Background
Men and women show differences in the age distribution of post traumatic stress disorder (PTSD) prevalence during their lifespan [
1]. Although this is supported by a large and thorough epidemiological study, The National Comorbidity Survey (NCS), the finding is limited by the fact that it only involves participants at the age of 15 to 54 years. This must be regarded as a considerable limitation. The average age of living has been increasing in the modern Western world for more than 200 years [
2] and passes far beyond the age of 54. It therefore seems reasonable to include a broader range of age when estimating the lifespan distribution of PTSD among men and women. The latest updates show that men living in a country within the European Union (EU) have a life expectancy of approximately 76 years, whereas, women have a life expectancy of nearly 82 years [
3]. Therefore, individuals now live for an increased number of years compared to that of previous generations; however, as a result individuals also have more years in which they can experience traumatic events or be affected by the potentially negative consequences that follow traumatic experiences.
It therefore becomes important to pay attention to the risk of PTSD in relation to different stages in the lifespan. This will aid in the progression of age adjusted assessment and treatment methods as well as improving the individual coping strategies of PTSD. Men and women show differences in the biological aspects of brain development, thus differences in behavioural development throughout the lifespan could influence the way risk factors or trauma exposures are met [
4]. Additionally, gender has been found to be an important biological determinant of vulnerability to psychosocial stress [
5]. Therefore, focusing on the lifespan distribution of PTSD makes sense when it accounts for gender specific developmental details, and when it combines the effect of gender and age on PTSD.
Gender and age often function as demographic variables in PTSD or trauma research and as such they are both very commonly studied. However, for a great deal of studies neither age nor gender differences are the main area of focus. Numerous gender studies have been conducted with regards to PTSD. The main findings regard the fact that women, although less subjected to potentially traumatic events, develop PTSD more often than men [
6‐
8]. Other studies have reported a twofold increase in PTSD prevalence for women compared to men [
9]. Speculations have been made that the increased risk of PTSD among females is due to the higher likelihood of females to experience specific trauma types that appear to be particularly traumatic or PTSD inducing [
10]. However, it has been reported that the increased prevalence of PTSD in women remains even when trauma type is controlled for [
9]. Indications have been made that different trauma types show variations in the extent of gender differences in PTSD prevalence and as such gender shows variation in its effect on PTSD according to trauma type [
11].
Fewer age studies than gender studies are represented in the PTSD literature. Thompson, Norris, and Hanacek [
12] examined age differences in the psychological consequences of Hurricane Hugo and found that younger people exhibited the most distress in the absence of a disaster, whereas, middle-aged people exhibited most distress in the presence of a disaster. It is obvious to think that increased levels of distress are coherent with an increased risk of PTSD. Norris
et al. [
13] examined the effects of age on PTSD in a cultural context, and compared the effects of age after similar disasters in three different parts of the world. The findings showed no consistent effect of age on PTSD. Therefore, it was concluded that PTSD depended upon the social, economic, cultural, and historical context of the disaster-stricken setting more than it depended on age. They found inconsistent results among respondents from the USA, Mexico, and Poland, where the most distressed were the middle aged, the young, and the old, respectively. Thus, the age differences in PTSD prevalence tend to show some cultural variance.
In one of the most comprehensive epidemiological studies of PTSD conducted to date, the NCS [
1], results concluded both gender and age differences in PTSD. The results pertaining to gender differences concluded that women were approximately twice as likely as men to develop PTSD during their lifetime. This finding has since become well established with subsequent studies reaching similar conclusions [
6,
9]. Interestingly, Kessler and his colleagues concluded no age differences in lifetime rates of PTSD for males across different age groups. However, for women it was suggested that when age increases PTSD rates tend to decrease [
1]. The results showed that combining gender and age leads to further information about the prevalence of PTSD. Among the male participants the prevalence of PTSD was highest from their mid 40s to their mid 50s, whereas, the female participants showed the highest prevalence of PTSD from their mid 20s to their mid 30s. The results regarding PTSD prevalence are shown in Table
1.
Table 1
PTSD prevalence estimates from nationally representative studies
Age, years: | | | | |
15 to 24 | 2.8 | 10.3 | 3.8* | 5.9* |
25 to 34 | 5.6 | 11.2 | 2.5 | 4.6 |
35 to 44 | 5.0 | 10.6 | 2.0 | 3.1 |
45 to 54 | 7.6 | 8.9 | 2.2 | 3.7 |
55 to 64 | NA | NA | 2.0 | 1.5 |
65+ | NA | NA | 0.4 | 0.0 |
Total | 5.0 | 10.4 | 2.0 | 3.2 |
From the findings of previous research it appears that men and women have different developmental distributions of PTSD during their lifetime. Although the findings by Kessler and colleagues [
1] are interesting in this regard, they are still, as mentioned above, limited by the fact that the study only included data on participants between 15 and 54 years of age. Therefore, the study did not include the age extremities of childhood or late life. The relevance of including childhood in the developmental distribution of PTSD during lifetime may seem superfluous or controversial for several reasons. However, the inclusion of childhood would essentially comprise the basis of comparison among the age groups because of obvious biological and psychological differences between children and adults, which may be regarded as important for the perception of the trauma and coping strategies. The inclusion of the age extremities beyond the age of 54 seems otherwise relevant especially with an increasing tendency for the average person to reach old age.
Another comprehensive epidemiological study based on data from the Australian National Survey of Mental Health [
14] included participants beyond the age of 54. This study found incoherence for PTSD prevalence rates and exposure to trauma across the lifespan. Results showed that PTSD prevalence reduced with age for both men and women, whereas, a nearly symmetrically inverted U-shaped curvilinear pattern of lifetime exposure to trauma across the lifespan was found for women and a linear increase in lifetime exposure to trauma was found for men. Both male and female participants above the age of 65 reported negligible rates of PTSD. Women showed a higher level of PTSD prevalence in young age and in adulthood compared to men. This effect was seen until the mid 50s where men started to show a higher PTSD prevalence than women. The results of PTSD prevalence among men and women from the study are shown in Table
1. The findings suggested that the highest rates of PTSD prevalence among both men and women are found between the age of 18 and 24 years and the lowest among older people [
14]. However, it is important to note that the study only included participants above the age of 18. Some evidence points to the fact that potentially traumatic events as well as the risk of developing PTSD are as much a part of adolescence as it is part of adulthood [
15]. Interestingly, the tendency of PTSD prevalence rates declining from young age to old age follows the clinical picture found for PTSD in Danish normative data for the Millon Clinical Multiaxial Inventory III (MCMI-III) [
16]. Here, a linear decrease in PTSD prevalence rates according to age was found. This study also concluded a significantly higher score for women compared to men with regards to PTSD.
The finding of low PTSD prevalence in older people is consistent with some studies [
17] but inconsistent with others [
18,
19]. Maercker
et al. [
18] found a substantially higher prevalence of PTSD among participants in the age range of 60 to 93 years compared to the participants below 60 years of age. Thus, the results showed a linear increase in the prevalence of PTSD. However, the increase in prevalence of PTSD among older people could to a large extent be explained by World War II trauma, making the results interesting but also less representative and comparable to populations from non-World War II countries or countries less involved in the war. Elklit and O'Connor [
20] examined the occurrence of PTSD in a Danish population sample of older people who had been bereaved. They found that 27% met all four core criteria for PTSD 1 month after losing their spouse; this number was reduced to 17% 6 months after the loss. Findings showed that an additional 16% met a subclinical level of PTSD (missing one criterion) 1 month after the loss. This number had increased to 28% after 6 months. The study did not include gender-related findings. However, elsewhere suggestions have been made that increased age is a bigger risk factor of PTSD for men than it is for women [
21], and that younger age significantly predicts PTSD in women but not in men [
22]. This is concordant with findings indicating that the mean onset age of PTSD is higher among men than among women [
23].
Although the existence of a combined effect of gender and age on PTSD rates is implied by various studies, the results are ambiguous and the differences in lifespan distribution of PTSD among men and women remain unclear. It is the aim of the present study to expand previous research by including a larger number of participants 55 or older and examining the differences in lifespan distribution among men and women, respectively, along with the possible combined effect of gender and age on PTSD prevalence in order to clarify the extent and consequences of such an effect. Knowledge of the lifespan distribution of PTSD could contribute to the aetiology or phenomenology of PTSD. Furthermore, knowledge of such could be beneficial in relation to the assessment or treatment of PTSD. With the previous findings in the PTSD literature in mind, we find it relevant to examine the following hypotheses concerning the lifespan distribution of PTSD prevalence: (a) the lifespan distribution of PTSD will be different for men and women; (b) women will at all ages show a higher prevalence of PTSD than men; (c) men will show their highest level of PTSD prevalence later in life compared to women.
Methods
Procedure and participants
The criteria for including studies were: (a) the study included both male and female participants; (b) the Harvard Trauma Questionnaire (HTQ) was used for assessment in the study and thus could be a measure of comparison for the included studies. All studies that did not fulfil the abovementioned criteria were excluded from the study analysis. In addition, the participants (a) should have notified their gender; (b) be between 13 and 80 years of age; and (c) have given full information on the HTQ. Participants who did not fulfil these criteria were excluded.
Two sets of data were made for analysis. A total sample, which included participants from all the studies both epidemiological population samples and convenience samples of different trauma events, and a trauma sample, including only the participants from the convenience samples of different traumatic events within five trauma types; disasters and accidents, loss, malignant disease, non-malignant disease, and violence.
The data for the total sample consequently consisted of data from 25 different studies of trauma and PTSD that were conducted between 1996 and 2008 (Table
2). The final sample was composed of 6,548 participants, 2,768 (42.3%) men and 3,780 (57.7%) women. The age distribution of the participants ranged from 13 to 80 years of age. Of the included studies, 20 were carried out in Denmark, 4 in Iceland, and 1 in the Faroe Islands.
Table 2
Convenience and epidemiological samples included in the present study
Disaster and accident: | | | |
Earthquake victims | 33 | 40 | 73 |
Explosion affected residents | 226 | 235 | 461 |
Rescue personnel dealing with explosion | 397 | 28 | 425 |
Whiplash victims | 296 | 1,131 | 1,427 |
Violence: | | | |
Assault victims | 138 | 50 | 188 |
Knife homicide at a Danish gymnasium | 107 | 172 | 279 |
Robbery victims | 20 | 34 | 54 |
Malignant or severe disease: | | | |
Families with chronically ill children | 32 | 53 | 85 |
Parents of chronically ill children | 147 | 312 | 459 |
Non-malignant disease: | | | |
Adolescent and young adults surviving childhood cancer | 19 | 25 | 44 |
Cleft surgery patients | 18 | 4 | 22 |
Overweight persons | 15 | 141 | 156 |
Paraplegics | 147 | 69 | 216 |
Parents of prematurely born children | 18 | 40 | 58 |
Stutterers | 22 | 6 | 28 |
Loss: | | | |
Older people who have been bereaved (pilot study) | 20 | 38 | 58 |
Older people who have been bereaved (new study) | 248 | 314 | 562 |
Parents who have lost an infant (hospital) | 44 | 55 | 99 |
Parents who have lost an infant (parent association) | 264 | 353 | 617 |
Youth samples: | | | |
Danish national youth probability sample | 145 | 132 | 277 |
Faroese youth population total sample | 217 | 242 | 459 |
Icelandic national youth probability sample | 73 | 80 | 153 |
Students: | | | |
Social and Health Care College Students | 37 | 83 | 120 |
Others: | | | |
Control group from the study of parents who have lost an infant | 21 | 25 | 46 |
Trauma clients in treatment | 64 | 118 | 182 |
Total | 2,768 | 3,780 | 6,548 |
The data for the trauma sample consequently consisted of data from 17 different convenience samples of trauma and PTSD. The final sample consisted of 4,998 participants, 2,039 (40.8%) men, and 2,959 (59.2%) women. The age distribution of the participants ranged from 13 to 80 years of age. The frequency of the 13 to 15 year olds was low in the trauma sample.
For both the total sample and the trauma sample the participants were divided into 14 different age groups with a 5-year span in age for analysis except the age group of 13 to 15 year olds, which only had a 3-year span in age. Three of the included studies were undertaken in Iceland, and the others were completed in Denmark.
All studies included met the ethical guidelines for Nordic psychologists. Studies within the hospital sector were approved by a regional Helsinki committee.
Measures
The questionnaires used for measurement varied between the individual studies. All questionnaires in the selected studies requested data about gender and age of the participants. The HTQ [
24] was used in a Danish, Icelandic or Faroese edition. The HTQ estimates PTSD diagnosis according to the
Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) [
25] and at the same time it measures the severity of PTSD symptoms. The HTQ-IV hereby permits a dichotomous assessment of PTSD. The HTQ originally contained 30 items based on the 3 subscales of PTSD concerning a potentially distressing event. The answers are scored on a four-point Likert scale (1, 'not at all'; 2, 'a little'; 3, 'quite a bit'; 4, 'all the time'). Only scale items above or equal to 3 on the HTQ were considered for a PTSD diagnosis. In all the included studies an item was added or regarding feelings of guilt for something done or omitted. Some studies also divided item 16 (sudden emotional or physical reactions when reminded of the incident) into two questions. However, this additional item was not included in the HTQ total scores used for analysis in the present study, giving a total of 31 items with a possible total HTQ score in the range of 0 to 124. A total of 16 items were related to the 3 subscales of PTSD in DSM-IV: avoidance (7 items), re-experiencing (4 items), and arousal (5 items). Mollica
et al. [
24] have reported good internal consistency, test-retest reliability, and concurrent validity. The HTQ has been used extensively in the Nordic countries [
26].
Statistical analyses
Data were analysed using SPSS V.17.0 (SPSS, Chicago, IL, USA). Statistical tests included descriptive analyses performed on the data using mean scores, standard deviation (SD), and percentages. One-way analyses of variance (ANOVAs) with descriptive statistics were performed to compare the independent variables of gender and age, and the continuous dependent psychometric variable of the HTQ total score. Both the HTQ mean score as well as a categorical PTSD score were ranked by age groups. Both the HTQ mean scores as well as the categorical PTSD scores can be seen as a way to estimate the vulnerability to PTSD. The dimensional and categorical results of PTSD were both ranked by age groups in order to find the estimated distribution of PTSD prevalence according to age.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
AE conceived the study and participated in the design of the study. DND prepared the data file, performed the statistical analysis, and drafted the manuscript. Both authors read and approved the final manuscript.