Introduction
Recent updates in diagnostic criteria in Diagnostic and Statistical Manual of Mental Disorders – 5th Edition (DSM-5; American Psychiatric Association, 2013) and International Classification of Diseases – 11th Revision (ICD-11; World Health Organization, 2022) have refined the understanding and identification of post-traumatic stress disorder (PTSD). These revisions reflect evolving perspectives on the disorder and have introduced notable divergences between the two systems. Consequently, there is a pressing need to generate new prevalence data, as most existing studies still rely on ICD-Tenth Revision (ICD-10) and DSM-Fourth Revision (DSM-IV) criteria. Valid epidemiological data are necessary to quantify the impact of PTSD and are critical for estimating treatment needs and informing public health planning. We aim to present updated prevalence data on PTSD in the German general population, which will help clarify the divergences in diagnostic operationalization between ICD-11 and DSM-5 and their impact on case identification.
Data of the WHO World Mental Health Surveys indicate an average lifetime prevalence of PTSD of 3.9% in the general population worldwide and 5.6% among those exposed to trauma (Koenen et al., Reference Koenen, Ratanatharathorn, Ng, McLaughlin, Bromet, Stein, Karam, Meron Ruscio, Benjet, Scott, Atwoli, Petukhova, Lim, Aguilar-Gaxiola, Al-Hamzawi, Alonso, Bunting, Ciutan, de Girolamo, Degenhardt, Gureje, Haro, Huang, Kawakami, Lee, Navarro-Mateu, Pennell, Piazza, Sampson, ten Have, Torres, Viana, Williams, Xavier and Kessler2017). However, Atwoli et al. (Reference Atwoli, Stein, Koenen and McLaughlin2015) report that global lifetime prevalence rates for PTSD vary widely, ranging from 1.3% to 12.2%. In European Countries, prevalence rates range from 1% to 3%, with distinct variation between countries (Trautmann and Wittchen, Reference Trautmann, Wittchen, Nemeroff and Marmar2018). Given the variability in prevalence rates across countries, national studies are crucial for obtaining precise data to inform national healthcare policies. Studies regarding the prevalence of PTSD in Germany show varying results between 0.7% and 3.9% depending on the population under study, diagnostic methods and the time frame considered. Regarding 1-month prevalence, several studies using DSM-IV criteria have produced estimates between 1.5% and 3.9%. For instance, Spitzer et al. (Reference Spitzer, Barnow, Volzke, John, Freyberger and Joergen Grabe2008) reported a prevalence of 1.5% in a predominantly elderly population, sampled between 2002 and 2006, using SCID interviews and the CIDI screener. Maercker et al. (Reference Maercker, Forstmeier, Wagner, Glaesmer and Brähler2008) observed 2.3% in a general population sample (2005) using modified Perceived Stress Scale questionnaires. Glaesmer et al. (Reference Glaesmer, Gunzelmann, Braehler, Forstmeier and Maercker2010) reported 3.9% in a population aged 60 and older (sampled in 2008) based on self-report (Posttraumatic Stress Diagnostic Scale, PDS). Lukaschek et al. (Reference Lukaschek, Kruse, Emeny, Lacruz, von Eisenhart Rothe and Ladwig2013) reported 1.7% in a regional sample from southern Germany (2006–2008), combining interviews and self-report (Impact of Event Scale, PDS), while Hauffa et al. (Reference Hauffa, Rief, Brähler, Martin, Mewes and Glaesmer2011) found 2.9% in the general population, sampled in 2007 with only self-report (PDS). More recently, Maercker et al. (Reference Maercker, Hecker, Augsburger and Kliem2018) reported a 1-month prevalence of 1.5% using ICD-11 criteria and the International Trauma Questionnaire (ITQ) in a general population sample from 2016. Regarding 12-month and lifetime prevalence of PTSD in the German general population, earlier studies based on DSM-IV criteria and various assessment methods reported 12-month prevalence rates ranging from 0.7% to 2.3% (Jacobi et al., Reference Jacobi, Höfler, Siegert, Mack, Gerschler, Scholl, Busch, Hapke, Maske, Seiffert, Gaebel, Maier, Wagner, Zielasek and Wittchen2014; Perkonigg et al., Reference Perkonigg, Kessler, Storz and Wittchen2000; Schmidt et al., Reference Schmidt, Watzke, Schulz, Baumeister, Freyberger and Grabe2013) and lifetime prevalence rates between 1.2% and 3.1% (Perkonigg et al., Reference Perkonigg, Kessler, Storz and Wittchen2000; Spitzer et al., Reference Spitzer, Barnow, Volzke, John, Freyberger and Joergen Grabe2008). PTSD prevalence estimates using DSM-5 criteria for the German general population are lacking so far. The studies mentioned above highlight the variability in PTSD prevalence estimates, influenced by the population under study, the diagnostic framework (DSM-IV vs. ICD-10/-11) and the data collection methods (interview vs. self-report). To ensure accurate estimates, it is crucial to address each of these aspects.
One aspect influencing prevalence estimates is the specific population and geographic and political context under study, as trauma exposure rates and sample characteristics vary significantly across settings. The frequency of potentially traumatizing events like natural disasters or armed conflicts differs across regions and historical periods (Atwoli et al., Reference Atwoli, Stein, Koenen and McLaughlin2015; Koenen et al., Reference Koenen, Ratanatharathorn, Ng, McLaughlin, Bromet, Stein, Karam, Meron Ruscio, Benjet, Scott, Atwoli, Petukhova, Lim, Aguilar-Gaxiola, Al-Hamzawi, Alonso, Bunting, Ciutan, de Girolamo, Degenhardt, Gureje, Haro, Huang, Kawakami, Lee, Navarro-Mateu, Pennell, Piazza, Sampson, ten Have, Torres, Viana, Williams, Xavier and Kessler2017). Additionally, different prevalence rates are observed based on the composition of the sample, whether clinical, trauma-exposed, military or general population (Trautmann and Wittchen, Reference Trautmann, Wittchen, Nemeroff and Marmar2018). Cohort effects within Germany, attributable to the World War II generation, further contribute to variability (Maercker et al., Reference Maercker, Forstmeier, Wagner, Glaesmer and Brähler2008). Consequently, findings are diverse and often confined to the specific sample under examination, precluding broad generalizations.
Another important factor affecting PTSD prevalence estimates is the diagnostic framework used, as DSM-5 and ICD-11 both provide standardized criteria to facilitate diagnosis, yet significant differences exist between the two. The DSM-5 categorizes 20 symptoms into four clusters (re-experiencing, avoidance of reminders of trauma, negative alterations in cognition and mood, heightened arousal or reactivity) (American Psychiatric Association, 2013). The ICD-11 clusters only six symptoms into three categories (re-experiencing/avoidance of reminders of the trauma/sense of threat) (World Health Organization, 2022). Both frameworks require symptoms to persist and to significantly impair social, occupational or other important areas of functioning, yet the DSM-5 offers a more structured outline of symptom duration (DSM-5: more than a month, ICD-11: several weeks). The differences in operationalization lead to discrepancies in PTSD case identification between the two diagnostic frameworks, underscoring the need for further exploration into factors influencing diagnostic decisions. In addition, the definitions of what constitutes a traumatic event (TE) differ slightly: DSM-5 emphasizes objective threat and exposure routes, whereas ICD-11 focuses more on the subjective experience of extreme threat. Research using various assessment tools across different populations consistently shows lower PTSD prevalence rates according to ICD-11 compared to DSM-5 (Hansen et al., Reference Hansen, Hyland, Armour, Shevlin and Elklit2015; Heeke et al., Reference Heeke, O’Donald, Stammel and Böttche2020; Hyland et al., Reference Hyland, Shevlin, McNally, Murphy, Hansen and Elklit2016; O’Donnell et al., Reference O’Donnell, Alkemade, Nickerson, Creamer, McFarlane, Silove, Bryant and Forbes2014; Wisco et al., Reference Wisco, Miller, Wolf, Kilpatrick, Resnick, Badour, Marx, Keane, Rosen and Friedman2016), with only one study reporting slightly higher prevalence rates of PTSD according to ICD-11 (Stein et al., Reference Stein, McLaughlin, Koenen, Atwoli, Friedman, Hill, Maercker, Petukhova, Shahly, van Ommeren, Alonso, Borges, de Girolamo, de Jonge, Demyttenaere, Florescu, Karam, Kawakami, Matschinger, Okoliyski, Posada-Villa, Scott, Viana and Kessler2014). Studies find that more individuals meet the re-experiencing and hyperarousal criteria for DSM-5 compared to ICD-11 (Hyland et al., Reference Hyland, Shevlin, McNally, Murphy, Hansen and Elklit2016; O’Donnell et al., Reference O’Donnell, Alkemade, Nickerson, Creamer, McFarlane, Silove, Bryant and Forbes2014; Schellong et al., Reference Schellong, Hanschmidt, Ehring, Knaevelsrud, Schäfer, Rau, Dyer and Krüger-Gottschalk2019; Wisco et al., Reference Wisco, Miller, Wolf, Kilpatrick, Resnick, Badour, Marx, Keane, Rosen and Friedman2016). Specifically, intrusive memories play a crucial role in this disparity, with studies showing that additionally including intrusive memories in ICD-11 criteria aligns prevalence rates more closely with DSM-5 (O’Donnell et al., Reference O’Donnell, Alkemade, Nickerson, Creamer, McFarlane, Silove, Bryant and Forbes2014; Schellong et al., Reference Schellong, Hanschmidt, Ehring, Knaevelsrud, Schäfer, Rau, Dyer and Krüger-Gottschalk2019).
In addition to population and diagnostic framework, the method of data collection (the use of clinical interviews versus self-report instruments) can can substantially influence prevalence estimation. Clinical interviews are considered the gold standard for their accuracy but are challenging to implement on large-scale samples. Self-report instruments, though more feasible for widespread use and screening, only provide probable PTSD diagnosis (Bovin and Marx, Reference Bovin and Marx2023). Beyond the use of diagnostic algorithms based on self-report instruments, empirically derived cut-off-scores are a common procedure for case identification. However, when used for prevalence estimation, they require adjustments and critical reflection to ensure accuracy (Terhakopian et al., Reference Terhakopian, Sinaii, Engel, Schnurr and Hoge2008). Cut-off scores, optimized for sensitivity, are lenient regarding false positives, potentially leading to overestimation of positive cases, especially in populations with a low PTSD base rate. For example, in a population of 1000 people with only 4 true PTSD cases (0.4% prevalence), a test with perfect sensitivity (1.0) and 75% specificity would incorrectly identify 249 individuals as positive (25% of the 996 true negatives). This results in an inflated prevalence estimate of 25.3%. Therefore, prevalence estimates using self-report measures in populations with low base rates of PTSD, such as the general population, require careful evaluation of existing cut-off values to prevent overestimation. When self-report instruments are used without a clinical interview, it is recommended to apply a cut-off value established through comparison with clinical interview outcomes in an independent sample, prioritizing a high positive predictive value and accurate estimation of true prevalence (Pettrich et al., Reference Pettrich, Schellong, Dyer, Ehring, Knaevelsrud, Krüger-Gottschalk, Nesterko, Schäfer and Glaesmer2025; Terhakopian et al., Reference Terhakopian, Sinaii, Engel, Schnurr and Hoge2008).
Beyond these factors, stratification by age and gender appears to be crucial in understanding PTSD risk. Previous research suggests that women may be at a higher risk of developing PTSD, particularly in response to trauma types like sexual violence (Kessler et al., Reference Kessler, Rose, Koenen, Karam, Stang, Stein, Heeringa, Hill, Liberzon, McLaughlin, McLean, Pennell, Petukhova, Rosellini, Ruscio, Shahly, Shalev, Silove, Zaslavsky, Angermeyer, Bromet, de Almeida, de Girolamo, de Jonge, Demyttenaere, Florescu, Gureje, Haro, Hinkov, Kawakami, Kovess-Masfety, Lee, Medina-Mora, Murphy, Navarro-Mateu, Piazza, Posada-Villa, Scott, Torres and Carmen Viana2014; Tolin and Foa, Reference Tolin and Foa2008). However, this gender difference seems to be influenced by various mediating factors, such as pre-existing mental health conditions and neurobiological responses (Kornfield et al., Reference Kornfield, Hantsoo and Epperson2018; McLean and Anderson, Reference McLean and Anderson2009). Age-related risk factors also seem to play a role, with younger individuals possibly facing higher risk due to trauma experienced during vulnerable developmental stages (Teicher et al., Reference Teicher, Anderson, Ohashi, Khan, McGreenery, Bolger, Rohan and Vitaliano2018). Conversely, for older adults, PTSD may arise from the cumulative impact of lifetime trauma, such as war-related experiences, although PTSD is often underreported in this population due to age-related vulnerabilities (Brewin et al., Reference Brewin, Andrews and Valentine2000; Frans et al., Reference Frans, Rimmö, Åberg and Fredrikson2005; Ogle et al., Reference Ogle, Rubin and Siegler2014). These gender- and age-related disparities in PTSD risk underscore the multifactorial nature of trauma response, a central theme in the present study’s investigation of PTSD prevalence across different demographic groups.
The present study examines the prevalence of TEs and probable PTSD in a representative sample of the general German population and aims to assess the comparability of PTSD prevalence rates based on DSM-5 and ICD-11 criteria. We use a widely recognized self-report screening tool and apply various diagnostic algorithms to evaluate PTSD criteria. While ICD-11 distinguishes between PTSD and Complex PTSD (CPTSD), our study focused exclusively on core PTSD symptoms as operationalized by DSM-5 and ICD-11 definitions. Additionally, the study investigates trauma and PTSD prevalence stratified by age and gender, exploring disparities and risk factors within the German general population.
Methods
Subjects
This cross-sectional study is part of a representative survey of the general German population conducted in 2016 by an independent market and social research institute (USUMA, Berlin). The sampling procedure is described elsewhere in more detail (Pettrich et al., Reference Pettrich, Friedrich, Nesterko and Glaesmer2024). In total, 2510 people (aged 14–94 years) filled out the self-report questionnaire given to them. After excluding 86 underage participants and 20 subjects with missing data on all items of the PTSD Checklist for DSM-5 (PCL-5) and Life-Events-Checklist for DSM-5, the final sample consisted of 2404 participants aged 18–94 years. Characteristics of the study sample are presented in Table 1. In our pursuit to derive prevalence estimates based on the PCL-5, we used the weighted sample. The resulting analysis sample was representative regarding age (M = 50.24; SD = 18.07) and gender for the Federal Republic of Germany when compared to the microcensus of the Federal Office of Statistics of the sampling year. The Ethics Committee of the Medical Faculty of the University of Leipzig determined the proposed study to be outside their scope but expressed no ethical concerns for the survey.
Table 1. Socio-demographic characteristics in the sample (N = 2404)
Instruments
The LEC-5 (Krüger-Gottschalk et al., Reference Krüger-Gottschalk, Knaevelsrud, Rau, Dyer, Schäfer, Schellong and Ehring2017; Weathers et al., Reference Weathers, Litz, Keane, Palmieri, Marx and Schnurr2013) is a self-report measure assessing the lifetime prevalence of TEs. Sixteen predefined TEs and the option to describe any other very stressful event or experience not listed before are presented. The A-criterion of the PTSD diagnosis is endorsed when at least one of the events is answered by ‘Happened to me’, ‘Witnessed it’ and ‘Part of my job’. The answer options ‘Heard about it’, ‘Not sure’ and ‘Doesn’t apply’ are not considered affirmative. Although originally developed for DSM-5, the LEC-5 is widely used in ICD-11 PTSD research to ensure a consistent assessment of the A-criterion (e.g., Cloitre et al., Reference Cloitre, Shevlin, Brewin, Bisson, Roberts, Maercker, Karatzias and Hyland2018).
The PCL-5 (Krüger-Gottschalk et al., Reference Krüger-Gottschalk, Knaevelsrud, Rau, Dyer, Schäfer, Schellong and Ehring2017; Weathers et al., Reference Weathers, Litz, Keane, Palmieri, Marx and Schnurr2013) consists of 20 items covering the DSM-5 PTSD symptoms. Symptom severity in the past month can be rated on a 5-point Likert scale from 0 (‘Not at all’), 1 (‘A little bit’), 2 (‘Moderately’), 3 (‘Quite a bit’) to 4 (‘Extremely’). Following the DSM-5 diagnostic criteria, a diagnosis can be derived if one re-experiencing symptom, one avoidance symptom, two negative alterations in thinking and mood symptoms and two heightened arousal or reactivity symptoms are endorsed (American Psychiatric Association, 2013). Although the PCL-5 has been originally developed to assess DSM-5 criteria, a subset of items was created in accordance with previous research (Heeke et al., Reference Heeke, Franzen, Hofmann, Knaevelsrud and Lenferink2022, Reference Heeke, O’Donald, Stammel and Böttche2020; Kuester et al., Reference Kuester, Köhler, Ehring, Knaevelsrud, Kober, Krüger-Gottschalk, Schäfer, Schellong, Wesemann and Rau2017; Schellong et al., Reference Schellong, Hanschmidt, Ehring, Knaevelsrud, Schäfer, Rau, Dyer and Krüger-Gottschalk2019) to assess ICD-11 diagnostic criteria (items 2, 3, 6, 7, 17, 18) and a probable PTSD diagnosis can be given if one re-experiencing, one avoidance of reminders of the trauma and one sense of threat symptom are fulfilled (World Health Organization, 2022). Based on an empirically derived cut-off score, the PCL-5 screens positive for PTSD if the sum score of all items exceeds a cut-off score of 31–33 in the German population for clinical use (Krüger-Gottschalk et al., Reference Krüger-Gottschalk, Knaevelsrud, Rau, Dyer, Schäfer, Schellong and Ehring2017) and of 38 for prevalence estimation (sensitivity: 0.84, specificity: 0.703) (Pettrich et al., Reference Pettrich, Schellong, Dyer, Ehring, Knaevelsrud, Krüger-Gottschalk, Nesterko, Schäfer and Glaesmer2025).
Statistical procedure and considerations
All analyses were conducted using R (R Core Team, 2021). In this data set, in 4.83% of the subjects, the answers were incomplete (with a total of 0.6% missing data points throughout the whole data set). Little’s missing completely at random test indicated that data were missing systematically (χ 2(874) = 1263.7, p < .001). Therefore, the missing data treatment used the predictive mean matching method of the mice package (Buuren and Groothuis-Oudshoorn, Reference Buuren and Groothuis-Oudshoorn2011). We defined the A-criterion for estimating PTSD prevalence as met if participants endorsed at least one of the 16 predefined TEs on the LEC-5. Other stressful events were not included in the A-criterion, since the answers were very heterogenous and did not meet the A-criterion in many cases. Prevalence estimations included frequencies of probable diagnoses based on diagnostic algorithms according to DSM-5 and ICD-11 and a cut-off score of 38 (Pettrich et al., Reference Pettrich, Schellong, Dyer, Ehring, Knaevelsrud, Krüger-Gottschalk, Nesterko, Schäfer and Glaesmer2025), stratified by age and gender. Trauma frequencies were reported based on the worst event reported, with unspecified TEs or no indicated worst event categorized as ‘Other Stressful Event/Not Specified’. Chi-square tests with simulated p-values (Monte Carlo method) were used for difference testing of stratified prevalences where expected cell counts were low. McNemar’s test for paired samples was employed to compare case identification across different diagnostic frameworks.
Results
A total of 1135 subjects (47.21%) of the sample reported at least one TE. Of these, 438 (18.22%) had only one TE, 569 (23.67%) had between two and five, and 128 (5.32%) reported more than five. The most common TEs were transportation accident (n = 175, 7.28%), life-threatening illness/injury (n = 118, 4.91%), severe suffering (n = 109, 4.53%) and serious accident (n = 76, 3.16%). A total of 573 (48.81%) male and 562 (45.69 %) female subjects reported at least one TE. Overall exposure to at least one TE did not differ between gender (χ 2 = 2.34, p = 0.13) or age groups (χ 2 = 7.06, p = 0.072). Certain TEs were reported significantly more often by female subjects, including natural disasters, sexual assaults, unwanted sexual experiences and severe suffering. In contrast, male subjects reported higher exposure to fire/explosion, transportation accidents, serious accidents and physical assault. Older adults were more likely to report war-related TEs (for detailed information and test statistics, see Table 2).
Table 2. Prevalence and characteristics of index TE stratified by gender and age
Abbreviations: n = sample size, % = percentage of participants, χ 2 = McNemar’s chi-squared test statistic.
a Any TE: endorsement of at least 1 of the 16 predefined LEC-5 events, excluding ‘other event’ type. Significance Levels: *p < .05 = statistically significant, **p < .01 = highly statistically significant, ***p < .001 = extremely statistically significant.
The prevalence of PTSD in the sample was found to be 4.7% (n = 113) according to both the ICD-11 and the DSM-5 diagnostic criteria, 4.08% (n = 98) according to the traditional clinical cut-off scoring method (cut-off score = > 33) and 2.62% (n = 63) according to the prevalence cut-off scoring method (cut-off score = > 38), we favoured in this study. No significant differences were found for gender (DSM-5: χ 2(1) = 0.004, p = .951; ICD-11: χ 2(1) = 1.843, p = .175; cut-off: χ 2(1) = 0.696, p = 0.404) or age groups (DSM-5: χ 2(3) = 3.598, p = .308; ICD-11: χ 2(3) = 1.265, p = .737; cut-off: χ 2(3) = 4.01, p = .261) for any of the prevalence estimates. Probable PTSD prevalence in trauma-exposed individuals (n = 1135) ranged between 5.55% (cut-off scoring method) and 9.96% (DSM-5 and ICD-11 scoring method). Higher conditional PTSD prevalence rates were observed in subgroups exposed to sexual (16.22–32.43%) and physical assaults (2.38–21.43%), unwanted sexual experiences (14.29%) and causing harm to others (14.29%) depending on the diagnostic framework (see Table 3).
Table 3. PTSD prevalence by index TE and comparison of DSM-5, ICD-11 and cut-off case identification
Abbreviations: n = sample size, % = percentage of participants, M (SD) = Mean (standard deviation), χ 2 = McNemar’s chi-squared test statistic.
a Any TE: endorsement of at least 1 of the 16 predefined LEC-5 events, excluding ‘other event’ type.
b Agreement in case identification: the percentage reflects the agreement of positive cases identified by the diagnostic systems, relative to those identified as positive by either system, indicating diagnostic concordance. McNemar χ2 tests assess whether differences in identification between the diagnostic systems are statistically significant.
* Significance Levels: p < .05 = statistically significant, **p < .01 = highly statistically significant, ***p < .001 = extremely statistically significant.
While overall PTSD prevalence rates were identical between the DSM-5 and ICD-11 criteria, the two frameworks identified partially different cases (see Table 3). The agreement between the DSM-5 and ICD-11 criteria was substantial, though not perfect (Cohen’s κ = 0.62; 95% CI [0.54, 0.69]). Of the subjects who screened positive for probable PTSD according to either diagnostic algorithm, 48.68% were identified by both systems. McNemar’s test showed no significant difference between the two diagnostic criteria (χ 2(1) = 0, p = 1.00), suggesting that the discrepancies in case identification were evenly distributed between the two systems.
Upon closer examination of the symptom clusters, significant differences were found in the re-experiencing criteria between ICD-11 and DSM-5. Cohen’s kappa for the re-experiencing criteria was moderate (κ = 0.57; 95% CI [0.52, 0.61]). Of the subjects who screened positive in the re-experiencing symptom cluster according to either system, 52.18% were identified by both diagnostic frameworks. McNemar’s chi-square test showed a significant difference in the proportions of cases identified (χ 2(1) = 217, p < 0.001). In contrast, no significant differences were found in the fulfillment of the combined D- and E-criteria of the DSM-5 and the Sense of Threat criteria of the ICD-11.
The cut-off scoring method (cut-off score>38) showed higher agreement with the DSM-5-algorithm (Cohen’s κ = 0.60, 95% CI [0.51, 0.68]) than with the ICD-11-algorithm (Cohen’s κ = 0.42, 95% CI [0.33, 0.52]). Of the subjects who screened positive for PTSD by either the diagnostic algorithm or the cut-off scoring method, 45.45% were identified by both the DSM-5 and the cut-off method, while 30.37% were identified by both the ICD-11 and the cut-off method. For more detailed analyses of symptom clusters and contingency tables, see appendix.
Discussion
In large-scale population studies, it is often not feasible to conduct clinical interviews. Moreover, it is well-known that in samples with relatively low base rates of PTSD, self-report measures tend to overestimate prevalence rates. Therefore, in this study, we estimated the prevalence of PTSD in the German general population using a cut-off value specifically designed for prevalence estimation, as well as commonly used diagnostic algorithms based on DSM-5 and ICD-11 criteria. In addition to providing an updated prevalence estimate, we also aimed to further investigate the comparability of the DSM-5 and ICD-11 frameworks. Our findings yield several important conclusions.
Firstly, approximately 2.62% of the population screened positive for PTSD using a conservative cut-off (cut-off score = 38), whereas prevalence estimates reached up to 4.7% when applying the more lenient DSM-5/ICD-11 diagnostic algorithm. Additionally, almost half of the German general population has experienced at least one TE. Importantly, demographic factors such as age and gender did not significantly affect overall trauma exposure or PTSD diagnostic status, though specific trauma types were unevenly distributed. Women were more likely to report sexual assault and severe suffering, while men reported higher exposure to fire or explosions, transportation accidents, serious accidents and physical assault. In terms of PTSD prevalence comparability, although both DSM-5 and ICD-11 yielded identical prevalence rates, they identified partially different individual cases. With only about half of the identified cases supported by both diagnostic frameworks, it is evident that while both systems capture similar proportions of PTSD cases, their diagnostic criteria prioritize particularly the re-experiencing cluster differently. Furthermore, the comparison of DSM-5 and ICD-11 with the cut-off-scoring method revealed that DSM-5 had a higher agreement with the cut-off-based procedure (cut-off 38) than ICD-11. This highlights the need for continued refinement of diagnostic algorithms to ensure accurate identification of PTSD cases across varying assessment methods. Lastly, our findings indicate that using the PCL-5 as a self-report instrument, normed against a clinical interview for clinical purposes (diagnostic algorithms, cut-off = 33), results in much higher PTSD prevalence rates compared to using the same instrument with a cut-off value specifically designed for prevalence estimation (cut-off = 38), where the rates align more closely with expected prevalence.
The prevalence rates in our study are slightly higher than those reported in previous research. Even using more conservative, cut-off scores adjusted for prevalence estimation, our rates are on the higher end of those found in similar populations, such as 2.9% by Hauffa et al. (Reference Hauffa, Rief, Brähler, Martin, Mewes and Glaesmer2011) and 1.6% by Maercker et al. (Reference Maercker, Hecker, Augsburger and Kliem2018). The reasons for this upward trend remain unclear, but one potential explanation could be the different diagnostic systems used in the previous studies – with Hauffa et al. (Reference Hauffa, Rief, Brähler, Martin, Mewes and Glaesmer2011) using DSM-IV criteria and the PDS and Maercker et al. (Reference Maercker, Hecker, Augsburger and Kliem2018) using the ITQ based on ICD-11 criteria. Our findings contrast with much of the existing literature, where most studies report higher 1month PTSD prevalence rates using DSM-5 compared to ICD-11 criteria, while we observed equivalent rates with substantial agreement for both. Several studies, such as Hansen et al. (Reference Hansen, Hyland, Armour, Shevlin and Elklit2015) and Hyland et al. (Reference Hyland, Shevlin, Brewin, Cloitre, Downes, Jumbe, Karatzias, Bisson and Roberts2017), found lower PTSD diagnostic rates based on ICD-11 criteria in Danish (via Harvard Trauma Questionnaire symptom mapping) and British (ITQ and PCL-5) populations. Similarly, Heeke et al. (Reference Heeke, O’Donald, Stammel and Böttche2020) reported that the DSM-5 criteria identified more individuals with PTSD among traumatized refugees (PCL-5 symptom mapping). O’Donnell et al. (Reference O’Donnell, Alkemade, Nickerson, Creamer, McFarlane, Silove, Bryant and Forbes2014) observed DSM-5 producing twice the PTSD prevalence (6.7%) compared to ICD-11 (3.3%) via the Clinician-Administered PTSD Scale interview, and Wisco et al. (Reference Wisco, Miller, Wolf, Kilpatrick, Resnick, Badour, Marx, Keane, Rosen and Friedman2016) noted ICD-11 yields 10-30% lower prevalence rates due to its stricter symptom criteria (National Stressful Events Survey symptom mapping). Our results, however, align with studies like Morina et al. (Reference Morina, van Emmerik, Andrews and Brewin2014), who found no significant differences between DSM-IV and ICD-11 in war survivors (PDS symptom mapping), and Kuester et al. (Reference Kuester, Köhler, Ehring, Knaevelsrud, Kober, Krüger-Gottschalk, Schäfer, Schellong, Wesemann and Rau2017), where DSM-5 (56%) and ICD-11 (48%) prevalence rates were not statistically different (PCL-5 symptom mapping).
This study has several notable strengths. First, it utilizes a large, representative dataset, making it the first German PTSD prevalence study to assess DSM-5 criteria and directly comparing prevalence rates according to ICD-11 and DSM-5 in the general population. Our findings contrast with previous studies that typically reported higher PTSD prevalence under DSM-5 criteria; this effect was not observed in our sample among individuals with at least one TE. Additionally, the study demonstrates that despite showing substantial agreement, DSM-5 and ICD-11 criteria identify partially different groups of individuals with PTSD. The re-experiencing cluster seems to play a major role here. This finding highlights the importance of understanding how different diagnostic criteria may influence case identification. Finally, this study introduces a novel approach by using a cut-off for prevalence estimation (Pettrich et al., Reference Pettrich, Schellong, Dyer, Ehring, Knaevelsrud, Krüger-Gottschalk, Nesterko, Schäfer and Glaesmer2025) on a self-report instrument to mitigate the tendency for inflated results commonly associated with self-reports. Despite these strengths, there are limitations to consider. Notably, the ICD-11 measure was adapted from an instrument originally developed for DSM-5 criteria. Consequently, the findings and implications regarding the diagnostic performance of ICD-11 criteria are specific to this operationalization. Additionally, it should be considered when interpreting diagnostic comparisons, that trauma exposure in this study was assessed using the DSM-oriented LEC-5. An important limitation of our study is that it does not differentiate between PTSD and CPTSD as defined in ICD-11. Our ICD-11 approximation was based on selected PCL-5 items and did not assess Disturbances in Self-Organization, a key criterion for CPTSD. Therefore, some individuals who would meet criteria for CPTSD may be included in our ICD-11 PTSD estimate. Similarly, the dissociative subtype of PTSD described in DSM-5 was not assessed, as the PCL-5 does not include items related to depersonalization or derealization. While trauma types commonly associated with CPTSD were reported by a minority of the sample, we cannot exclude the possibility that a proportion of CPTSD cases are misclassified. Future prevalence studies should incorporate validated instruments such as the ITQ to capture both PTSD and CPTSD as defined in ICD-11. Additionally, both diagnostic systems were applied only at the symptom cluster level, and we did not account for the time criterion, subjective impairment or exclusion criteria of substance use or physical factors due to a lack of available information. The absence of a clinical interview restricts our conclusions to PTSD screening diagnoses. Furthermore, the suggested cut-off of 38 for prevalence estimation was normed on a clinical sample with a higher PTSD base rate, which may limit its applicability to the general population. Additionally, a potential limitation of our study could be that we did not exclude participants with cognitive impairments, which may have affected the accuracy of their self-report responses.
Future research should explore the mechanisms behind the identification of different cases at the symptom cluster level. Moreover, prevalence studies using clinical interviews should be conducted in the general German population to further validate our findings. A meta-analysis could also be beneficial to integrate existing knowledge from the growing body of prevalence studies conducted on various samples within the German population and beyond.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S2045796025100164.
Availability of data and materials
The data are not publicly available due to the informed consent given by the participants.
Acknowledgements
During the revision of this work, the author used ChatGPT v2 in order to refine language. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.
Financial support
This work is funded by the Open Access Publishing Fund of Leipzig University supported by the German Research Foundation within the program Open Access Publication Funding.
Competing interests
None.
Ethical standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2013.
Open access
