (0.339, 1.274); 0.214	1.410 (0.571, 3.480); 0.456	0.645 (0.310, 1.344); 0.242	-
Supportive neighborhood	0.460 (0.218, 0.972); 0.042a	0.710 (0.395, 1.275); 0.251	0.591 (0.287, 1.218); 0.154	-	0.749 (0.355, 1.582); 0.449
Afterschool activity	0.790 (0.386, 1.618); 0.520	0.853 (0.486, 1.496); 0.578	1.025 (0.502, 2.094); 0.945	0.489 (0.267, 0.897); 0.021a	0.317 (0.157, 0.642); 0.001a
Volunteerism	1.010 (0.471, 2.163); 0.980	1.354 (0.743, 2.467); 0.322	0.738 (0.345, 1.581); 0.434	-	-
Share ideas with the caregiver	0.132 (0.050, 0.351); < 0.01a	0.226 (0.105, 0.488); < 0.01a	0.066 (0.026, 0.167); < 0.001a	0.217 (0.101, 0.468); < 0.001a	0.294 (0.123, 0.707); 0.006a
Number of PCEs ≥ 4	0.256 (0.079, 0.830); 0.023a	0.400 (0.158, 1.015); 0.054	0.324 (0.107, 0.982); 0.046a	2.039 (0.816, 5.092); 0.127	2.165 (0.794, 5.898); 0.131

^aP < 0.05, factors that were statistically significant; data were not included due to P > 0.2 in the univariate logistic regression analysis.

¹Data are presented as the odds ratio (95% confidence interval).

GPAX: Cumulative grade-point average; PCEs: Positive childhood experiences.

Discriminative ability of PCEs

ROC analysis demonstrated that a cutoff of ≥ 4 PCEs had acceptable discriminative performance for distinguishing adolescents with and without SDQ risk/problematic behavioral outcomes across domains. AUC values were 0.835 for emotional problems, 0.818 for conduct problems, 0.743 for hyperactivity, 0.708 for peer problems, and 0.754 for low prosocial behavior (all P < 0.001). Sensitivity ranged from 84.8% to 93.1%, and specificity from 41.9% to 60.8% (Figure 2).

Figure 2 Receiver operating characteristic analyses of positive childhood experiences and child behavioral-emotional outcomes. A: Emotional problems: Area under the curve (AUC) = 0.835 [95% confidence interval (CI): 0.790-0.881]; B: Conduct problems: AUC = 0.818 (95%CI: 0.767-0.868); C: Hyperactive behavior: AUC = 0.743 (95%CI: 0.685-0.801); D: Peer problems: AUC = 0.708 (95%CI: 0.640-0.776); E: Prosocial behavior: AUC = 0.754 (95%CI: 0.685-0.823). All P < 0.001. The optimal cutoff for identifying children at lower risk was ≥ 4 positive childhood experiences, yielding high sensitivity (84.8%-93.1%) and moderate specificity (41.9%-60.8%) across all domains. Collectively, these findings highlight ≥ 4 positive childhood experiences as a consistent protective threshold associated with fewer emotional and behavioral difficulties and better prosocial functioning. ROC: Receiver operating characteristic.

DISCUSSION

Regarding the prior literature, similarities can be noted between American adults in the prior study by Sege et al[11] and Thai adolescents in the present study, particularly in the distribution of PCEs. In both contexts, protective relationships and supportive family emerged as the most prevalent PCEs, whereas engagement in community traditions and volunteerism were the least common. However, in the Thai sample, positive relational experiences were substantially more frequent than indicators of community connectedness. This distribution underscores the role of cultural context in shaping PCEs. Specifically, the high prevalence of family-based PCEs such as resilience, caregiver communication, and shared problem-solving may reflect collectivist cultural norms in Thailand, which emphasize intergenerational support, parental authority, and family cohesion. This pattern is consistent with antecedent studies from Asian societies. For example, research in China and South Korea demonstrates that family is often the primary source of psychosocial support, with adolescents reporting high levels of parental involvement but comparatively limited exposure to structured mentorship or community-based protective factors[12,13]. This disparity between family and community engagement may be partially explained by the tendency of Asian youth to prioritize academic achievement and family obligations[14], which leaves fewer opportunities for participation in volunteerism or civic traditions that would foster community-level PCEs.

By contrast, studies from Western contexts emphasize the salience of community-based PCEs. Blum et al[7] and Crouch et al[4] found that structured community programs and accessible nonfamily mentorship opportunities significantly contribute to resilience and overall well-being among American and European adolescents. In our cohort, mentorship was common, whereas other indicators of community connectedness (e.g., volunteerism and supportive neighborhoods) were less prevalent, suggesting that community-level PCEs may be unevenly distributed in this Thai urban sample. This cross-cultural difference in PCE distribution likely reflects underlying cultural orientations: Collectivist cultures, such as Thailand, often prioritize vertical family ties over horizontal community linkages, which may explain the lower prevalence of community-level PCEs in the present Thai cohort.

While strong family-level PCEs in Thailand may effectively buffer children against adversity, they may not fully substitute for the broader developmental benefits of peer support and civic participation. Western societies, with more institutionalized community engagement, appear to provide a more balanced PCE profile that combines family and community resources. This raises important implications for intervention: Although family-based PCEs in Thailand remain robust, expanding opportunities for volunteering and broader community participation and strengthening school-community partnerships may help address existing gaps and promote more comprehensive protective experiences.

Given the identified gap in community-level PCEs among Thai adolescents, it is important to highlight protective factors beyond the family that can extend similar benefits. In this study, mentorship primarily reflects supportive nonfamily adults (e.g., teachers, coaches, relatives, or other trusted adults) who provide guidance and encouragement and facilitate access to supportive routines and resources; caregivers may also play a mentoring role. Consistent with this interpretation, our findings show that mentorship is associated with lower risks of emotional difficulties, hyperactivity, and prosocial challenges, reflecting the broader role of caring adults in children’s lives[4,5]. These mentoring relationships provide additional coping strategies and emotional support that help children manage stress and navigate developmental challenges more effectively, particularly in settings where access to structured community programs is limited. Taken together, the high prevalence of mentorship alongside lower volunteerism and supportive-neighborhood indicators suggests that, in this urban Thai sample, mentorship may be expressed mainly through school-linked and other informal nonfamily relationships, rather than through broader civic participation, aligning with the observed imbalance between relational PCEs and community connectedness.

Family resilience and caregiver-child communication emerged as other strong protective factors, reinforcing the centrality of family processes in shaping children’s psychological adjustment. As Anderson et al[9] note, open communication fosters emotional security by creating a safe space in which children feel heard, validated, and supported. This communicative openness facilitates collaborative problem-solving and strengthens belonging and relational trust - features consistently associated with lower behavioral and emotional difficulties. In parallel, resilience within the family system provides children with models of adaptive coping and constructive responses to adversity; families that normalize challenges and frame them as manageable can mitigate stress exposure and promote healthier emotion regulation and confidence[15].

The opportunity for children to share ideas with caregivers represents an additional protective factor that strengthens family functioning. Such exchanges foster open dialogue, mutual respect, and the validation of children’s perspectives, which together enhance emotional security and a sense of belonging. When caregivers actively listen and engage with their children’s ideas, they encourage autonomy and collaborative problem-solving, allowing children to feel both supported and empowered. This dynamic is associated with a lower likelihood of emotional and behavioral difficulties and with skills for constructive communication and adaptive coping beyond the family context. Moreover, the act of sharing ideas reinforces trust within the caregiver-child relationship, creating a foundation for resilience that enables children to navigate stressors with greater confidence and regulation. Consistent with this, a recent systematic review found that higher-quality parent-child communication was negatively associated with a broad range of mental health difficulties, underscoring its role as a robust protective factor[16].

Our findings reinforce and extend the growing literature on the protective role of PCEs. While much attention has historically focused on the detrimental effects of ACEs on mental health, recent research emphasizes that PCEs not only mitigate the negative impact of ACEs but also actively promote resilience and well-being. Bethell et al[6] reported that PCEs were strongly associated with better adult mental and relational health - even among individuals exposed to multiple ACEs - supporting a cumulative resilience model. In line with our observation of relatively low percentages of behavioral and emotional problems (SDQ), the recent study of Logan and Lewis-de Los Angeles in 2025[17] found that greater PCEs weakened the association between ACEs and problems measured by the Childhood Behavioral Checklist. Crouch et al[4] and Choi et al[5] demonstrated that mentorship, family cohesion, and supportive caregiver-child interactions were associated with lower internalizing and externalizing symptoms in pre-adolescents. Consistent with this body of evidence, our data show that having ≥ 4 PCEs was significantly associated with lower odds of emotional and hyperactive problems in Thai adolescents, providing cross-cultural support for the cumulative resilience framework. This threshold pattern may reflect an additive (dose-response) effect, whereby multiple co-occurring supportive experiences collectively enhance emotion regulation and coping and buffer stress-related behavioral difficulties. Although ROC analyses suggested acceptable discrimination across domains, adjusted associations for some domains may be less stable due to low event rates (e.g., no “problem” level in peer and prosocial domains). Notably, the distribution of cumulative PCEs also differed: Whereas only 27.6% of Thai participants reported 6-7 PCEs, Sege et al[11] found that 53.1% of American adults endorsed this higher range, suggesting broader access to community-level supports in Western contexts and a correspondingly higher overall PCE profile.

Another factor affecting behavioral problems in adolescents that warrants consideration is socioeconomic position (SEP). Prior work shows that lower childhood SEP is associated with a greater risk of ACEs/maltreatment[18]. In our study, however, the strong protective effects of family-level PCEs suggest that even in contexts of higher SEP, relational and emotional supports remain critical determinants of child mental health. The associations we observed - higher GPAX, parental education, and income relating to fewer behavioral and emotional problems - are consistent with socioeconomic models of child development that emphasize material resources, enriched environments, and parental involvement, as noted by Blum et al[7]. Notably, children of caregivers older than 50 years had a significantly higher risk of emotional problems (aOR = 17.098), although the estimate was imprecise, as reflected by a wide CI. This pattern is less frequently documented in the literature. Possible explanations include generational gaps in parenting style, reduced energy or availability for active engagement, and differences in attitudes toward emotional expression; these mechanisms merit further investigation in future studies.

In univariate analysis, female sex was associated with lower odds of conduct problems (OR = 0.653, P = 0.047), consistent with sex differences commonly observed in externalizing behaviors[10]. However, sex was not statistically significant in the multivariable model; therefore, this association should be interpreted cautiously.

In summary, these findings corroborate and extend prior research by demonstrating that PCEs are associated with reduced behavioral and emotional difficulties in a non-Western, Southeast Asian context. They also underscore the importance of interpreting the prevalence and impact of PCEs through both cultural norms and socioeconomic conditions, situating our results within a broader cross-cultural and socioeconomic framework without presupposing specific implementation strategies at this stage.

This study has several strengths, including its large sample size (n = 540), the use of validated instruments such as the Thai SDQ[10], and rigorous statistical analyses, including ROC modeling to evaluate discriminative thresholds. The inclusion of multiple domains of PCEs also enabled a comprehensive assessment of both family- and community-level protective factors.

Despite these strengths, several limitations warrant consideration. First, the cross-sectional design limits causal inference; while PCEs are strongly associated with lower behavioral problems, reverse causality cannot be ruled out (e.g., children with fewer behavioral issues may elicit more positive experiences). Moreover, ROC/AUC results reflect cross-sectional discriminative performance and should not be interpreted as predictive or causal. Second, we observed evidence of complete separation between caregiver mental health and adolescents’ emotional problems, which can produce non-convergence in logistic regression when every caregiver with mental health problems has an adolescent with emotional issues. In such circumstances, estimates of the OR are not reliable; future work should consider penalized likelihood approaches, Firth’s logistic regression, or exact logistic regression, and larger samples to improve stability and generalizability.

With respect to generalizability, the study was conducted exclusively in urban Bangkok and relied on convenience sampling from participating schools. Therefore, prevalence estimates may not be nationally representative due to potential sampling bias. In addition, restricting the sample to a school-based non-clinical population by excluding adolescents with prior clinician-diagnosed neurodevelopmental or psychiatric disorders may underestimate the prevalence of behavioral-emotional problems and limit generalizability to these groups. Crouch et al[19] reported higher community-based PCEs in rural settings, potentially reflecting greater volunteering and supportive neighborhoods; however, our findings may not generalize to rural areas or other regions of Thailand, and this should be examined in future studies. Additionally, the high socioeconomic profile of the sample (89.4% reporting household income > 1544 United States dollars) may limit applicability to more disadvantaged groups, for whom access to and the impacts of PCEs could differ. Finally, reliance on self-report introduces risks of recall and social desirability biases, particularly for sensitive constructs (e.g., family resilience, caregiver mental health). Future studies could triangulate reports across multiple informants (e.g., teachers, peers) and employ longitudinal designs to clarify temporal relationships.

Practical implications

The findings carry several practical implications. At the family level, interventions that promote caregiver-child communication and enhance family resilience could reduce risks for behavioral and emotional difficulties. At the community level, programs that foster mentorship opportunities, such as school-based mentoring or structured extracurricular activities, may be beneficial, especially for children from vulnerable households. Expanding opportunities for volunteerism and neighborhood cohesion may also strengthen protective factors that appear underrepresented in Thai children’s experiences. From a policy perspective, embedding PCE promotion within national child health and education strategies could contribute to broader improvements in population mental health. For example, integrating PCE-oriented questions into pediatric visits or school health assessments could help identify children who may be at higher or lower risk. Given the discriminative pattern observed at ≥ 4 PCEs in this cross-sectional sample, this threshold may help flag lower-risk profiles; however, external validation and prospective evaluation of predictive performance are required before any clinical screening use.

Future research directions

Future studies should adopt longitudinal designs to examine how PCEs are associated with, and may shape, developmental trajectories into adolescence and adulthood, and to test whether cumulative thresholds remain consistently associated with outcomes across life stages. Investigating interactions between PCEs and ACEs in Thai populations would clarify whether PCEs attenuate the adverse effects of trauma, as reported in Western cohorts[6]. Cross-cultural and multi-site research is warranted to evaluate how integrating family and community PCEs maximizes protective effects on child and adolescent well-being across diverse contexts. Additionally, qualitative studies that explore children’s and caregivers’ perspectives, including implementation barriers and facilitators, could provide deeper insight into cultural meanings and real-world applications of PCEs in Thailand.

CONCLUSION

This study shows that PCEs are common among Thai children and are robustly associated with fewer behavioral and emotional difficulties. Mentorship, sharing ideas with caregivers, and family resilience were especially influential, and a cumulative threshold of ≥ 4 PCEs showed acceptable-to-good discrimination for current behavioral risk, underscoring the value of strengthening family- and community-based protective experiences.

ACKNOWLEDGEMENTS

The authors would like to thank Ms. Julaporn Pooliam of the Office for Research and Development for her assistance with the statistical analyses.