Anxiety after a fall in elderly subjects and subsequent risk of developing post traumatic stress disorder at two months. A pilot study

Abstract

Background

The psychological burden suffered by elderly subjects after a fall adds to rehabilitation difficulties. We hypothesised that Post Traumatic Stress Disorder (PTSD) may be at the root of these psychological post-fall problems. The immediate psychological state was examined after 2 months post-fall in order to determine the prevalence of anxiety and PTSD and their possible correlates.

Method

A single centre prospective survey including elderly subjects referred to an emergency department (ED) after a fall with a 2-month follow-up.

Results

30.5% of patients showed a significant level of anxiety after their fall and (26%) showed evidence supporting the presence of PTSD at 2-months. Immediate anxiety, ability to re-establish upright posture and a history of repeated falls were significant predictors of PTSD at 2-months.

Conclusion

persistent psychological problems post fall may be related to PTSD which is significantly correlated to immediate anxiety. This study can be seen as a first step in patient characterisation, and predicts who may benefit from which of the various approaches of rehabilitation.

Introduction

Fall-related morbidity is a serious public health issue in older adults (1, 2, 3) referred to emergency departments. The current literature on rehabilitation of falls mainly focuses on patients who have sustained post fall fractures and will experience various degrees of subsequent disability and a further high risk of falling (4, 5). Rehabilitation of gait and balance disorders in this group of patients is now well accepted as part of the clinical management programme for the prevention of further falls (6).

Most physicians however give their primary attention to the immediate injuries and traumatic consequences of falls, often under-estimating their psychological consequences. Nevertheless, a fall can cause significant psychological trauma to the elderly and this psychological factor makes rehabilitation more difficult because of the need to take both motor and psychological components into account.

The immediate psychological consequences are most often described as fear, self-efficacy, activity avoidance, loss of confidence, impairment of body image, insecurity (7) and in its most severe form, anxiety 8, 9. All these immediate psychological outcomes of falling can be associated with a motor component and are often grouped under the poorly-defined ‘post-fall syndrome’ 10, 11.

The progression of these symptoms is unclear. It has been reported that some patients might develop a psychomotor disadaptation syndrome (PDS) (12) or the ‘fear of falling’ syndrome (FoF) (13). However, it is difficult to predict how this syndrome will progress for each patient. Nevertheless, if we analyse the possible symptoms in the light of the definition of post-traumatic stress disorder (PTSD) according to the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM IV) (14), we can find substantial similarities. Indeed, in order for PTSD to occur the subject must have been the victim of a traumatic event and their response must include intense fear, helplessness or feeling of horror. There is, in this definition, a subjective estimation of trauma intensity, and the focus is on the patient's response, which allows the possibility of defining a fall as a trauma. PTSD reactions among the elderly have also been shown, in general, to be quite similar to those displayed by younger people (15). PTSD, which has been studied in relation to a broad range of traumatic events, has rarely been studied specifically in relation to falls in older people (16).

In this context, we hypothesized that PTSD may be a delayed psychological outcome after falls in elderly patients. In this article, we present the results of a pilot study in which we examine the anxiety of elderly subjects post fall, after discharge and after 2 months in order to determine the prevalence of anxiety and PTSD and their possible correlates in this population.

Method

Study Design and Objectives

The study was an observational, prospective, single centre, community-based cohort survey. Our primary goal was to evaluate the anxiety of patients aged 75 years or older who had experienced a fall, at emergency department referral and after 2-months.

The study was designed and carried out by the authors, who also analysed the data; the authors take full responsibility for the data, the analysis, and the completeness and accuracy of this article.

Participant Enrolment and Follow-up

Eligible participants were men and women aged 75 years or over subsequently referred to the ED in the presence of the assistant investigator after a fall, defined as ‘any unexpected event in which the person comes to rest on the ground, floor or a lower level’ (17). Exclusion criteria were age below 75 years, if health status was not compatible with the interview or if they waived their consent for participation.

A standardised questionnaire with close-ended questions was used including biographical and social data, autonomy, mental status according to the patient or his family, risk factors and circumstances about the fall and its consequences.

At inclusion, the anxiety was evaluated using the “State-Trait Anxiety Inventory” (STAI) (18) which differentiates between temporary or emotional state anxiety and long-standing personality trait anxiety using 20 items for assessing trait anxiety and 20 for state anxiety.

The assistant investigator who gathered information about the patient's outcome and evaluated the presence of PTSD carried out the 2-month follow-up phone interview, using the PTSD Checklist Scale (PCLS) (19), a self-report questionnaire consisting of 17 items validated for French speakers (20). This scale differentiates well between patients presenting PTSD from other anxiety disorder subjects and non-clinical subjects (21). A score over 44 demonstrates good diagnostic efficiency (0.90) for PTSD when used with Motor Vehicle Accident victims (19). For this reason this same cut-off point was adopted for the study. Trait anxiety was also evaluated with the STAI.

The study did not interfere with the usual care procedures delivered to the patient. All of them gave oral informed consent in accordance with the guidelines of the local hospital ethics committee.

Statistical Analysis

Continuous variables were expressed as mean (±sd) or median (interquartile range), and qualitative variables as percentages. Characteristics of the study population were compared using Fisher's Exact test for distributions and the Chi-square test for means. For correlations between quantitative variables, Spearman's Rank correlation test was used. Significance was assumed at p<0.05. The StatView software suite version 5.0.1 (SAS Inc., Cary, NC, USA) was used for all analyses.

Results

Population

The study ran from February to April 2012; the follow-up ended in June 2012. Over this period, 158 elderly patients were referred after a fall, 52 of them were enrolled according to the inclusion criteria with 48 answering the questionnaire. The 2-month follow-up was not available for 12 patients, who were also excluded from the final analysis therefore leaving 36 patients. Twenty-three participants agreed to answer the questionnaire at 2 months; four patients died and 9 patients refused or were unable to participate (Figure 1).

Figure 1.

Study Flowchart

The age range for patients was 75 to 101 years, a third were older than 90 years and 31 (86.1%) were women. Twenty seven (75%) were living alone at home without assistance and 8.3% with home helps. The majority (89%) were considered to be in good cognitive state and 4 (11%) showed moderate decline. Eleven patients (30.5%) had a history of repeated falls. Other medical histories that could be related to falling (Cardiac, neurological, metabolic and osteoarticular factors) are shown in Table 1 with the characteristics of the population at inclusion.

Table 1.

Patient characteristics and anxiety measures at inclusion (n=36)

Age in years, mean (sd)	87.3 (6.7)
Sex (Female), number (percent)	31 (86.1)
History of repeated falls, number (percent)	11 (30.5)
Good cognitive condition, number (percent)	32 (88.9)
Medical history, number (percent)
- Neurological disease	10 (27.8)
- Cardiac disorders	20 (55.6)
- Metabolic disorders	5(13.9)
- Osteoarticular problems	21 (58.3)
Reason for the fall, number (percent)
- Accident	25 (69.4)
- Fainting Fit	8 (22.2)
- Unknown aetiology	3 (8.3)
Score STAI mean (sd)
- Trait anxiety	43.2 (8.4)
- State anxiety	46.4 (14.0)
Degree of anxiety, number (percent)
- The subject shows greater than normal anxiety	11 (30.5)
- The subject feels comfortable in his immediate environment	7 (16.7)
- Anxiety is not greater than usual	18 (52.8)

For 25 patients (69.4%) the fall was extrinsic i.e. accidental, 8 (22.2%) patients had suffered a fainting fit and for 3 (8.3%) of them the fall was of unknown aetiology.

Concerning the consequences of the fall, 8 patients (22.2%) were unable to get up on their own without assistance and 16 (44.4%) needed help to recover. Eleven patients (30.5%) suffered minor injury (haematoma, minor wounds or pain), 16 (44.4%) patients had severe trauma (hip fracture, other fractures, head trauma…) and one suffered an infectious complication whereas only 8 (22,2%) had no complications. Sixteen (44.4%) were hospitalised, of whom 19% required surgery.

Anxiety at inclusion and at 2-months

According to the STAI-State, 30.5% of patients had a significant level of anxiety after their fall. The anxiety characteristics of the population are presented in Table 1. The mean PCLS score at 2 months was 31.7 ± 11.8. Six patients (26%) had a score over 44, suggesting PTSD. As expected, there was no significant difference between the STAI-trait mean score at baseline and at 2 months (43.2 ±8.4 versus 44.0 ±8.2).

We observed a significant correlation between the STAI-state score at baseline and the PCLS score at 2 months (r = 0.45, p<0.05); confirmed by a tendency for a higher average score in the PTSD group (54.8 ± 12.5 versus 45.9 ± 13.2).

A trend for a negative correlation between the PCLS score and age was found (r = -0.2), supported by a lower age of patients with PTSD (82.0 ± 4.6 versus 87.8 ± 6.1, p = 0.05). No significant differences were found according to gender or cognitive status.

There was a significantly higher PCLS score in the patient group of with a history of repeated falls (28.6 ±9.4 versus 43.0 ± 13.6; p<0.01) confirmed by a significantly greater number of patients with a history of repeated falls in the group of patients with PTSD. No significant differences were found according to other medical conditions considered as risk factors.

Focussing on the consequences of the fall, we were unable to find a significant difference depending on whether the patient was hospitalised or not after their fall, or whether or not there were severe traumatic consequences of the fall.

In contrast, a significantly higher PCLS score was found for patients able to recover on their own compared to recovery with assistance (36.6 ± 12.0 versus 24.1 ± 6.2).

Discussion

This study provides evidence for the development of PTSD in 26% of a population of elderly patients, two months after a fall. Chung et al., found a 25% rate of full or partial PTSD in their study (16) but were unable to show any pattern of factors, at baseline, predictive of PTSD. Our major results are that immediate post fall anxiety, a history of repeated falls and ability to get up from the ground unassisted are significant predictors of PTSD at 2-months. Concerning post fall anxiety, our results replicate those of Bui et al. (22) in a population of elderly subjects, one month after a discharge in an ED where traumatic distress was evaluated, whatever the trauma and not only post fall.

The association between a history of repeated falls and development of a PTSD at 2-months in our study seems more complex as it is difficult to determine if the psychological disorder is connected to the recent fall or to a previous one resulting in the emergence of a syndrome of fear of falling, which itself is correlated with anxiety (23).

The most likely explanation for the higher PCLS score among patients who were able to get up unassisted can be that a fall in these patients, often in better health (24), can be a more disturbing sign, causing greater anxiety. However, we were unable to demonstrate any other correlations between baseline patient characteristics or initial consequences and subsequent development of PTSD.

The tendency for a negative correlation between PCLS score and age may seem unexpected but is in accordance with a previous similar result by Chung et al. (16) who showed a significant correlation of -0.18 between age of patients and their PTSD score at four months. Similarly, when looking at the general population Parto et al. (25) showed that younger subjects showed more frequent positive screening score for PTSD than the elderly.

We acknowledge that conclusions of our study are limited by the sample size. Including patients in this study was difficult even though our criteria were not very selective. This is partly explained by the organisation of the ED, which leaves little space to implement this sort of questionnaire, considered as a non-priority action in the case of high workloads. Nevertheless, we were able to contact 70% of the patients at two months, which is a good percentage of follow-up.

In our population, most patients seem to have normal mental status and no dependency. These data could lead to the assumption, that the sample is highly selective. Nevertheless they were very similar to a previous but larger sample of fallers recruited in our hospital (26). We assume that it may be comparable the ED recruitment of any urban area in industrialised countries and reflect a small part of the elderly population suffering falls. Further extrapolation of these results would be very difficult, but the present study gives weight to the setting up of a multi-centre study to confirm these observations.

Our study suggests that a substantial minority of older people have symptoms of full or partial PTSD following a fall. All these arguments support the idea that full or partial PTSD may be a delayed consequence of fall. Overall, we are faced with different forms of anxiety that will develop differently: simple anxiety or more complex states assimilated to a post fall syndrome. These symptoms can resolve spontaneously; develop into a true PTSD or a FoF and will require a variety of different types of rehabilitation. It is important to precisely characterise each case presented, so as not to miss PTSD. This could then lead to a failure of adequate rehabilitation.

The next step will be to explore and test hypotheses as to how we can improve treatment, determine when to treat patients and what patient characteristics may predict who is likely to complete and benefit from various rehabilitation approaches. We should be able to choose several techniques integrating or not psychological care strategies, which could take several forms. We could, for example, test the use of virtual reality techniques for rehabilitation, whose efficacy as been proven in post-war PTSD.

Disclosure statement: the authors declare no conflict of interest.