Open Journal of Depression
2013. Vol.2, No.2, 7-10
Published Online May 2013 in SciRes (http://www.scirp.org/journal/ojd) http://dx.doi.org/10.4236/ojd.2013.22002
Could Resilience Predict the Outcome of Psychiatric
Rehabilitation Patients?
——A Pilot Study*
Birk Engmann#
Department of Neurology and Psychosomatics, Brandis Rehabilitation Centre, Am Wald, Brandis, Germany
Received March 12th, 2013; revised April 17th, 2013; accepted April 25th, 2013
Copyright © 2013 Birk Engmann. This is an open access article distributed under the Creative Commons Attri-
bution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the
original work is properly cited.
Background: Individual resistance to burdens such as stress, adversity, and even disasters is called “resil-
ience”. Whereas most studies of resilience investigate post-traumatic stress disorder, fewer studies treat
anxiety and depressiveness in mentally healthy populations. The present pilot study investigates whether
resilience lessens the severity of depressive symptoms. Aim was to figure out whether further research is
needed on the relation of resilience both to depression and to rehabilitation from it in regard to fitness for
work. Design/Setting/Population: A retrospective, non-blind, non-randomized analysis of charts of 503
stationary inpatient-patients was performed. Patients without age limits who suffered from diseases of the
depressive spectrum (F32, F33) and also adjustment disorder (F43.2), dysthymia (F34.1), and neuras-
thenia (F48) according to ICD 10 were included. BDI and resilience scale, state of fitness or unfitness for
work, were analysed with correlation analysis and descriptive statistics by the SPSS program. Results:
Analysis revealed that resilience and depressive symptoms are inversely related. The higher the resilience
the lower the BDI score, indicating a lower burden of depressive symptoms. Resilience also seems to in-
fluence important outcome factors of rehabilitation such as reintegration in work and remission of depres-
sive symptoms. Patients who either were fit for work at the beginning of rehabilitation or were dismissed
as fit for work had higher resilience scores than those who were unfit for work. Conclusion: This pilot
study encourages further investigation of the relation of resilience not only to depressive disorders, but
also to rehabilitation and social reintegration after depression.
Keywords: Resilience; Rehabilitation; Depression; BDI
Introduction
Psychic diseases are being diagnosed more often, and be-
coming more prominent in absenteeism from work. Therefore,
investigating factors which affect rehabilitation from psychic
disease may help to assess outcome. Resilience could be one of
such factors. According to Rutter (2006), “resilience is an in-
teractive concept that refers to a relative resistance to environ-
mental risk experiences, or the overcoming of stress or adver-
sity. As such, it differs from both social competence and posi-
tive mental health. Resilience differs from traditional concepts
of risk and protection in its focus on individual variations in
response to comparable experiences”. Others such as Luthar et
al. (2006) and Cicchetti (2010) emphasize an additional influ-
ence of neurobiological and genetic factors.
Most of the studies focused on the relationship of posttrau-
matic stress disorder and resilience, such as Mealer et al. (2012)
who showed that nurses with higher resilience could better cope
with circumstances at intensive care units. Also, coping with
disasters such as bombings was investigated. According to a
study of disaster survivors in North and Cloninger (2012), de-
pression was not a robust marker of low resilience in a study of
disaster survivors. Contrary to that, occurrence of posttraumatic
stress disorder was associated with low resilience. A similar
finding is in Ahmand et al. (2010), which points out that earth-
quake survivors with higher resilience scores had less post-
traumatic symptom levels. Peres et al. (2011) investigated po-
lice officers with traumatic memories by fMRI scan. Policemen
with high resilience, or those who were in a psychotherapy
group, had decreased amygdala activity during traumatic mem-
ory retrieval in comparison to those without current psycho-
therapy.
Fewer studies dealt with the question of how depressive dis-
orders relate to resilience. Hjemdal et al. (2011) compared re-
silience with anxiety and depression scales in psychically
healthy teenagers in highschools in Norway. The authors show-
ed that higher resilience is connected with lower scores for
depression, anxiety, stress and compulsory behaviour. Pierini
and Stuifbergen (2010) investigated resilience factors and de-
pressiveness in patients with post-polio syndrome. They found
spiritual growth as the main predictor of lesser depressive
symptoms. A Korean study compared a healthy population with
psychiatric outpatients. “Greater resilience was found to be
associated with less perceived stress, anxiety and depression
*Conflict of interest: The author declares that there is no conflict of interest.
#Corresponding author.

B. ENGMANN
and with higher levels of positive affect and purpose in life”,
according to the authors (Jung et al., 2012).
In somatic diseases, resilience also improves outcome as
shown in a study by Strauss et al. (2007) of cancer patients with
fatigue who received radiation therapy. Furthermore, elderly
people with higher resilience have fewer subjective body com-
plaints, as reported in a study by Leppert et al. (2005).
All in all, previous studies with regard both to psychic and
somatic symptoms showed the beneficial influence of resilience
for a lower burden of psychic symptoms.
The present study investigates following hypotheses:
1) Patients with high resilience have lower BDI score at be-
ginning of rehabilitation.
2) Patients with high resilience have more remissions of de-
pressive symptoms within 3 weeks of starting rehabilitation
than patients with low resilience.
3) Patients with high resilience are more likely to be classi-
fied as “fit for work” after rehabilitation.
Methods
This is a pilot study to assess further needs and possibilities
for investigation. A retrospective, non-blind, non-randomized
analysis of charts of stationary patients, i.e. only inpatient, no
outpatient or daycare rehabilitation, from August 2011 to De-
cember 2011, was performed. In total, 503 cases fulfilled crite-
ria for inclusion and exclusion (see below). The duration of
rehabilitation lasted from 3 weeks to 5 weeks. All patients re-
ceived multimodal treatment of both psychotherapy in one-
to-one and group conversations, and furthermore, physio-
therapy.
Included in the analysis were all diseases of the depressive
spectrum (F32, F33) and also adjustment disorder (F43.2),
dysthymia (F34.1), and neurasthenia (F48) according to ICD 10.
Excluded were depressive states in bipolar disorders (F30, F31),
schizophrenia (F20-F29), alcohol and drug addiction (F10-F19),
brain diseases such as dementia or Parkinson’s.
In Brandis hospital, BDI and resilience scale are standard
tests for all patients. Remission or response in BDI is used as an
outcome marker of a successful rehabilitation.
The following assessment criteria of Beck Depression In-
ventory (BDI) were used:
BDI ≤ 10: no depressive symptoms.
BDI = 11 - 17: mild to moderate depressive symptoms.
BDI ≥ 18: clinically relevant depressive symptoms (Haut-
zinger et al., 1995).
Remission is defined as BDI ≤ 10. Response is defined as
BDI score reduction of 50%.
BDI’s were filled out at beginning of rehabilitation and after
3 weeks.
For the assessment of resilience, the 11-item resilience scale
RS-11 was used. This is a short version of 25-item resilience
scale of Wagnild and Young (1993). It was modified and trans-
lated into German by Schumacher et al. (2005). Minimal score
which stands for lowest resilience is 11 (1 point for each item),
highest score is 77 (7 points for each item) which stands for a
very high resilience. Patients filled out this resilience scale at
beginning of rehabilitation.
Statistical analysis was conducted with SPSS.
Results
34.4% (n = 173) of the patients were male, 65.6% (n = 330)
were female (total: n = 503).
Minimum age was 19 years, maximum age 84 years. Mean
age 51.8 years, median age 52.0 years (SD: 11.2). During reha-
bilitation psychiatric medication was altered in 21.8% (n = 110)
of the subjects. Medication was stable in 77.8% (n = 393). No
data in 0.4% (n = 2).
Is there a correlation between resilience and BDI score at be-
ginning of rehabilitation? In 448 of 503 cases all data were
available. Pearson correlation coefficient is −0.628 (p <
0.0001).
The graph (Figure 1) shows a good reverse correlation be-
tween resilience score and BDI at beginning of rehabilitation.
Another question is “does resilience influence duration of
absenteeism from work by sick note?”
At the beginning of rehabilitation (Figure 2), 46.5% (n = 235)
of the patients were absent from work by sick note, and only
32.3% (n = 163) were fit for work. Others were old age and
disability pensioners (20.8%, n = 105).
At dismissal from rehabilitation 48.9% (n = 247) were still
counted as being unfit for work, and 29.9% were dismissed as
fit for work (pensioners 20.8%, n = 105).
Before rehabilitation, the longest period of being unfit for
work by a sick note was 1540 days, the minimal was 2 days,
and the mean was 212.5 days, and the median was 180.0 days
(SD: 161.7). That means that the average amount of time that a
person was unfit for work by a sick note before rehabilitation
was approximately 4 months!
Is there a relation between fit or unfit for work and res-
ilience?
First, resilience scores were tested for normal distribution by
Kolmogorov-Smirnov-Test (2-tailed p = 0.370; near by alpha =
0.4) which was fulfilled.
The mean resilience score in the whole sample (448 patients,
57 missing) is 46.8 (median 47, SD: 15.5).
Patients who came to rehabilitation unfit for work had a mean
resilience score of 43.1 (median 43.5, SD: 14.8; n = 214, miss-
ing resilience score in 21 subjects). Contrary to that, patients who
came fit for work to rehabilitation had a mean resilience score of
51.1 (median 52.0, n = 145, SD: 14.6; missing resilience score
in 18 subjects).
BDI intitial
6050403020100
initia l re s i lien c e score
80
70
60
50
40
30
20
10
0
Figure 1.
Correlation between BDI and resilience score at beginning of rehabili-
tation.
8

B. ENGMANN
unfit for work at admission
pensioner or othersnoyesMissing
Count
300
200
100
0
Figure 2.
Status of being fit or unfit for work at the beginning of rehabilitation.
Similar findings exist for dismissal. In patients who were re-
leased unfit for work from rehabilitation mean resilience score
was 42.97 (median 43.0, n = 223, SD: 14.8; missing resilience
score in 14 subjects). Patients who were released fit for work
had a mean resilience score of 51.99 (median 52.50, n = 136,
SD: 14.1; missing resilience score in 15 subjects).
In conclusion, there is a trend that patients who either came
fit for work or left rehabilitation fit for work had higher resil-
ience scores.
Does resilience predict outcome of depressive symptoms?
Patients who had a remission after 3 weeks rehabilitation had
a mean resilience score of 53.56 (median 54.0, SD: 12.3; n = 61,
0 missing).
Patients who had a response but no remission after 3 weeks
rehabilitation had a mean resilience score of 40.11 (median
39.0, n = 17, SD: 7.7; 0 missing).
Patients who had neither remission nor response (“no effort”)
had a mean resilience score of 41.77 (median 42.0, n = 252, SD:
14.4; 4 missing).
In conclusion, a trend could be described that patients with
remission are accompanied by higher resilience compared to
patients with response or “no effort”.
Discussion
The mean resilience score (46.8) in the whole sample is
lower than the medium resilience score of a healthy population
(58.03 according to Schumacher et al., 2005).
The study revealed a good reverse correlation between resil-
ience and BDI. The higher the resilience the lower the BDI
score, indicating a lower burden of depressive symptoms.
Two outcome items were investigated in relation to resilience:
being fit or unfit for work and remission of depressive symp-
toms. There is a trend that patients who either were fit for work
at the beginning of rehabilitation or were dismissed as fit for
work had higher resilience scores that those who were unfit for
work. The same is for patients with higher resilience who were
more likely to have a remission than those with lower resilience.
This finding could only be described as a trend because number
of subjects in subgroups differs to a high degree and thus could
lead to bias.
Patients with depressive disorders in rehabilitation have al-
ready been absent from work approximately 4 months before
the beginning of rehabilitation. This raises a question of when
symptoms may be considered chronic.
It is interesting that there is no major effect of rehabilitation
in relation to whether patients are fit or unfit for work before
and after rehabilitation and the outcome is counterintuitive, in
that there was a slight increase of patients judged unfit for work
after rehabilitation. Perhaps this outcome may be explained by
differences between the physicians responsible for admitting
the patient and the physicians of the rehabilitation clinic, in that
they may have different criteria for being fit for work. Another
question is if different approaches of involved physicians affect
ICD-10 criteria and thus inclusion criteria. Other limitation
factors could be the relatively small sample size, or the severity
of the disease which goes with prolonged absenteeism from
work before rehabilitation. Furthermore the study has no fol-
low-up to assess the duration of absenteeism after dismissal nor
does it contain information about how many patients had been
included in a gradual reintegration program (called “stufen-
weise Wiedereingliederung” or “Hamburg model”) in the old
working place. This program deems one unfit for work until
one achieves a full number of working hours. Such a program
avoids the oversimplification of defining the value of rehabili-
tation entirely by results present at the time of dismissal.
Another factor which weakens the results of the present
study is that psychiatric medications were altered during reha-
bilitation. In addition to it, even cases with no alteration during
rehabilitation could benefit from delayed effects of a medica-
tion which had been started before rehabilitation. Moreover,
patients did not receive a completely “superposable” psycho-
logical and physical treatment. Also, different age groups could
lead to a bias in the study.
In conclusion, the present pilot study encourages further in-
vestigation of the relation of resilience to depressive disorders,
and rehabilitation from them, and reintegration in social life.
Acknowledgements
Professor Dr. Wilfried Mau, Institute of Rehabilitation Me-
dicine, University of Halle, Germany, and Professor Dr. Ger-
hard Schmid-Ott, Berolina Clinic, Bad Oeynhausen, Germany,
for introduction to the topic of resilience.
REFERENCES
Ahmand, S., Feder, A., Lee, E. J., Wang, Y., Southwick, S. M.,
Schlackman, E., Buchholz, K., Alonso, A., & Charney, D. S. (2010).
Earthquake impact in a remote South Asian population: Psychosocial
factors and posttraumatic symptoms. Journal of Traumatic Stress, 23,
408-412.
Cicchetti, D. (2010). Resilience under conditions of extreme stress: A
multilevel perspective. W orld P s y c h i at r y , 9 , 145-154.
Hautzinger, M., Bailer, M., Worall, H., & Keller, F. (1995). Beck-
Depressionis-Inventar (BDI) . Testhandbuch (2nd ed.). Bern: Hans
Huber.
Hjemdal, O., Vogel, P. A., Solem, S., Hagen, K., & Stiles, T. C. (2011).
The relationship between resilience and levels of anxiety, depression,
and obsessive-compulsive symptoms in adolescents. Clinical Psy-
chology & Psychotherapy, 18, 314-321. doi:10.1002/cpp.719
Jung, Y. E., Min, J. A., Shin, A. Y., Han, S. Y., Lee, K. U., Kim, T. S.,
Park, J. E., Choi, S. W., Lee, S. H., Choi, K. S., Park, Y. M., Woo, J.
M., Bhang, S. Y., Kang, E. H., Kim, W., Yu, J. J., & Chae, J. H.
(2012). Positiveness research team of Korea: The Korean version of
the connor-davidson resilience scale: An extended validation. Stress

B. ENGMANN
10
Health, 28, 319-326. doi:10.1002/smi.1436
Leppert, K., Gunzelmann, T., Schumacher, J., Strauss, B., & Brähler, E.
(2005). Resilience as a protective personality characteristic in the
elderly. Psychotherapie, Psychosomatik, Medizinischle Psychologier,
55, 365-369. doi:10.1055/s-2005-866873
Luthar, S. S., Sawyer, J. A., & Brown, P. J. (2006). Conceptual issues
in studies of resilience: Past, present, and future research. Annals of
the New York Academy of Sciences, 1094, 105-115.
doi:10.1196/annals.1376.009
Mealer, M., Jones, J., & Moss, M. (2012). A qualitative study of resil-
ience and posttraumatic stress disorder in United States ICU nurses.
Intensive Care Medicine, 38, 1445-1451.
doi:10.1007/s00134-012-2600-6
North, C. S., & Cloninger, C. R. (2012) Personality and major depress-
sion among directly exposed survivors of the Oklahoma city bomb-
ing. Depression Research and Treatment, 2012, Article ID: 204741.
Peres, J. F., Foerster, B., Santana, L. G., Fereira, M. D., Nasello, A. G.,
Savoia, M., Moreira-Almeida, A., & Lederman, H. (2011). Police of-
ficers under attack: Resilience implications of an fMRI study. Jour-
nal of Psychiatric Research, 45, 727-734.
doi:10.1016/j.jpsychires.2010.11.004
Pierini, D., & Stuifbergen, A. K. (2010). Psychological resilience and
depressive symptoms in older adults diagnosed with post-polio syn-
drome. Rehabilitation Nursing, 35, 167-175.
doi:10.1002/j.2048-7940.2010.tb00043.x
Rutter, M. (2006). Implications of resilience concepts for scientific
understanding. Annals of the New York Academy of Sciences, 1094,
1-12. doi:10.1196/annals.1376.002
Schumacher, J., Leppert, K., Gunzelmann, Th., Strauß, B., & Brähler, E.
(2005). Die resilienzskala—Ein fragebogen zur erfassung der psy-
chischen widerstandsfähigkeit als personmerkmal. Zeitschrift für
klinische Psychologie, Psychiatrie und Psychotherapie, 53, 16-19.
Strauss, B., Brix, C., Fischer, S., Leppert, K., Füller, J., Roehrig, B.,
Schleussner, C., & Wendt, T. G. (2007). The influence of resilience
on fatigue in cancer patients undergoing radiation therapy (RT).
Journal of Cancer Research and Clinical Oncology, 133, 511-518.
doi:10.1007/s00432-007-0195-z
Wagnild, G. M., & Young, H. M. (1993). Development and psycho-
metric evaluation of the Resilience Scale. Journal of Nursing Meas-
urement, 1, 165-178.