Alexithymia in oncologic disease: association with cancer invasion and hemoglobin levels.

Abstract

The literature suggests that alexithymia is the result of individual differences and/or biological mechanisms. Both individual differences and disease mechanisms may play a role among individuals with medical or surgical conditions. The relative weight of clinical and individual differences factors related to alexithymia has not been studied in patients with cancer. The purpose of this study was to examine the extent to which individual differences in perceived stress and biological markers of illness severity are associated with alexithymia among patients with cancer.
Treated oncologic outpatients (N = 37) were assessed using the 20-item Toronto Alexithymia Scale and Perceived Stress Scale. Alexithymia was examined in relation to perceived stress, tumor staging, and hemoglobin levels.
Among the patients studied, 34.2% endorsed the established cutoff score (≥61) for alexithymia. Higher alexithymia scores were found in patients with more advanced stages of cancer invasion. Alexithymia correlated directly with perceived stress and indirectly with hemoglobin levels. Hemoglobin levels and cancer invasion significantly correlated with alexithymia when controlling for perceived stress.
A significant component of alexithymia in cancer patients may be construed as acquired. But awareness of health status influencing perceived stress might partially mediate the role of cancer invasion and hemoglobin on alexithymia.

Full text

Page 1

AnnAls of CliniCAl PsyChiAtry
AACP.com
Annals of Clinical Psychiatry | Vol. 23 No. 2 | May 2011
125
AnnAls of CliniCAl PsyChiAtry 2011;23(2):125-130
ReseaRch aRticle
Background: The literature suggests that alexithymia is the result of
individual differences and/or biological mechanisms. Both individual
differences and disease mechanisms may play a role among individuals
with medical or surgical conditions. The relative weight of clinical and
individual differences factors related to alexithymia has not been stud-
ied in patients with cancer. The purpose of this study was to examine
the extent to which individual differences in perceived stress and bio-
logical markers of illness severity are associated with alexithymia among
patients with cancer.
Methods: Treated oncologic outpatients (N = 37) were assessed using the
20-item Toronto Alexithymia Scale and Perceived Stress Scale. Alexithymia
was examined in relation to perceived stress, tumor staging, and hemoglo-
bin levels.
results: Among the patients studied, 34.2% endorsed the established cut-
off score (≥61) for alexithymia. Higher alexithymia scores were found in
patients with more advanced stages of cancer invasion. Alexithymia cor-
related directly with perceived stress and indirectly with hemoglobin lev-
els. Hemoglobin levels and cancer invasion significantly correlated with
alexithymia when controlling for perceived stress.
conclusions: A significant component of alexithymia in cancer patients
may be construed as acquired. But awareness of health status influencing
perceived stress might partially mediate the role of cancer invasion and
hemoglobin on alexithymia.
keywords: alexithymia, cancer, hemoglobin, perceived stress, personal-
ity, tumor staging
Antonino Messina, MD
Anna Maria Fogliani, MD
Scuola di Specializzazione in Psichiatria
Università di Catania
Catania, Italy
Sergio Paradiso, MD, PhD
Department of Psychiatry
University of Iowa
Iowa City, IA, USA
Alexithymia in oncologic disease: Association
with cancer invasion and hemoglobin levels
CorrespondenCe
Sergio Paradiso, MD, PhD
University of Iowa
Department of Psychiatry
200 Hawkins Drive, W278 GH
Iowa City, IA 52242 USA
e-MAIL
sergio-paradiso@uiowa.edu

Page 2

AlexithymiA in onCologiC diseAse
May 2011 | Vol. 23 No. 2 | Annals of Clinical Psychiatry
126
intRoduction
Individuals with limited introspective capacity and a
cognitive style characterized by difficulty expressing
emotions and distinguishing between emotions and
bodily sensations are commonly identified as alex-
ithymic.1 Alexithymia is associated with psychiatric
disorders including substance,2 somatoform,3 panic,4
posttraumatic stress,5 and eating disorders6; and with
medical illnesses, such as gastrointestinal dysmotility,7
hypertension,8 and inflammatory bowel syndrome.9
Alexithymia is a stable personality trait (ie, primary
alexithymia)10 but also can be acquired or secondary.
Mechanisms for acquired alexithymia may be brain
damage,11 age-related gray matter reduction in the
anterior cingulate cortex (ACC),12 or dysfunction in the
interhemispheric transfer of information (eg, resection
of the corpus callosum).13,14
Malignancies have been associated with ele-
vated levels of alexithymia.15 Whereas identifying risk
factors for alexithymia in oncologic illness may help
in targeting medical interventions and may ultimately
lead to improved quality of life,16 research in this
area has been scant. The present study represents
an initial step in identifying potential risk factors for
alexithymia in cancer patients. It was posited that
alexithymia in oncologic disease may have primary
(related to individual differences)10 and/or second-
ary (acquired or biological) mechanisms.11,17 As an
individual difference factor, the choice fell on stress
because stress is perceived differently depending on
the personality of the individual,18 and the degree to
which stress is perceived is considered a heritable
psychological factor.19 In addition, the trying life con-
ditions of individuals with cancer invariably result
in a wide range of stress responses, limiting foresee-
able floor and ceiling effects on the perceived stress
measure.
As biological factors, anemia and degree of cancer
invasion were chosen because they are clear indicators
of illness severity.20 In addition, almost all individuals
with cancer will develop some degree of anemia, with
up to 80% of patients experiencing serious hemoglobin
reduction.21-23
This study examined the extent to which alexi-
thymia among cancer patients is associated with psy-
chological (ie, perceived stress) and/or biological (ie,
anemia and tumor staging) factors.
Methods
From March 2007 to June 2007, cognitively intact out-
patients (N = 37; 20 women) with solid and hemato-
logic malignancies were enrolled from the oncology
services of the Cannizzaro and Ferrarotto Hospitals in
Catania, Italy, based solely on their willingness to par-
ticipate in the study and their ability to consent accord-
ing to the Declaration of Helsinki. Malignancy type,
staging, and hemoglobin levels were extracted from the
medical record at the time of alexithymia assessment.
Alexithymia was measured using the Italian version
of the Toronto Alexithymia Scale (TAS-20).24 The TAS-
20 is a well-validated, widely used scale that measures
alexithymia.25 Cutoff scores are as follows: ≤50, no alexi-
thymia; 51 to 60, borderline alexithymia; and ≥61, alexi-
thymia.25 Severity of depression was measured using
the Italian version of the Beck Depression Inventory-II
(BDI-II), with scores as follows: 0 to 13, minimal depres-
sion; 14 to 19, mild depression; 20 to 28, moderate
depression; and 29 to 63, severe depression.26,27
Perceived stress was measured using the 10-item
Perceived Stress Scale (PSS),28 a widely used instru-
ment that measures stress perception for the previous
month. Response options are assessed using a 5-level
Likert scale (0, never to 4, very often). PSS scores range
from 0 to 40, with higher scores indicating greater stress.
Moreover, each patient was assigned a cancer stage
according to the TNM classification, which consid-
ers size and extent of tumor (T), involvement of lymph
nodes (N), and presence of metastasis (M).
data analysis
The effects of sex and cancer type on alexithymia,
perceived stress, and hemoglobin levels were exam-
ined for descriptive purposes using ranks (data were
not normally distributed), and t and F statistics.
Hypothesis testing (ie, associations of alexithymia with
hemoglobin level and perceived stress) was carried out
using ranks, correlation, and regression techniques.
One-way analysis of variance (ANOVA) was used to
examine the effect of illness staging on ranked alexi-
thymia scores and hemoglobin levels. Least square
differences (LSD) were examined to report post hoc
significant differences when the omnibus model
was significant. All analyses were carried out using
the Statistical Package for the Social Sciences (SPSS)
version 17.0.1.

Page 3

AACP.com
Annals of Clinical Psychiatry | Vol. 23 No. 2 | May 2011
AnnAls of CliniCAl PsyChiAtry
127
Results
The patients’ age range was 32 to 75 years (mean age,
59.8; SD = 10.2). Most patients (84.2%) were married,
15.8% were single or widowed; the mean education was
10.9 years (SD = 4.8); and employment was as follows:
5.3%, primary sector; 13.2%, secondary sector; 39.5%,
tertiary sector; and 26.3%, retired. According to the
BDI-II cutoff scores, 4 participants (10.8%) had moder-
ate depression, and 14 (37.8%) had severe depression.
Patient malignancies were as follows: 13, gastroenteric;
10, breast; 6, urogenital; 5, hematologic; and 3, pulmo-
nary. Proportions of patients by TNM staging were as
follows: stage I, 5.4%; stage II, 43.2%; stage III, 37.8%;
and stage IV, 13.5%. Patients with hematologic cancer
were at stage II (80%) or III (20%); patients with gastro-
enteric cancer were at stage II (31%), stage III (46%), or
stage IV (23%); patients with breast cancer were at stage
I (20%), stage II (40%), stage III (20%), or stage IV (20%);
patients with urogenital cancer were at stage II (80%) or
stage III (20%); and all pulmonary cancer patients were
at stage III.
Prevalence of alexithymia (TAS-20 ≥61) was 34.2%.
Mean alexithymia score was 55.3 (SD = 9.7; range, 40 to
79). For women, the mean score was 54.4 (SD = 11.5)
and for men it was 56.5 (SD = 7.4) (t = .63; df = 36; P >
.5). Mean perceived stress was 21.6 (SD = 6.0), with men
showing a mean of 21.5 (SD = 5.5) and women show-
ing 21.7 (SD = 6.5) (t = –.11; df = 36; P > .4). Hemoglobin
mean values were 10.7 mg/dL (SD = 1.8; range, 8 to 14
mg/dL). Mean values for men (11.1 mg/dL; SD = 1.9)
were not different from mean values for women (10.3
mg/dL; SD = 1.6) (t = 1.4; df = 36; P > .2).
Effects of malignancy type on alexithymia, perceived
stress, and hemoglobin levels were examined for descrip-
tive purposes (TAbLe 1). A significant effect of cancer type
on alexithymia was found (F[4,32] = 3.6; P < .02), while
the effect on perceived stress just failed to meet signifi-
cance (F[4,32] = 2.6; P = .051) and there was no significant
effect of cancer type on hemoglobin level (F[4,32) = .8;
P > .4). Patients with breast cancer had the lowest alexi-
thymia scores, whereas patients with lung cancer had the
highest alexithymia scores (TAbLe 1). Patients with hema-
tologic (LSD = 12.2; P < .03) and lung malignancies (LSD
= 22; P = .001) showed significantly higher alexithymia
compared with patients with breast cancer. No other dif-
ferences were found to be statistically significant.
The influence of stage of tissue invasion on alexi-
thymia, perceived stress, and hemoglobin levels are shown
in TAbLe 2. The effect of cancer invasion on alexithymia was
statistically significant (F[3,33] = 5.8; P = .002) and showed
that alexithymia scores were higher with greater tissue
invasion (see TAbLe 2 for details on post hoc analyses). The
effect of cancer staging on perceived stress also was signifi-
cant (F[3,33] = 3.5; P = .025) (see TAbLe 2 for details on post
hoc analyses). The effect of cancer staging on hemoglobin
level just failed to reach statistical significance (F[3,33] =
2.7; P = .063). When patients with hematologic cancer (n
= 5) were excluded from the analysis, the effect of cancer
staging on hemoglobin level became statistically signifi-
cant (F[3,28] = 3.5; P = .028), confirming that stage of cancer
invasion and anemia are related indices of illness severity.
TAbLe 1
Alexithymia, perceived stress, and hemoglobin levels by malignancy type
Cancer typen
TAsa
pssb
Hemoglobinc
Meansd MeansdMean sd
Hematologic560.410.8 26.4 6.589.802.05
Gastroenteric 1356.08.621.86.72 10.621.94
Genitourinary655.5 5.8 19.22.9911.171.83
Breast 1048.510.2 19.65.8711.201.55
Lungs366.3 1.525.32.529.672.08
ToTAL 3755.359.7821.466.0010.681.83
aANOVA for TAS: F(4,32) = 3.6; P < .02.
bANOVA for PSS: F(4,32) = 2.6; P = .051.
cANOVA for hemoglobin level: F(4,32) = .8; P > .4.
Post hoc analyses: Patients with hematologic malignancies (LSD = 12.2; P < .03) and lung malignancies (LSD = 22; P = .001) showed significantly higher alexithymia compared
with patients with breast cancer.
ANOVA: analysis of variance; LSD: least squares difference; PSS: Perceived Stress Scale; SD: standard deviation; TAS: Toronto Alexithymia Scale.

Page 4

AlexithymiA in onCologiC diseAse
May 2011 | Vol. 23 No. 2 | Annals of Clinical Psychiatry
128
Hemoglobin level inversely correlated with alexi-
thymia (r = –.51; P = .001) and with perceived stress (r =
–.42; P = .005), which directly correlated with alexithymia
(r = .52; P = .001). When patients with hematologic cancer
were excluded, correlations between hemoglobin levels
and alexithymia remained essentially the same (r = –.49;
P = .004; n = 32) as well as the association between alexi-
thymia and perceived stress (r = .50; P = .003; n = 32), but
the correlation between hemoglobin level and perceived
stress was no longer significant (r = –.27; P > .1; n = 32).
These 2 sets of analyses indicate that lower hemoglobin
levels predicted severity of alexithymia, which was also
associated with perception of stress, but in nonhema-
tologic oncologic disease, perceived stress was not pre-
dicted by anemia.
Perceived stress, cancer staging, and hemoglobin
level were then examined in a regression model predict-
ing alexithymia. The model showed an overall r = .70 (R2
= .49) (F[3,33] = 10.6; P < .0001). Standardized beta for
perceived stress was .25 (t[1,36] = 1.7; P > .10), while for
cancer staging it was .37 (t[1,36] = 2.6; P < .02), and for
hemoglobin it was –.29 (t[1,36] = –2.05; P < .05). Perceived
stress, hemoglobin, and cancer staging were also entered
independently in the hierarchical model. In this case, the
R2 for perceived stress was .29 (F[1,35] = 14.4; P = .001),
whereas the R2 changed to .38 after hemoglobin (F[1,34]
= 5.1; P < .05) was entered, and to .49 (F[1,33] = 6.9; P <
.02) after cancer staging was entered. The order in which
the independent variables were entered did not affect
the results. In summary, perceived stress, cancer inva-
sion, and hemoglobin level together accounted for about
50% of the variance in alexithymia. Hemoglobin level and
cancer staging were significant predictors of alexithymia,
independent of perceived stress.
discussion
The present study found a 34.2% prevalence of alexi-
thymia in cancer patients. This is consistent with the 36.4%
prevalence found in women with breast cancer15 and is
about 8-fold greater than in the general population of Italy
(4.5%).9 There is broad consensus that alexithymia may
stem from individual developmental differences,10 age-
related brain changes,12 or medical or surgical disease.17
In this sample of cancer patients, psychological factors (ie,
stress perception) accounted for about 30% of the variance
in alexithymia. Hemoglobin level and degree of cancer
invasion were independently associated with alexithymia,
and each accounted for an additional 10% of the variance.
In addition, cancer invasion and lower levels of hemoglo-
bin were associated with greater perceived stress. When
patients with hematologic cancer were excluded from the
analysis, the association of low hemoglobin levels with
perceived stress was no longer significant. These findings
confirm that overall health status has significant implica-
tions on psychological well-being. We do not know how
many patients were fully aware of their survival prognosis
based on their cancer progression, which raises the possi-
bility that illness severity and perceived stress were at least
TAbLe 2
Alexithymia, perceived stress, and hemoglobin levels by cancer staging
Cancer stagen
TAsa
pssb
Hemoglobinc
Mean sdMean sdMeansd
I244.0010.50 0.7110.50 .71
II16 50.9 2.321.31 5.03 11.561.82
III 1459.0 1.920.175.899.861.56
IV563.6 4.226.205.7210.201.92
ToTAL 3755.359.78 21.666.0510.681.83
aANOVA for TAS: F(3,33) = 5.8; P = .002.
bANOVA for PSS: F(3,33) = 3.5; P = .025.
cANOVA for hemoglobin level: F(3,33) = 2.7; P = .063.
Post hoc analyses for TAS showed significant differences between stages I and III (LSD = 18; P < .02) and I and IV (LSD = 21.6; P <.01), and between stages II and III (LSD =
10.1; P < .006) and IV (LSD = 13.3; P <.01). No significant differences were found between stages I and II (LSD = 8.3; P > .2) or between III and IV (LSD = 3.1; P >.5).
Post hoc analyses for PSS showed significant differences between stages I and III (LSD = 17.6; P < .03) and I and IV (LSD = 25.7; P < .005) and between stages II and IV (LSD =
10.9; P < .04). No significant difference was found between stages I and II (LSD = 14.8; P = .055) or between III and IV (LSD = 8.0; P =.13).
ANOVA: analysis of variance; LSD: least squares difference; PSS: Perceived Stress Scale; SD: standard deviation; TAS: Toronto Alexithymia Scale.

Page 5

AACP.com
Annals of Clinical Psychiatry | Vol. 23 No. 2 | May 2011
AnnAls of CliniCAl PsyChiAtry
129
in part associated because of the patients’ knowledge of
their prognosis.
Some limitations should be acknowledged before
further discussion of the findings in this study. The sam-
ple size was limited, and enrollment was notably influ-
enced by patients’ willingness to take part in the study.
These findings—including the association between can-
cer type and alexithymia—will need replication with a
larger, randomly selected sample of patients. In addition,
the sample size limited the number of possible analyses.
Only the TAS-20 total score, and not the 3 factors—dif-
ficulty identifying emotions, difficulty describing emo-
tions, and externally oriented thinking—was examined,
and no analyses were conducted to explore the associa-
tion of alexithymia and its factors with depressive symp-
toms. This association has been examined extensively in
the literature, showing that overlap and independence of
the 2 constructs exist.29,30
Similarly, the number of psychological illness sever-
ity factors was limited. A different set of illness severity
predictors may have yielded different results. No longitu-
dinal measures of alexithymia were available, allowing a
definite causal association with biological or psychologi-
cal factors. In this regard, contrary to the implications of
the results of the present study, relative stability of alexi-
thymia in cancer has been reported.31 In that study, sta-
bility of alexithymia was suggested based on an initial
evaluation carried out the day before surgery and another
evaluation attained 6 months after surgery; no measures
of illness severity were reported at the initial or follow-up
evaluations.31 A longitudinal design with assessment of
alexithymia before onset of the malignancy may resolve
this question.
In addition, the findings in the present study may not
generalize to childhood malignancies, as 5-year cancer
survivors (diagnosed before age 19) showed no signifi-
cant differences in alexithymia compared with healthy
comparisons, and no indication of medical severity—not
including hemoglobin levels—was found to be associ-
ated with alexithymia.32
The relationship between alexithymia and cancer is
complex and bidirectional. By altering the expression of
immune function, alexithymia may have a role in the eti-
ology of cancer.33 Levels of IL-1β, IL-2, and IL-4 and TH1/
TH2 lymphocytes (IL-2/IL-10) and CD4:CD8 ratios, as
well as CD4, are decreased in women with alexithymia.34
This may in part be the indirect result of elevated steroid
response to stress as a function of alexithymia.35 TH1 lym-
phocytes induce the proliferation of lymphocytes, CD8+,
and natural killer cells and therefore may have an impor-
tant role in stimulating cell-mediated defense against
cancerous cells. On the other hand, alexithymia may be a
psychological consequence of stress and suffering,36 com-
mon especially in advanced stages of tissue invasion. The
present study has shown that in addition to these psycho-
logical mechanisms, cancer may play a significant role in
alexithymia by altering the normal somatic physiology,
through both invasion of healthy tissue and anemia.
The ACC, a cerebral structure involved in the mech-
anisms of alexithymia,12,37-39 is among the brain regions
most vulnerable to hypoxic injury.40 In humans, the ACC
shows early functional decline,41 perhaps due to vascular
vulnerability.42 Decline in function and structure of the
ACC has been considered a mechanism for the greater
degree of alexithymia observed in older age.12 Therefore,
a potential mechanism by which hemoglobin is associ-
ated with alexithymia is a reduction of oxygen transport
to the brain and, specifically, to the anterior cingulate.12,42
The corpus callosum also has been posited as a brain
structure involved in alexithymia.43,44 This important
route of interhemispheric communication also is vulner-
able to low oxygen, as hypobaric hypoxia affects the mor-
phologic and functional maturation of the corpus callo-
sum in rats during postnatal development.45
The findings of the present study confirm the well-
established and general notion that lower hemoglobin
levels are associated with impaired psychological func-
tioning.46 It is noteworthy that correction of low hemo-
globin levels with iron therapy improves cognitive per-
formance in otherwise healthy adults.47 The hypothesis
that treatment of anemia improves alexithymia warrants
empirical testing. Improvement of alexithymia with
relatively simple interventions aimed at reducing ane-
mia may alleviate the negative effects of alexithymia on
immune response.
In summary, patients with cancer in the present
study showed greatly elevated rates of alexithymia in rela-
tion with rates in the general population.9 Alexithymia
was associated with progression of cancer invasion.
Alexithymia correlated directly with perceived stress and
indirectly with hemoglobin levels. Hemoglobin levels
and cancer invasion significantly correlated with alexi-
thymia when controlling for perceived stress. Perceived
stress, strongly influenced by personality, was associated
with cancer invasion but less definitively with lower levels
of hemoglobin. These findings suggest that alexithymia