1. Introduction
The core psychopathology of delusional disorder (DD) is the presence of a persistent non-bizarre delusional system. DSM-IV describes seven different types of DD attending mainly to the content of the delusions (persecutory, jealous, somatic, grandiose, erotomanic, mixed and not otherwise specified (NOS)). [Reference American Psychiatry Association1] DD is a surprisingly poorly researched psychotic disorder on which both clinical and epidemiological studies are extremely rare. In addition, studies to date using current diagnostic criteria are even rarer, as the majority of such studies have not used DSM-IV criteria [Reference Kendler20]. The lack of evidence about this disorder is partly due to an assumed low prevalence of DD and also to the fact that many DD patients do not seek psychiatric help unpromptedly [Reference Munro24].
1.1. Sociodemographics of DD
Kendler's meta-analysis continues to be the seminal descriptive study on the epidemiology of DD [Reference Kendler20]. It was based on 17 existing studies about the frequency of DD, mostly based on data from patients being admitted to hospital with a diagnosis of paranoia. Estimated prevalence for DD, inferred from admissions of patients with a diagnosis of paranoia, ranged between 24 and 30 cases in every 100,000 inhabitants. The mean age of onset ranged between 35 and 45 years and female patients slightly outnumbered male patients (female/male ratio is 1.2/1). Over the last decade, there have been four other smaller epidemiological studies on DD showing an even higher female/male ratio (female/male ratios ranging from 3/1.9) [21,22,25,31]. Conversely, another study showed minimal excess numbers of male patients (female/male ratio of 0.86) [Reference Hsiao, Liu, Yang and Yeh12]. Regarding marital status, 60–75% of the patients were married when they were first admitted into hospital, whilst between a quarter and a third of the remaining sample were widowers, separated or divorced [Reference Kendler20]. However, Maina et al., studying an out-patient sample, found a lower percentage of married patients (47.8%), a higher percentage of single individuals (43.5%) and a lower percentage of divorced or separated individuals (6.5%) and widowers (2.2%) [Reference Maina, Albert, Bada and Bogetto21].
1.2. Risk factors for DD
Social risk factors of DD include low socioeconomic status, older age, family history of psychiatric disorders, immigration, sensory deficit and exposure to stressful events. Kendler found that two-thirds of DD patients belong to low socioeconomic status and estimated that 20–25% of the new cases of DD take place in old age [Reference Kendler20]. Relatives of DD patients do not seem to be at higher risk of either schizophrenia or mood disorders but appear to show an increased rate of paranoid personality disorder and/or a premorbid tendency to jealousy, suspiciousness, and secretiveness [14–17,19,28–30]. An English study by Hitch and Rack suggested a prevalence of 16% of paranoid ideas among immigrants compared to a significantly smaller rate of 4% among English natives [Reference Hitch and Rack11]. As mentioned earlier, sensory deficits have also been reported as a risk factor for paranoia. Cooper et al. found that 21% of deaf people tend to develop delusions and other psychotic symptoms, whilst other studies have related deafness to late paraphrenia or schizophrenia [5,6,8,26]. Deafness can more likely be seen as a correlate of psychosis rather than a specific risk factor for DD.
1.3. Comorbid psychopathology
Maina et al. studied a series of 64 cases of DSM-IV DD out-patients establishing that 31.3% had at least one comorbid axis I disorder (21.9% had a mood disorder and 3.1% had at least one comorbid anxiety disorder [Reference Maina, Albert, Bada and Bogetto21]. In another study, Hsiao et al. found that hallucinations were also common among DD patients with non-prominent auditory hallucinations being the most frequent type (11.6%), followed by the tactile hallucinations (5.8%), non-prominent visual hallucinations (2.3%) and olfactory hallucinations (2.3%) [Reference Hsiao, Liu, Yang and Yeh12]. Finally, reported comorbid personality disorders include paranoid, schizotypal and schizoid personality disorders [13,18,23].
1.4. DD types
Most studies report that by and large the most prevalent DD presentation is the so-called persecutory type [21,25,31]. Excluding the mixed type, Yamada found that, among their 54 DSM-III-R DD cases, the persecutory type was the most frequent (51%), followed by the somatic type (27.5%) and the jealous type (13.7%) [Reference Yamada, Nakajima and Noguchi31]. Similarly, Hsiao et al., in a retrospective study with 86 cases of DD according to DSM-IV criteria, also found that the persecutory type was the most frequent (70%), followed by the mixed cases (14%) and those with jealous type (8%) [Reference Hsiao, Liu, Yang and Yeh12]. The latter study also found no significant differences between these four most frequent types in terms of sex, age of onset, frequency of hallucinations and the presence of depression. Maina, with patients diagnosed using DSM-IV criteria, also found a higher prevalence of the persecutory type (54.4%), followed by somatic (17.4%), mixed (15.2%) and jealous types (6.5%) [Reference Maina, Albert, Bada and Bogetto21]. This study also reported a higher frequency of comorbid mental disorder (mainly mood disorders) among persecutory cases (54.4%) and a lower rate of psychiatric comorbid conditions among the mixed cases (66.7%) [Reference Maina, Albert, Bada and Bogetto21].
1.5. Aims of the study
Provided the conspicuous shortage of empirical descriptive studies on DD, this study provides a unique opportunity to explore and describe clinical correlates of DD based on a relatively large case register of DD patients. We aim to describe and quantify frequencies of DD types, empirically describe sociodemographic and psychopathological features of DD, explore the comorbidity and global functioning of DD patients and identify clinical and risk correlates of specific DD types.
2. Methods
2.1. Setting and design
Sant Joan de Déu – Serveis de Salut Mental (SJD – SSM) is a public mental health services provider covering both community and in-patient psychiatric care for a well-defined catchment area in the South of the province of Barcelona (Spain). Psychiatric care provision within this geographic area during the study period (2001–2004) was implemented via six community mental health out-patient teams, three day-hospitals, three acute in-patient care sub-units and a rehabilitation unit. Such groups of resources served some 607,494 inhabitants. In Spain, referral to psychiatric care is generally issued by general practitioners (GP); access to these referrals is free and universal. Therefore, most cases detected by GP would be referred to their local community mental health team for further assessment and treatment, enabling us to establish a fairly thorough case register of all DD patients being cared for by such a large community care resource setting.
In addition, SJD – SSM has developed a sophisticated computerised psychiatric medical note software package named HCI that ensures a highly structured method for medical-notes taking and psychopathological assessment procedures. It also enforces multi-axial DSM-IV diagnoses in all cases attending SJD – SSM resources [Reference Autonell and Mochón2]. All SJD – SSM psychiatrists had been formally trained to compulsorily incorporate, as the result of each individual consultation, all axes' DSM-IV diagnoses, structured mental-state psychopathology assessments and medication updates to such software. There is no alternative way to add notes to medical records within all SJD – SSM resources. The HCI software package also records consultation types to register psychiatric resources used by each patient.
2.2. The sample
The initial sample consisted of all 467 patients who were clinically recorded as having a diagnosis of DD. A systematic structured assessment including a checklist of all DSM-IV criteria was used in a retrospective thorough evaluation of medical records included in the HCI software from January 2000 to December 2003. Following such accurate diagnostic review procedure, 370 patients were finally found to completely fulfil DSM-IV criteria for DD and constituted our final DD case-register sample (n = 370).
2.3. Diagnostic procedure for inclusion in the delusional disorder case register
A structured symptom checklist based on DSM-IV criteria for DD was used to perform a thorough retrospective inspection of all entries on HCI for each patient clinically registered with a diagnosis of DD (see Table 1). Retrospective assessments to check the adequacy of DD diagnoses were made twice by a fully trained psychiatrist and a trained pre-doctoral research psychologist. Both professionals discussed their findings together within an expert panel which included an additional senior psychiatric supervisor before deciding, following DSM-IV criteria, whether to consider a patient as a certain DD case to be included in the study. This panel also assigned patients to the appropriate DSM-IV DD type considering all information gathered both from the HCI software and from the retrospectively completed symptom checklist. In summary, all cases who did not strictly fulfil all DSM-IV criteria for DD were excluded. Excluded cases were mostly patients who had been originally diagnosed as DD but ended up having sufficient criteria for a diagnosis of schizophrenia or other psychoses.
Table 1 Symptom checklist used to register clinical data from computerised medical records

2.4. Assessment of independent variables
Should patients fulfil DSM-IV criteria for DD, an additional retrospective assessment was made using a structured risk factor questionnaire for DD based on risk factors suggested by previous research (Table 1). Thus, information was gathered on a dichotomous manner (Yes versus No) from the medical records, on sociodemographic variables, associated phenomenology, family and personal history of psychoses, comorbidity with other axis I disorders, comorbid personality disorder, global functioning using the Global Assessment of Functioning (GAF) scale, medication use and psychiatric service use over a three-year period at different levels of care (out-patient, day-hospital, emergencies and in-patient admissions).
2.5. Statistical methods
Univariate associations between DD types and independent variables were explored by the use of parametric (ANOVA) or non-parametric (Kruskal–Wallis test) tests as appropriate.
3. Results
3.1. Sample description
Following a systematic review of medical records of all 467 DD patients originally included in our case register as DD cases, we confirmed DSM-IV diagnostic criteria for DD in 370, who were finally included in this study (n = 370). Table 2 shows the samples' sociodemographic characteristics. In brief, DD cases are more frequently middle aged married housewives with a higher than expected frequency of family history of schizophrenia. The female/male ratio in the sample was 1.29. The sample showed a low level of global functioning (mean GAF 51.7; ±14.18 SD) and a high level of mental health services use (e.g., mean number of out-patient consultations over a three-year period: 32.12; ±43.96 SD). A percentage (81.4%) of DD patients were on antipsychotic medication and of those 60% were taking an atypical antipsychotic.
Table 2 Correlates of independent variables with delusional disorder types

*statistically significant.
3.2. Correlates of DD types
The most frequent specific DD type in the sample was persecutory type (47%), followed by mixed (11%) and jealous (9%) types. Twenty-three percent of the patients were diagnosed as DD NOS. Tables 2–4 show a summary of how different DD types compare regarding all independent variables. We present the following three consecutive analyses: (a) comparing all types (Table 2); (b) comparing persecutory versus all other types (Table 3); and (c) jealous subtypes versus the rest (Table 4). When all types were compared with each other some significant differences emerged (Table 2). Thus, compared to all other DD types, patients with grandiose type were more likely to be single, to have an onset following a stressful precipitating factor and to have a significantly worse global functioning and a higher psychiatric services use. Conversely, patients with somatic type significantly presented a higher frequency of rare perceptual abnormalities such as tactile, olfactory and non-prominent visual hallucinations. On the other hand, those with erotomanic type were significantly more likely divorced or widowed women and tended to have a poorer global functioning (see Table 2 for details).
Table 3 Persecutory versus other DD types by independent variables

*statistically significant.
Table 4 Jealous versus other DD types by independent variables

*statistically significant.
Persecutory type was significantly associated with not being married, using more antidepressants and having the highest percentage of atypical antipsychotic use (Table 3). Persecutory type versus all other DD cases lumped together, we found the persecutory type, when compared with all other types lumped together, to be significantly associated with patients who tended significantly to have worse functionality, had a high personal history of depression and use more atypical antipsychotics (Table 3).
Finally, jealous type compared with all other types tended significantly to be more frequently married, have a comparatively better global functioning, and also have a higher frequency of antidepressant use (Table 3). Additionally, when compared with the aggregate of all other DD types, we found jealous type to be significantly associated with higher frequencies of being male, married, having a family history of schizophrenia, having immigrated, having a personal history of depression and having a high comorbidity with depressive symptoms and more antidepressant and atypical antipsychotic use (Table 4).
4. Discussion
This is one of few reports empirically exploring both clinical and psychosocial features of delusional disorder. Our objective was to empirically describe sociodemographic and clinical characteristics in a large sample of DD patients and explore potential correlates of DD types.
4.1. Strengths and limitations
The main strength of this report is that, in the light of current relative absence of empirically based knowledge, it does provide a rare clinically based description of DSM-IV DD and its types. In spite of the novelty and rareness of our findings, we have to acknowledge many potential limitations that have to be considered and therefore results should be interpreted with caution particularly on those findings based on low prevalence of symptoms or risk factors. First, this is a medical record based retrospective study exploring medical records recorded by many different, yet fully trained, psychiatrists. We did not, therefore, interview patients directly or administer diagnostic scales. Instead we gathered our study data from a systematic and structured retrospective exploration of medical records. Record accuracy can vary widely across different professionals hence providing potentially biased or even partially subjective information.
However, the fact that professionals were all trained to enter their diagnostic records on an equally structured computerised medical record system does minimise recording variability. For instance, all psychiatrists had to enter a complete multi-axis DSM-IV diagnosis, including functionality axis V, and record symptoms and medications using the same options within the software. As mentioned earlier, all psychiatrists entering data in the computerised medical records system had a three-day training course to use the software and did reliability testing of their diagnostic records. Yet, even when these and other limitations exist, the relative absence of previous similar reports make our findings an array of clinically and epidemiologically unique information on DD that may lead to more thorough and methodologically sound empirical studies. A potential advantage of the study is that it may naturalistically represent fairly well a large sample of DD patients with a mean age and other sociodemographic features within those that could be expected for a late-onset disorder.
4.2. The case register
We used computerised medical records being registered at public mental health centres in a well-defined catchment area serving some 607,494 inhabitants in South Barcelona (Spain). Such procedure enabled us to build a fairly reliable clinically based case register of delusional disorder as this public mental health service network is by and large the only public mental health option for all people living in the area. Hence, family doctors in the area very rarely would not transfer patients with a disabling disorder such as DD to our public psychiatric system. Therefore, we can be reasonably confident that most cases with DD would have been actually referred to our system and were subsequently included in our register. In addition, even when all original DD cases were diagnosed by fully trained psychiatrists, we ensured the validity of DD diagnoses in cases finally included in the register by performing a thorough retrospective inspection of medical records applying a self-designed checklist (Table 1) to confirm whether patients did actually fulfil DSM-IV-RT criteria for DD.
4.3. Estimated prevalence, DD types frequencies and sociodemographic findings
We fully acknowledge that our design is inappropriate for calculating the prevalence of DD as indicated by most previous reports. Nonetheless, if we were to estimate the attended prevalence of DD in our sample, considering the population census of the areas covered by our catchment area, it would virtually double estimates as reported by previous studies (0.006%), even when such estimates are based on more selected in-patient samples [Reference Kendler20]. Such a finding may suggest that the disorder is possibly not as rare as previously thought of and also that many DD patients might not seek professional help. However, prevalence discrepancies could also reflect differences in diagnostic criteria and variability among studied populations (i.e., community based versus hospital in-patient based). Our argument, admittedly potentially speculative, is however not based on data from a poorer quality than most of those previously reported and, in the absence of better evidence, suggest the idea that DD, yet uncommon, might not be as rare as previously thought.
We found, in agreement with most previous studies [21,22,25,31], that the most prevalent DD type is the persecutory type. However, in contrast to previous reports our findings suggest that the jealous type is the second most prevalent DD type in our sample whilst the somatic type showed sensibly lower prevalence rates [21,22,31]. Nevertheless, such previous studies are based on considerably smaller sample sizes. Our study, based on both out- and in-patients with DD, shows that DD is a middle to late life psychoses which is more frequent among women (female/male ratio was 1.29). This finding is consistent with that reported by the largest study on the topic [Reference Kendler20] although smaller studies, among different populations, indeed suggest an even higher female/male ratio (female/male ratios ranging from 1.91 to 3) [21,22,25,31]. In agreement to previous reports [20,21] about half of our DD patients were married, yet a larger than expected proportion of unmarried patients suggest that DD sufferers have more difficulties to either start or maintain stable relationships.
4.4. Risk correlates of DD
Our findings suggest that DD patients may share vulnerability with schizophrenia and other psychotic patients, including an excess of DD among immigrants or among people with sensory deficits [6,11]. Interestingly, opposite to what has been suggested earlier [14–17,19,28–30], we found a high frequency of schizophrenia among relatives of our DD patients. Indeed, such frequency does not differ much from that expected among relatives of schizophrenia patients [Reference Gottesman and Irving10]. Thus, it could be argued that our results support the notion that psychotic disorders run in families possibly as a general susceptibility for a psychotic syndrome rather than for specific psychotic disorder [4,7]. Family history of schizophrenia is significantly more frequent among jealous DD type, in line with descriptions of family patterns of jealous behaviour [Reference Vauhkonen27]. Patients with persecutory type tended to use more atypical antipsychotics, a finding supporting that of a previous study [Reference Bömer and Brüne3].
4.5. Psychopathological correlates
Depressive symptoms were common in our DD patients in line with results reported earlier [12,21,22]. Depressive symptoms were significantly more prevalent amongst persecutory and jealous DD types compared to the remaining types. Along with previous reports [13,18,22,30] DD also tends to be frequently comorbid with personality disorder, particularly those from the so-called cluster A. We report virtually parallel results to those described by Hsiao et al. regarding hallucinations [Reference Hsiao, Liu, Yang and Yeh12]. Hence, the most frequent hallucinations in our sample were auditory not prominent, olfactory and tactile. Both the higher incidence of delusions in married women and the higher rates of pathological jealousy among men might reflect some Mediterranean cultural-bound influence. The so-called Othello syndrome has, nonetheless, being described after a Shakespeare's play of a morbid type of jealousy particularly frequent in men [Reference Enoch, Ball, Enoch and Ball9].
4.6. Future research directions
Our descriptive study may open up future research lines such as exploring the role of neuropsychological testing in better defining the very concept of DD. Similarly, a thorough psychopathological study is needed to rely more on empirical presence of symptoms in real patients than in a priori categorical descriptions. Finally, genetics or neurophysiological studies may also help to find endophenotypes to better describe DD and other psychotic categories.
Acknowledgements
We would like to thank all Sant Joan de Déu-Serveis de Salut Mental psychiatric staff for their collaboration in getting trained and completing clinical data in the computerised HCI medical records systems. We also thank the hard work by Francisco Mochon in designing the HCI and Marisol Novio in monitoring and extracting from HCI some of the data used for this study. The study has been partially funded by a Spanish Grant from the Fondo de Investigaciones Sanitarias (FIS: PI021813). This is a collaborative study of several research centres included in the CIBER Salud Mental of the Spanish Ministry of Health.
Comments
No Comments have been published for this article.