INTRODUCTION
Posttraumatic stress disorder (PTSD) is a complex and debilitating mental health condition that frequently arises following significant traumatic events. Particularly, traumatic physical injuries from accidents, acts of violence, or other distressing events are well-documented triggers for PTSD, often leading to persistent functional impairments and reduced quality of life [1]. Identifying predictors for the onset and progression of PTSD is crucial for the disorder’s prevention and management.
Age is frequently examined as a potential factor influencing PTSD development; however, findings regarding its role remain inconsistent, highlighting the complexity of this relationship. Younger individuals may exhibit heightened vulnerability to PTSD due to factors such as less developed coping strategies, greater emotional sensitivity, and trauma’s pronounced impact on the developing brain [2]. Conversely, older adults could exhibit greater resilience and more favorable recovery trajectories, potentially related to accumulated life experiences, emotional maturity, and stronger social support networks [3]. Nevertheless, these age-related patterns are not universal, as multiple confounding factors, including cumulative trauma exposure, previous psychiatric disorders, physical health conditions, and the type and severity of trauma experienced, substantially influence PTSD risk and recovery across all age groups [4,5]. Indeed, a systematic review of PTSD predictors among road traffic accident survivors underscores these complexities, reporting inconsistent findings on the role of age and calling for more comprehensive investigations [6].
Alcohol consumption is another important factor intricately linked with both age and PTSD onset. Older adults generally consume alcohol less frequently, possibly due to increased physiological sensitivity to alcohol that accompanies aging, leading to quicker intoxication and prolonged recovery periods [7]. Such age-related metabolic changes and shifting social behaviors highlight the importance of understanding drinking patterns across different age groups.
Moreover, alcohol consumption itself is a recognized factor influencing PTSD risk. Chronic alcohol use, defined quantitatively as consuming an average of five or more drinks per day on 5 or more days within a month, increases the likelihood of experiencing traumatic events and may initiate a detrimental cycle of PTSD symptoms, dependence, and worsening mental health outcomes [8,9].
Considering these interrelations, we hypothesize that alcohol drinking status negatively modulates the association between age and PTSD development, predicting a higher incidence of PTSD among younger adults who consume alcohol compared to older or non-drinking individuals. We test this hypothesis prospectively in a 2-year cohort of Korean patients with physical injuries, carefully controlling for critical confounders such as previous psychiatric disorders, cumulative trauma exposure, and psychosocial variables. By clarifying whether observed associations can be directly attributed to age and alcohol use, this study addresses existing gaps in the literature and may inform targeted interventions for PTSD prevention and management.
METHODS
Study outline
This investigation forms a specific segment of the broader Biomarker-based Diagnostic Algorithm for Post-Traumatic Syndrome (BioPTS) study, which seeks to refine diagnostic models for PTSD. We focused our analysis on a subset of participants drawn from the larger BioPTS cohort. The comprehensive details of the overall study design have been elaborately outlined in a previously published design paper [10]. Participants included in this analysis were individuals who had sustained physical injuries and were enrolled in BioPTS from June 2015 to January 2021 at the Trauma Center of Chonnam National University Hospital (CNHU) in Gwangju, South Korea. Initial assessments were conducted within 1 month of the participants’ hospital admission. These evaluations were extensive and involved gathering socio-demographic data, analyzing clinical profiles, and conducting detailed assessments of each participant’s alcohol consumption status. Additionally, the assessments included measures known to be associated with the onset of PTSD [6], providing critical insights into the complex interactions between psychological resilience, social support, injury severity, and mental health outcomes. The longitudinal aspect of the study is highlighted by the successive PTSD assessments conducted via structured telephone interviews at 3, 6, 12, and 24 months following the injury. The timeline of questionnaires administered throughout the study is visualized in Supplementary Figure 1. The study protocol, including its ethical considerations and participant consent processes, received approval from the CNHU Institutional Review Board (IRB Approval No. CNUH 2015-148).
Participants
Inclusion criteria were: 1) individuals aged 18 years or older at the index injury, 2) patients hospitalised for more than 24 hours after sustaining a moderate to severe physical injury measured by the Injury Severity Score (ISS) ≥9 [11], and 3) individuals sufficiently proficient in the Korean language to comprehend the study protocol. Exclusion criteria were: 1) moderate or severe brain injury measured by the Glasgow Coma Scale (GCS) <10 [12]; 2) physical injuries resulting from suicide attempts; 3) conditions hindering comprehensive psychiatric evaluation due to severe physical ailments; 4) prior history of psychiatric disorders including psychotic disorder, bipolar disorder, or alcohol or substance use disorders other than depressive and anxiety disorders; 5) significant cognitive impairments due to organic mental or neurocognitive disorders; and 6) pre-existing convulsive disorders or a history of anticonvulsant use. All participants were fully informed about the study’s purpose, procedures, potential risks, and benefits before providing written informed consent.
Baseline evaluations
Alcohol drinking status
Participants were classified based on their alcohol consumption at baseline into two groups: non-current drinkers, which included both never drinkers and ex-drinkers (who previously drank but had not consumed alcohol in the past 12 months), and current drinkers. An assessment of alcohol use, drinking patterns, and related issues during the preceding 12 months was conducted using the Alcohol Use Disorders Identification Test (AUDIT) scale [13]. Participants completed the AUDIT shortly after hospital admission, recognizing that recent hospitalization could potentially influence reported consumption patterns. Participants were further divided based on their AUDIT scores into those with scores below 8 (indicative of low-risk drinking) and those with scores of 8 or higher (suggestive of hazardous drinking), according to World Health Organization (WHO) guidelines [14].
Age group
The age of participants at baseline enrollment was recorded and initially categorized into two groups: younger (<60 years) and older (≥60 years), following the precedent set by previous research protocols [15]. This categorization was strategically chosen to balance the number of participants in each group, maximizing statistical power for comparative analyses. Additionally, to address concerns regarding the appropriateness of this dichotomous age categorization and to provide a more comprehensive analysis, age was also treated as a continuous variable in further analyses.
Socio-demographic characteristics
Collected baseline socio-demographic data included participants’ sex, duration of education, cohabitation status (whether living alone or with others), and current employment status.
Pre-trauma characteristics
Documented histories of psychiatric disorders included depressive disorders, panic disorder, agoraphobia, social phobia, and generalized anxiety disorder. The experiences of participants with previous lifetime traumatic events and childhood abuse (emotional/psychological, physical, and sexual abuse before age 16 were assessed using the Nemesis Childhood Trauma Interview [16]. For analysis, a broad definition of “childhood abuse” (experiencing at least one type of abuse) was employed. Resilience and social support were measured using the Connor-Davidson Resilience Scale (CDRS) [17] and the Multidimensional Scale of Perceived Social Support (MSPSS) [18], respectively, with lower scores indicating higher symptomatology.
Trauma related characteristics
The severity of the injuries sustained by participants was quantified using the ISS, as outlined in the eligibility criteria. Higher ISS scores indicate more severe trauma.
Peri-trauma characteristics
During the peri-trauma period, depressive and anxiety symptoms were evaluated using the Hospital Anxiety and Depression Scale (HADS), Anxiety subscale (HADS-A), and Depression subscale (HADS-D) [19].
Follow-up diagnosis of PTSD
To diagnose PTSD, this study utilized the Clinician-Administered PTSD Scale for DSM-5 (CAPS-5) [20]. CAPS-5 is widely recognized for its high reliability and validity, providing a comprehensive and robust framework for PTSD evaluation, even when administered via telephone [21]. For a diagnosis to be established, participants had to meet specific criteria across various symptom clusters: at least one symptom from Cluster B (intrusion symptoms), one from Cluster C (avoidance), two from Cluster D (negative alterations in cognitions and mood), and one from Cluster E (alterations in arousal and reactivity). Additionally, the diagnosis required fulfilling the duration and functional significance criteria outlined in Clusters F (duration of symptoms) and G (functional significance). The study’s outcome variables were defined as the occurrence of any PTSD diagnosis at least once during the follow-up period and the presence of PTSD at each respective follow-up interval.
Statistical analysis
This analysis incorporated participants who completed at least one follow-up evaluation subsequent to the initial baseline assessment. This approach is consistent with DSM-5 guidelines [22], which recognize PTSD diagnosis as valid one-month post-trauma. Baseline data were stratified based on the development of PTSD during the follow-up period (absent vs. present), alcohol drinking status (current non-drinker vs. current drinker), and age groups (≥60 vs. <60 years). Continuous variables were compared using t-tests if normally distributed, and non-parametric tests were employed for those not meeting this criterion; categorical variables were assessed through χ2 tests. Covariates for further analyses were chosen based on their significant associations in these initial comparisons (p<0.05) and an assessment of collinearity among variables. To illustrate the relationships between the study variables, a correlation matrix using Spearman’s rho tests was generated and included in the analysis, providing insight into the strength and direction of associations among the variables. The individual associations of age groups (≥60 vs. <60 years), alcohol drinking status (current non-drinker vs. current drinker), and hazardous alcohol consumption (AUDIT score <8 vs. ≥8) with PTSD development were first explored using binary logistic regression models. These models were applied both before and after adjustment for the selected covariates. To investigate the potential modifying effects of alcohol drinking status on the relationship between age groups and PTSD occurrence, multinomial logistic regression was employed, adjusting for the predetermined covariates. This analysis was repeated for PTSD occurrence at each follow-up interval to identify any modified association dependent on alcohol drinking status. Additionally, multinomial logistic regression analyses were also conducted with age treated as a continuous variable to provide a comprehensive evaluation of its impact on PTSD development. All statistical tests were two-sided, with the significance threshold set at 0.05. The analyses were performed using SPSS software, version 21.0 (IBM Corp.).
RESULTS
Recruitment and baseline characteristics
The recruitment process from the initial assessment to the 24-month follow-up and the prevalence of PTSD are illustrated in Figure 1. Of the 1,142 patients who met the eligibility criteria and consented to participate, 95 (8.3%) discontinued before the 3-month evaluation, resulting in 1,047 patients (91.7%) being included in the final analysis. Statistical comparisons of baseline characteristics between those who continued and those who discontinued showed no significant differences (all p-values>0.1). Within the analyzed cohort, 122 patients (11.7%) were diagnosed with PTSD at some point during 24 months. Detailed comparisons of baseline characteristics between patients with and without PTSD are presented in Table 1. PTSD was significantly associated with several factors, including education, cohabitation status, previous psychiatric history, previous traumatic experiences, childhood abuse, and levels of anxiety and depression. Further analyses were conducted by current drinking status and age group, as presented in Supplementary Tables 1 and 2. Participants identified as current drinkers reported consuming alcohol on an average of 3.7 (SD=1.0) days per week. Current drinking was significantly correlated with lower age, male sex, higher education, employment, childhood abuse, higher resilience, higher social support, greater injury severity, and lower levels of anxiety and depression. Similarly, lower age was significantly associated with current alcohol consumption, male sex, higher education, unmarried status, employment, childhood abuse, higher resilience, higher social support, greater injury severity, and lower depression. Considering these associations and the potential for collinearity, eight variables were selected as covariates for subsequent analyses: sex, unemployment status, previous psychiatric disorders, previous traumatic events, any childhood abuse, and scores on the CDRS, MSPSS, ISS, and HADS-D. The correlation matrix among the baseline variables is presented in Supplementary Table 3.
Individual associations of age group and alcohol drinking status with PTSD diagnosis at any point
The associations between age group and alcohol drinking status with PTSD diagnosis at any point were explored through binary logistic regression analyses, with the results summarized in Table 1. Being in the younger age group was significantly associated with the occurrence of PTSD, both before and after adjusting for covariates. Conversely, no significant associations were observed with either current or hazardous alcohol consumption before or after covariate adjustment.
Modifying effects of alcohol drinking status on age group-PTSD associations
The potential modifying effects of alcohol drinking status on the relationship between age group and PTSD diagnosis were examined using multinomial logistic regression after adjustment. The outcomes of these analyses are visually depicted in Figures 2, 3, and Supplementary Figure 2. For current and hazardous alcohol drinkers, the associations between age group and PTSD were not significant. However, for current low-risk drinkers, younger age was significantly associated with PTSD diagnosis at any point, evidenced by significant interaction terms (Figure 2). The prevalence of PTSD was 8.2% at the 3-month, 7.2% at the 6-month, 4.7% at the 12-month, and 3.8% at the 24-month evaluation points. The relationship between age group and PTSD at each follow-up interval largely mirrored the overall findings. However, the modifying influence of current alcohol consumption was specifically apparent for PTSD diagnoses at 12 and 24 months, as indicated by significant interaction terms, but this effect was not observed at the 3- and 6-month evaluations (Figure 3). Similarly, hazardous alcohol consumption demonstrated modifying effects for PTSD at 6, 12, and 24 months, with significant interaction terms, but not at the 3-month mark (Supplementary Figure 2). When age was analyzed as a continuous variable, similar modifying effects were observed. The interaction between current alcohol consumption and age was significant for PTSD over the 2-year period and at the 12- and 24-month follow-up points. Additionally, the interaction between hazardous alcohol drinking and age showed significant effects for PTSD over the 2-year period and at the 6-, 12-, and 24-month intervals (Supplementary Table 4).
DISCUSSION
This 2-year longitudinal study of patients with physical injuries has revealed distinct alcohol drinking-dependent associations between age and PTSD. The significant association between younger age (<60 years) and subsequent PTSD development, observed in all participants, diminished in current and hazardous alcohol drinkers but was more pronounced in current low-risk drinkers, as indicated by significant interaction terms. This pattern remained consistent in later followup points (from 6 to 24 months) but not at the early 3-month evaluation.
These novel findings may contribute to the understanding of PTSD etiology. The study’s credibility is reinforced by the replication of associations previously reported in PTSD research. Particularly, this includes the well-established links between younger age and PTSD [2,3] and between younger age and alcohol drinking status [7]. Although alcohol drinking status alone was not significantly associated with PTSD in this study, it modified the well-known association between age and PTSD, suggesting a complex relationship between alcohol consumption and PTSD as reported in prior research [8]. This divergence from expected patterns highlights the unique contributions of our study and its significance in the field.
Several explanations are plausible for the observed alcohol drinking-dependent associations. Firstly, alcohol may affect the brain’s stress response system, and chronic use could lead to altered stress processing in individuals of all ages [23]. If alcohol similarly impacts younger and older adults, it might equalize susceptibility to PTSD among drinkers, regardless of age. Secondly, alcohol’s influence on memory formation and emotional regulation could lead to a more consistent development and manifestation of PTSD symptoms among drinkers across different age groups [24]. Thirdly, individuals might use alcohol to self-medicate or cope with PTSD symptoms [25]. If alcohol masks or temporarily alleviates these symptoms, the apparent incidence of PTSD might seem similar across age groups among drinkers. However, this doesn’t necessarily imply a lesser severity of underlying PTSD but rather that alcohol use might obscure or delay its recognition and diagnosis.
The significant modifying effects of alcohol drinking observed only for long-term PTSD (12-month and longer for baseline current alcohol drinking and 6-month and longer for hazardous alcohol drinking) can be interpreted through the hypothesis of alcohol’s cumulative effects. Participants who consumed alcohol at baseline likely continued during the study period, contrasting with current low-risk drinkers. The physiological and psychological effects of alcohol accumulate over time [26], potentially explaining the increasing divergence in interaction terms during long-term follow-up. In particular, the alcohol-dependent modifying effects appeared from the 6-month follow-up for hazardous alcohol drinking and from the 12-month follow-up for current alcohol drinking, further supporting the notion of cumulative effects. However, as the study did not monitor participants’ alcohol consumption throughout, further research is needed to substantiate these findings.
One of the primary constraints of this study is its exclusive focus on individuals who have sustained physical injuries. While this population is highly relevant to PTSD research—given that traumatic physical injuries are well-known triggers for PTSD, often leading to significant impairments in functioning and quality of life [1]—the applicability of our findings may not extend universally to populations that have experienced different forms of trauma, such as psychological or emotional traumas. Additionally, the recruitment of participants solely from a single trauma center, while ensuring consistency in evaluation and follow-up, may also introduce limitations regarding the broader applicability of the results. The study did not systematically collect information on the medication usage of participants. This omission represents a limitation, as medications can significantly influence PTSD symptoms and may interact with factors such as alcohol use and age. The absence of this data restricts our ability to fully account for these influences, potentially impacting the interpretation of the relationships explored in this study [27]. Furthermore, while baseline assessments revealed specific alcohol drinking statuses, it is important to note that these statuses were evaluated at a single point in time. Although some research suggests that alcohol drinking patterns are generally stable [28], recent findings indicate that trauma exposure can significantly alter these patterns. Studies have demonstrated that traumatic events can lead to changes in alcohol use, with some individuals increasing consumption as a coping mechanism, while others might decrease their use due to the negative impacts of alcohol on trauma recovery [29-31]. Another important limitation, therefore, is that alcohol consumption patterns following hospital discharge were not documented. This restricts our understanding of how potential post-injury changes in alcohol use might have influenced PTSD outcomes. Future studies should include repeated alcohol assessments throughout the follow-up period to better capture dynamic changes in drinking behaviors and more accurately interpret their relationships with PTSD development. Additionally, although logistic regression models adjusted for key confounding variables—including previous psychiatric conditions, psychosocial factors, and prior trauma exposure—the complexity of relationships among age, alcohol consumption, and PTSD may not be fully captured by this analytical approach alone. Consequently, residual confounding could still influence our findings, highlighting the need for future studies employing more advanced longitudinal or structural equation modeling techniques to comprehensively elucidate these complex interactions. Additionally, our categorization of non-current drinkers did not distinguish clearly between never drinkers and ex-drinkers, whose lifetime alcohol consumption patterns, especially among older adults, could differ substantially. The lack of detailed historical data regarding previous drinking habits or the duration and reasons for abstinence among ex-drinkers limits our ability to fully understand how these distinct groups might differently impact PTSD outcomes. Moreover, although hazardous alcohol use was defined according to AUDIT scores based on alcohol consumption in the preceding 12 months, the assessment was conducted shortly after hospitalization, which could influence reported drinking patterns and might not reflect typical consumption. Lastly, individuals diagnosed with alcohol use disorders were excluded according to the study’s exclusion criteria, further limiting the generalizability of our findings to broader populations. Finally, while follow-up evaluations were conducted via telephone interviews—an approach validated in previous studies as being as effective as face-to-face interviews [20]—it’s important to acknowledge the inherent limitations of this method.
A key strength of our study is the consecutive recruitment of participants from the entire eligible patient population recently affected by physical injuries. This approach significantly minimized selection bias, ensuring that our sample robustly represented the population of interest. The study’s rigorous design incorporated frequent, standardized follow-up assessments at predetermined intervals, markedly reducing the potential for bias associated with variable examination times. Moreover, our study’s longitudinal approach is pioneering in the field, specifically addressing the associations between age, alcohol drinking, and PTSD. Consistent evaluations and data collection were maintained through a structured research protocol, enhancing the consistency and reliability of the assessments across all participants. The utilization of well-established and standardized scales, including the AUDIT for assessing hazardous alcohol consumption and the CAPS-5 for diagnosing PTSD, significantly strengthened the methodological rigor of the study. Additionally, we collected a comprehensive range of potential covariates for PTSD at baseline, providing a robust framework for our analysis. Long-term follow-up rates were reasonable, and analyses revealed no evidence of selective attrition, further bolstering the credibility and reliability of our findings.
In conclusion, this study offers significant insights into the complex associations between age groups, alcohol drinking status, and PTSD development in patients with physical injuries. For public health implications, our findings suggest the necessity of adopting alcohol drinking and age group-sensitive approaches in PTSD prevention strategies among individuals who have experienced physical trauma. Public health initiatives should focus on educating the community about the roles of alcohol in trauma response, particularly for younger low-risk drinkers who are more susceptible to developing PTSD. For clinical implications, this study suggests the need for healthcare professionals to recognize and adjust to the varying patterns and manifestations of PTSD by age group and alcohol drinking status. This differentiation may require distinct therapeutic strategies and interventions, with an emphasis on early and comprehensive screening for alcohol drinking status. Such targeted approaches will enable timely and appropriate mental health interventions, enhancing the effectiveness of PTSD treatment. For future research, it is crucial to expand the scope of studies to multi-center designs and to include individuals who have experienced a broader spectrum of traumatic events beyond physical injuries. Our finding may serve ground for understanding of the complex relationships between alcohol drinking status, age, and PTSD, paving the way for more effective treatments and preventive strategies.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Supplement1
Print
