Losing a loved one is often accompanied by intense feelings of grief and longing for the deceased. Most bereaved people are able to adapt to the loss over time without professional support (1). However, when grief reactions persist and impede daily functioning, a disorder called Prolonged Grief Disorder (PGD) should be considered.

The conceptualization, diagnostic criteria, and assessment of PGD have been the topics of debate among researchers for many years (2). Some have called for the inclusion of PGD in the International Statistical Classification of Diseases and Related Health Problems (ICD) and the Diagnostic and Statistical Manual of Mental Disorders (DSM) while proposing different diagnostic criteria-sets [e.g., (3–5)]. The DSM-5 workgroup on Trauma/Stress-Related and Dissociative Disorders first decided to include PGD as Persistent Complex Bereavement Disorder (PCBD) in Section III, “Conditions for Further Study” within the DSM-5 to encourage research into the condition (2, 6, 7). Later, the ICD-11 workgroup on Stress-Associated Disorders found the evidence for PGD sufficient and included PGD in the ICD-11 as a new diagnostic entity (8). Based on this inclusion and the collected evidence, the DSM Steering Committee reviewed a proposal to modify the criteria for PCBD. The American Psychiatric Association (APA) then approved the inclusion of PGD in Section II of the DSM-5-TR, thus replacing the criteria for PCBD (9, 10).

Although both classification systems included PGD as a disorder and used the same name, the criteria content and diagnostic approach vary. While the DSM-5-TR provides an explicit diagnostic algorithm, the ICD-11 only uses a typological approach without strict requirements for the number of symptoms that must be present to meet the diagnostic threshold (11). Table 1 provides an overview of the diagnostic criteria. A PGD_DSM-5-TR diagnosis requires one of two separation distress symptoms (persistent and pervasive longing for the deceased and/or persistent and pervasive preoccupation with the deceased), and at least three out of eight accessory symptoms to a functionally impairing degree, with a minimum of 12 months since the loss (12). A PGD_ICD-11 diagnosis, however, only requires at least one of two separation distress symptoms, combined with any of ten accessory symptoms to a functionally impairing degree for an atypically long period of time (at least six months) after the loss (8).

Some researchers have argued that the ICD-11’s typological approach, with simple diagnostic descriptions and no strict requirement for the number of symptoms needed to meet a diagnostic threshold, has the advantage of increasing cross-cultural applicability and being helpful in clinical practice as it results in greater sensitivity in case identification (13). Others, however, criticize this approach for being too liberal since the application of the minimal criteria leads to much higher prevalence rates than the prior used criteria sets for PGD (such as PCBD) or the PGD_DSM-5-TR criteria (11). For instance, Boelen and colleagues compared the PGD_ICD-11 criteria to the former PCBD_DSM-5 criteria and found prevalence rates of 18% vs. 6.4%, respectively. Furthermore, Rosner and colleagues found prevalence rates of 4.2% for PGD_ICD-11 compared to 3.3% for PGD_DSM-5-TR in a representative bereaved sample in Germany (14). A recent study extended these findings and demonstrated limited content overlap between both criteria sets and their predecessors (15). However, these studies are all based on outdated assessment tools not designed to measure PGD according to the DSM-5-TR and ICD-11 criteria. Therefore, researchers recommend the use of validated instruments that capture both criteria sets to assess the prevalence and determine which PGD_ICD-11 algorithm leads to the greatest concordance with PGD_DSM-5-TR (15).

As demonstrated, the information on prevalence rates strongly depends on the chosen diagnostic algorithm. However, accurate estimates of prevalence rates are necessary to understand the health burden and then allocate economic and professional resources accordingly. When estimating the prevalence of PGD, not only the number of symptoms present might be decisive, but also the application of the time or duration criterion. The DSM-5-TR specifies that the loss of a loved one must have occurred at least 12 months ago before a diagnosis can be made. In ICD-11, however, the grief reaction must persist for an atypically long period of time following the loss, with a minimum of 6 months. This wording indicates an alternative application of the duration criterion, that is, that the grief response in itself and not just the bereavement period should last for at least six months. However, in some studies, only the time since loss was assessed, while in others, it is unclear whether the duration of symptoms or simply the time since the loss was assessed (e.g., 16–18).

To avoid medicalizing normal grief and overestimating the prevalence of PGD (19), determining the optimal diagnostic threshold for PGD_ICD11 is essential (20). The aims of the current study were therefore: 1) to estimate the probable point prevalence of PGD according to the DSM-5-TR and ICD-11 criteria using a validated instrument designed to capture both criteria sets and examine the diagnostic agreement, 2) to assess the frequency of occurrence of each symptom of prolonged grief, 3) to investigate the difference between time since loss and symptom duration for a PGD_ICD-11 diagnosis and 4) to determine which PGD_ICD-11 algorithm results in the greatest concordance with PGD_DSM-5-TR when varying the number of required accessory symptoms for a PGD_ICD-11 diagnosis in a representative sample of the population in Germany.

Of the 2,509 participants between the ages of 16 and 95 years, 50.9% were female. The mean age was 49.48 years (SD=17.81). Table 3 provides an overview of the participant characteristics. Of all participants, 1,071 (54.2%) reported having experienced the loss of a close person (e.g., a partner, relative or good friend). Nine participants had to be excluded due to missing data leading to a final data set of n=1,062. Bereaved individuals were primarily middle-aged, and 55.6% were female. The majority reported having lost a parent (40.9%), or other relatives besides children, parent, or partner (28.8%), and natural nonviolent deaths were reported most frequently (89.0%) as the cause of death (see Table 4 ).

The conditional prevalence of PGD using the DSM-5-TR diagnostic algorithm was 4.7% (n = 50), within the population-based sample (including non-bereaved), 2.0%. Using the diagnostic algorithm of the ICD-11 led to a conditional prevalence of PGD of 5.4% (n=57), within the population-based sample (including non-bereaved) 2.3%.

The PGD_DSM-5-TR diagnostic rate was significantly lower than the rate of PGD_ICD-11 (Fisher’s exact test, p <.001). The pairwise agreement between PGD_DSM-5-TR and PGD_ICD-11 was substantial with κ = 0.75, 95% CI [0.66–0.85]. There were nine unique PGD_DSM-5-TR cases (i.e., meeting PGD_DSM-5-TR but not PGD_ICD-11 criteria) and 16 unique PGD_ICD-11 cases (i.e., meeting PGD_ICD-11 but not PGD_DSM-5-TR criteria). The remaining 41 participants (3.9%) met both PGD criteria (see Figure 1 ).

The most frequently indicated symptoms of the TGI-SR+ were having intrusive thoughts or images related to the person who died (30.7%), followed by longing or yearning for the person who died (23.4%), having trouble accepting the loss (23.3%) and feeling stunned, shocked, or dazed by his/her death (19.6%; see Table 2 ). Functional impairment due to the death was reported by 8.9% (n=94).

When simply examining time since the loss (i.e., six months) instead of symptom duration, the prevalence of PGD using the ICD-11 criteria B, C, and D, here referred to as PGD_ICD-11-6M increased to 6.8% (n=72), which is significantly higher than the rate of 5.4% for PGD_ICD-11 (Fisher’s exact test, p <.001) and higher than the rate of 4.7% for PGD_DSM-5-TR (Fisher’s exact test, p <.001). Within the population-based sample (including non-bereaved), the prevalence increased to 2.9%. The pairwise agreement between PGD_DSM-5-TR and PGD_ICD-11-6M with time since loss six months was substantial with κ = 0.81, 95% CI [0.73-0.89]. There were no unique PGD_DSM-5-TR cases and 22 unique PGD_ICD-11-6M cases. The remaining 50 participants (4.7%) met both PGD criteria (see Figure 2 ).

The number of required diagnostic criteria for PGD_ICD-11 was subsequently increased in additional analyses. The conditional prevalence decreased from 5.4% with one accessory symptom to 0.9% with seven accessory symptoms (see Table 5 ). The highest pairwise agreement with PGD_DSM-5-TR was achieved when the number of accompanying symptoms was increased to 3 with κ = 0.79, 95% CI [0.70-0.88]. Increasing the number of accompanying symptoms to 3 for a PGD_ICD-11 diagnosis led to a prevalence of 4.7% (n=50). There were 10 unique PGD_DSM-5-TR cases and 10 unique PGD_ICD-11 cases in this case. The remaining 40 participants (3.8%) met both PGD criteria (see Figure 3 ).

One aim of this study was to estimate the probable point prevalence of PGD using a validated instrument designed to capture the DSM-5-TR and ICD-11 criteria sets and evaluate their diagnostic agreement in a population-based sample. Previous prevalence estimates were predominantly based on old criteria sets or outdated assessment tools. Previous studies also showed little content overlap between diagnostic criteria and found varying prevalence rates. Determining the optimal diagnostic threshold seems essential to avoid medicalizing normal grief and overestimating the prevalence of PGD.

The first main finding was a conditional prevalence of PGD_DSM-5-TR of 4.7%, which was significantly lower than the rate of PGD_ICD-11 of 5.4%. These estimates are slightly higher than the ones found in other population-based studies by Rosner et al. (PGD_DSM-5-TR 3.3%, PGD_ICD-11 4.2%) (14) and Treml et al. (PGD_DSM-5-TR 3.4%) (23). These differences might emerge from the fact that both prior studies used outdated instruments designed for previous diagnostic sets, thus not assessing all criteria, while the TGI-SR+ used in the current study was specifically designed to capture all current diagnostic criteria sets. The pairwise agreement between criteria sets in the present study was substantial (κ = 0.75), with 3.9% of the bereaved meeting both PGD criteria. However, there were still nine unique PGD_DSM-5-TR cases and 16 unique PGD_ICD-11 cases, demonstrating that there is a need for further convergence of the diagnostic criteria and their algorithm.

Another aim was to assess the frequency of occurrence of each symptom of prolonged grief. The advantage of the TGI-SR+ is that it captures not only the current diagnostic criteria but also former criteria for PGD such as PCBD or the criteria set proposed by Prigerson in 2009 (PGD₂₀₀₉), on which much research was based (4). The most frequently indicated symptoms were having intrusive thoughts or images related to the person who died (criterion B2, 30.7%) and longing or yearning for the person who died (criterion B1, 23.4%), which are considered the core criteria for PGD (15). The core criteria were followed by having trouble accepting the loss (23.3%). The last-mentioned symptom, however, is only a criterion within the ICD-11 and not listed within the DSM-5-TR, even though it was listed within all previous diagnostic criteria sets (for an overview see 15). Since this criterion is frequently reported, it could contribute to the higher prevalence rate of PGD_ICD-11 compared to PGD_DSM-5-TR. The next most common symptom, reported by nearly 20%, was feeling stunned, shocked, or dazed by his/her death. This symptom originates from the previous PGD₂₀₀₉ diagnostic criteria by Prigerson (4). In their study, 19.2% indicated feeling stunned, shocked, or dazed about the death, which led the authors to propose this symptom as a diagnostic criterion for PGD. Every study using for instance the PG-13 or the Inventory of Complicated Grief (ICG) subsequently assessed this item as one of the diagnostic criteria. However, the symptom was later not included in the ICD-11 or DSM-5-TR. Some studies used the item “feeling stunned, shocked, or dazed by his/her death” to assess emotional numbness (e.g., 14). In the current study, only 14% reported emotional numbness, indicating that these symptoms might be similar but do not assess the same feelings.

The third aim was to investigate the difference between the time since loss and symptom duration for a PGD_ICD-11 diagnosis. The prevalence rate of PGD_ICD-11 was significantly higher when only the bereavement period rather than the symptom duration was examined (6.8% vs 5.4%). In this case, the diagnostic agreement between PGD_ICD-11-6M and PGD_DSM-5-TR increased (κ = 0.81), since all individuals meeting the DSM-5-TR criteria were also enclosed when assessing the PGD_ICD-11-6M criteria, and there were no unique PGD_DSM-5-TR cases (see Figure 2 ). However, this algorithm led to 22 unique PGD_ICD-11-6M cases and a high prevalence rate of 6.8%. This prevalence rate is likely an overestimation of the true prevalence since all people bereaved for six months who meet the minimum criteria according to ICD-11 are included, even those whose symptoms have persisted for a shorter period of time. This finding indicates that the bereavement period alone might not be a reliable indicator for PGD but rather the symptom duration. Only considering bereaved with symptoms for at least six months leads to a much smaller and presumably more precise estimate that corresponds more closely to the DMS-5-TR. Nevertheless, the intersection between PGD_ICD-11 and PGD_DSM-5-TR still needs improvement. For the clinical practice, we recommend assessing the symptom duration as a vital part in diagnosing PGD instead of simply assessing the bereavement period.

The last aim was to determine which PGD_ICD-11 algorithm results in the greatest concordance with PGD_DSM-5-TR when varying the number of required accessory symptoms for a PGD_ICD-11 diagnosis. The results show that the highest agreement with PGD_DSM-5-TR was achieved when the number of accompanying symptoms for a PGD_ICD-11 diagnosis was increased to 3, which resulted in a lower prevalence of 4.7%. This result aligns with other studies that demonstrated that the minimal criteria of PGD_ICD-11 might be too liberal, as they lead to relatively high prevalence rates (20, 24), and a more conservative scoring rule might be beneficial. For instance, Rosner and colleagues found the highest agreement between both criteria sets with four accessory symptoms (14), while others recommend even five or more accessory symptoms for a PGD_ICD-11 diagnosis (22, 24). Increasing the accompanying symptoms for a PGD_ICD-11 diagnosis leads to a better agreement with PGD_DSM-5-TR, and in our case, even to the same prevalence rate. Nonetheless, there were still 10 unique PGD_DSM-5-TR cases and 10 unique PGD_ICD-11 cases, indicating that there is still variability within the groups of people meeting grief disorder criteria. A less heterogeneous diagnosis would be more beneficial for clinical practice and research.

Taken together, our results have implications for clinical practice. PGD is a serious disorder in the general population that is associated with significant impairments. Healthcare providers should be aware of the diagnosis and screen for it if indicated. However, this should be done using up-to-date measurement tools that capture the current diagnostic criteria, such as the TGI-SR+. When diagnosing PGD, the symptom duration, in particular, should be taken into account. Furthermore, the minimum criteria according to ICD-11 seem to be too liberal, as they lead to relatively high prevalence rates, and a more conservative scoring rule seems appropriate in order to avoid misdiagnosis.

Our findings have to be interpreted in light of some limitations. First, our results are based on self-reported data rather than clinician-administered structured interviews. The exclusive use of self-report measures could have led to bias due to misinterpretation of questions. Furthermore, a cross-sectional design was applied, eliminating any causal conclusions and conclusions about the stability of prolonged grief symptoms. Since grief naturally fluctuates in response to external stressors (e.g., anniversaries, birthdays, holidays), a longitudinal approach using multiple assessments beyond six and twelve months might be more appropriate to avoid diagnosing temporary distress. In our study, the data was collected between October and December, i.e. in some cases shortly before Christmas, which could be such an external stressor. However, the question of whether the symptoms have persisted for at least six months may provide some indication of symptom stability. Additionally, our results regarding the prevalence might be specific to the German general population. Hence, attempts should be made to replicate the findings in different cultures, in order to confirm the generalizability of our findings to other populations. Lastly, general mental health problems and psychiatric comorbidities were not examined. Therefore, the extent to which possible comorbid psychopathology influenced the results is unknown, challenging the interpretation of the results.

A major strength of the current study is the population-based setting, with a sample constructed to be representative in terms of age, gender, and education. The representativeness of the sample is especially relevant since most studies on PGD include predominately female participants leading to an underrepresentation of men.

The results demonstrate that the prevalence of PGD varies significantly depending on the application of the diagnostic algorithm and criteria. The prevalence ranged from 4.7% for PGD_DSM-5-TR to 5.4% for PGD_ICD-11 to even 6.8% for PGD_ICD-11-6M when only considering the bereavement period. Yet, at the same time, the diagnostic agreement between all sets was substantial. Increasing the accessory symptoms for a PGD_ICD-11 diagnosis to 3 lowered the prevalence to 4.7%, which is equivalent to the DSM-5-TR prevalence. However, the overlap between the two systems could to be improved by harmonizing the criteria. PGD remains a substantial disorder among the bereaved and should not be neglected by healthcare providers.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The study and the procedures were approved by the local ethical review board (Leipzig University, Medical Faculty; AZ: 298/21-ek, 12.7.2021). The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation in this study was provided by the participants themselves or by legal guardians/next of kin in case of minors.

JT: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing. KL: Data curation, Formal analysis, Methodology, Writing – review & editing. EB: Investigation, Methodology, Validation, Writing – review & editing. AK: Conceptualization, Resources, Supervision, Validation, Writing – review & editing.