Psychology
2013. Vol.4, No.6A1, 50-56
Published Online June 2013 in SciRes (http://www.scirp.org/journal/psych) http://dx.doi.org/10.4236/psych.2013.46A1008
50
The Health of Paid Caregivers:
A Cross-Sectional Study in Spanish Population
Isabel Cuéllar-Flores1, Rosa M. Limiñana-Gras2, M. Pilar Sánchez-López1
1Research Group of Psychological Styles, Gender and Health (EPSY), Red Hygeia (Health & Gender
International Alliance), Complutense University of Madrid, Madrid, Spain
2Department of Personality and Psychological Assessment and Treatment, Red Hygeia (Health & Gender
International Alliance), University of Murcia, Murcia, Spain
Received April 8th, 2013; revised May 12th, 2013; accepted June 10th, 2013
Copyright © 2013 Isabel Cuéllar-Flores et al. This is an open access article distributed under the Creative
Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium,
provided the original work is properly cited.
In countries with a welfare system, such as Spain, paid caregiving is an important profession and investi-
gating the health of people who care for the health of others is becoming an issue of increasing interest.
This paper reports on a study assessing the health of people who care for the health of others. It investi-
gates whether caregiving jobs carry with them a higher risk of psychological and physical health problems
than other unrelated occupations, and it examines whether risk profiles differ by sex and educational level.
A descriptive correlational approach and a cross-sectional design were used. The sample included 330
participants. Half of the participants worked as caregivers, and the other half worked in professions unre-
lated to caregiving services. The results indicate that paid caregivers have poorer health than their
matched counterparts. The variable of caregiving explains this difference better than sex or educational
level and professional category.
Keywords: Paid Caregivers; Physical Health; Psychological Health; Psychological Stress; Physical Work;
Job Satisfaction; Nursing
Introduction
Investigating the health of people who care for the health of
others is becoming an issue of increasing interest. The magni-
tude of the need for chronic and acute care and the presence of
developing welfare states is leading to concern about “caring
for the carers”. According to the Encuesta sobre Discapaci-
dades, Autonomía personal y situaciones de Dependencia (Spa-
nish survey on disabilities, personal autonomy, and dependency)
from the Spanish National Institute of Statistics (INE, 2008),
9% of the population has some type of disability, and it is esti-
mated that only 12% of the time devoted annually to caregiving
for health reasons is provided by health professionals. The de-
mand for this care will increase in the future (Durán, 2000), as
evidenced by the increase in the proportion of health profes-
sionals per inhabitant in Spain since the 1990s (INE, 2007). Ac-
cording to the Spanish Active Population Survey (INE, 2007),
there were 106,843 nursing assistants and 126,409 nurses in the
Spanish health sector in 2006.
Despite these numbers, research on the health of professional
or paid caregivers is far less abundant than research on the
health of family caregivers. Some of the available data suggest
that working in the health sector involves some cost to caregiv-
ers. For example, the data on morbidity and mortality among
nurses shows that they are at greater risk of dying from cancer,
viral hepatitis, falls, suicide, and substance abuse than other
professions. Nurses also have increased rates of diabetes and
heart disease (National Occupational Mortality Surveillance
System, USA, cited by McNeely, 2005). According to the US
Bureau of Labor Statistics, in 2006, nurses working in institu-
tions for the elderly had higher rates of backache, concussions,
tendonitis and sprains than the average worker in all private
industries (Bureau of Task Statistics [BLS], 2006). Although it
seems that the prevalence of disease and lesions varies accord-
ing to the occupational group under study (Clarke et al., 2002;
Edwards & Burnard, 2003), nursing is usually considered a
very stressful occupation with a high prevalence of psycholo-
gical disorders related to stress (Celentano & Johnson, 1987;
Revicki & May, 1989).
There are difficulties involved in defining and delimiting the
term professional caregiver. This term includes a wide range of
categories, from qualified nursing professionals to unqualified
home caregivers with no formal regulations to govern their ac-
tivities. This range partly accounts for the difficulty in identify-
ing the labour and social situation in the caregiving sector.
However, there are some indicators of the difficulties faced by
these workers. According to the US Bureau of Labor Statistics,
in the United States in 2003, caregiving was one of the jobs
with the highest rate of pathology; 10.1% of caregivers working
in centres for the elderly and other dependent populations had
some kind of pathology. In contrast, the rate of pathology was
6.8% in the building sector, and 5% in all other private em-
ployment (BLS, 2003). Eriksen (2003a, 2003b, 2006) and Erik-
sen et al. (2008) have shown that nursing assistants often pre-
sent backache, a frequent complaint within this occupational
group. According to these authors, sleep-related problems, psy-
chological distress, chronic fatigue, and high rates of labour
absenteeism are also quite common among nursing assistants.

I. CUÉLLAR-FLORES ET AL.
Several research publications report a high frequency of dis-
tress and other psychological problems among healthcare per-
sonnel (Bakker, 2000; Mealer et al., 2009). In fact, most studies
have focused on mental health rather than physical health, es-
pecially when investigating nurses (McNeely, 2005; Winwood
et al., 2006a), although the two factors are complementary. For
example, emotional health has been identified as an important
and significant predictor of fatigue (Rella et al., 2009). Anxiety,
self-perceived satisfaction, and self-esteem have been related to
physical health (Hansen, 2003; Stinson et al., 2008), especially
among women (Sánchez-López et al., 2006).
Every professional category involves different sources of
stress and different health conditions (Foxall et al., 1990; Hardy
et al., 1997). Therefore, the kind of job performed is likely to
influence workers’ health. Comparative studies show that
among health care staff, nurses, especially oncologic nurses and
nurses working with AIDS patients, run the highest risk of
burnout (Bernardi et al., 2005). Although caregiving-related
jobs have a variety of specific working conditions (shifts, salary,
recognition, etc.), they share certain stress-generating factors,
such as a high level of emotional involvement and high respon-
sibility. In fact, health service workers have greater levels of
minor psychiatric disorders and fatigue than the general public,
and psychiatric morbidity is highest among managers, doctors,
nurses and caregivers (Hardy et al., 1997; Wall et al., 1997).
Therefore, it is relevant to ask whether paid caregiving, regard-
less of the level of formal training, the characteristics of the
care situation, or the people cared for, involves a stressor that
influences physical and psychological health indexes.
In addition, some variables, such as age and female gender,
have been suggested as health risk factors in health-related jobs
(Chang et al., 2006; Lambert et al., 2004; Wall et al., 1997;
Winwood et al., 2006b). In other words, the results of some
studies suggest that women working as caregivers and older
women have poorer health. It has also been suggested that some
professional categories in the caregiving sector incur greater
health risks. For example, nursing assistants usually have very
low wages, high objective workloads, rather unfavourable la-
bour conditions (rotating night shifts or a lack of contracts),
limited training, lack of control over their work, and limited
opportunities for job improvement (Mickus et al., 2004). Some
studies have associated these factors with greater psychological
distress (Eriksen et al., 2003; Hochwälder, 2008).
The present study aims to investigate whether there is a
higher risk of psychological and physical health problems in
caregiving service occupations than in non-caregiving occupa-
tions, and whether risk profiles differ by sex and educational
level (related to specific caregiving service professions in
Spain).
The objectives of this study are as follows: 1) to analyse
whether there are differences in the physical and psychological
well-being and health of people working as caregivers and
those who work in other fields. Thus, we investigate whether
the caregiving factor itself, performed in different conditions,
may affect health; and 2) to test whether the variables of sex
and educational level modulate health differences between
caregivers and non-caregivers. To this end, two subgroups of
caregiving professionals were established: those with higher
education and those with secondary school or equivalent train-
ing only.
Based on the reviewed literature, we proposed the following
hypotheses: 1) paid caregivers have worse health than matched
non-caregivers in the indexes analysed; 2) caregiving women
present worse health than men (whether they work as caregiv-
ers or not) and non-caregiving women; 3) caregivers with sec-
ondary school education (nursing assistants or geriatric assis-
tants) have worse health than non-caregivers (regardless of their
educational level) and caregivers with higher education (such as
physicians and nurses).
Methods
Participants
Based on data on the prevalence of psychological disorders
(Mealer et al., 2009), a sample size of approximately 165 was
estimated to achieve 80% power at a 5% level of statistical
significance. The sample consisted of 330 participants, 60 men
(18.2%) and 270 women (81.8%), half of whom worked as
caregivers or health care providers and half of whom worked in
areas unrelated to caregiving. The age of the participants ranged
from 21 to 66 years, with a mean of 36.65 years (SD = 10.28).
The two groups were matched according to the following
characteristics: age, sex, educational level, socioeconomic level,
marital status and having or not having children. All of the par-
ticipants had a medium socioeconomic level, 63% of the par-
ticipants had a higher education level, and 37% of the partici-
pants had a secondary education level. Of the participants in the
sample, 41.5% were single, 49.3% were married or partnered,
8.5% were separated or divorced, and 0.6% were widowed.
The caregivers worked in centres for the elderly and in hos-
pitals in the Community of Madrid (Spain) in the following
areas: nursing (48.2%), geriatric or nursing assistant (37%), and
medical, physiotherapy, psychology, or occupational therapy
staff (14.8%). The caregivers were contacted in their work-
places and offered the opportunity to participate voluntarily and
anonymously in this study.
The participants in the control group worked in different oc-
cupations: clerical work (24.2%), technicians (16.4%), services
(15.8%), liberal professions (14.5%), managers/self-employed
(11.5%), unskilled workers (10.3%) and specialist workers and
operators (7.3%). The control group participants were contacted
at home and were selected from a wider sample of middle-class
1001 participants (as defined by the Spanish National Popula-
tion Census in the CPN-94 classification and based on the type
of work performed (INE, 2001) based on matches to the care-
givers for the examined variables.
Design
This study employed a descriptive correlational approach and
a cross-sectional design.
Procedure
The participants were recruited using two approaches. Care-
givers were recruited through contact with several institutions
for people with disabilities, and non-caregivers were identified
by college students from Universidad Complutense, Madrid,
who recruited participants with diverse demographic character-
istics (snowball sampling). The data were collected by several
trained evaluators during two 45-minute sessions. The sessions
took place in the homes or workplaces of the participants from
October 2007 to April 2008. All participants were invited to
take part in the study following a short description of the nature

I. CUÉLLAR-FLORES ET AL.
of the study (i.e., “to better understand people’s health”). All
participants gave their informed consent after the researchers
explained the purpose of the investigation, provided a descrip-
tion of the study procedures and alternatives to participation,
guaranteed the participants’ freedom to withdraw from any part
of the study without consequences and guaranteed the anonym-
ity of their data. Based on the total sample, two groups were
selected (caregivers and the control group) and randomised into
pairs according to the following variables: age (10-year ranges);
sex; educational level (two levels: secondary school and higher
education); marital status (four categories: single, married or
partnered, widow/widower and divorced or separated); and
children (two categories: with or without children). All partici-
pants were classified as having a mid-range socioeconomic
level. The pairing variables were selected to control for their
potential confounding effect on the results because they could
interact with the selection variables and create differences in
the health variables of the participants.
Instruments
The social and demographic questionnaire included items on
age, sex, marital status, educational level, employment status,
and profession. Previous research employed this instrument
(Sánchez-López et al., 2006; Limiñana, Corbalán, & Sánchez-
López, 2009; Sánchez-López et al., 2008), and satisfactory re-
sults were obtained.
A questionnaire on physical health asked participants to in-
dicate their number of chronic diseases, ailments, psychophar-
macological medications, frequency of substance consumption
(including cigarettes and alcohol), visits to the doctor and self-
perceived physical health. This questionnaire was used previ-
ously with satisfactory results (Sánchez-López et al., 2006,
Sánchez-López et al., 2008).
Psychological health data were obtained from The Rosenberg
Self-Esteem Scale (Cronbach’s alpha of .88 in Spanish popula-
tion; see Baños & Guillén, 2000), the Escala Reducida de An-
siedad (Reduced Anxiety Scale; Martínez-Sánchez et al., 1995;
Cronbach’s alpha .90 in Spanish population) and the Cuestion-
ario de Satisfacción Autopercibida: laboral, familiar y global
(Self-perceived satisfaction questionnaire: work, family, and
overall satisfaction). These scales were validated by previous
research (see, for example, Díaz-Morales, & Sánchez-López,
2002; Sánchez-López et al., 2008).
Ethical Considerations
This study obtained approval from the institutional review
board of the Universidad Complutense, Madrid. All participants
gave their informed consent after the purpose of the research
was explained and the voluntary basis of participation was in-
dicated. The anonymity of the data was ensured.
According to Spanish legislation on the study of human be-
ings (Law 14/2007 of July 3, of biometric research) and based
on the Helsinki Declaration, “consent” is defined as the “mani-
festation of free and conscious will validly emitted by an able
person or by his/her authorised representative, preceded by
adequate information”. In Spain, written and signed informed
consent is only applied in biomedical studies (articles 4 and 13
of Law 14/2007).
Data Analysis
Data were analysed using the statistical software package
SPSS 15.0 for Windows. A multivariate analysis of covariance
(MANCOVA) was performed to compare health variables be-
tween caregivers and non-caregivers and to explore the modu-
lating effect of sex and educational level on the differences in
health variables between caregivers and controls. Being a care-
giver, sex and educational level were introduced as independent
variables. The dependent variables were related to physical
health (number of chronic diseases, number of ailments, num-
ber of prescribed drugs taken, frequency of substance consump-
tion, visits to the doctor and self-assessment of physical health)
and psychological health (anxiety, self-esteem, work satisfac-
tion, family satisfaction and overall satisfaction). Furthermore,
the variable age was introduced into the model as a covariate.
Although age was controlled for the independent variable of
being a caregiver when pairing participants, it was believed to
have a confounding effect on the other two factors (sex and
educational level). Before performing the MANCOVA, the
normality and homogeneity of variance (Box Test, Levene’s test)
and the correlation between the dependent variables (Bartlett
test of sphericity) were verified. The indexes for the effect size
were also calculated using η2, which indicates the proportion of
variance explained by each variation source.
Results
Table 1 shows the descriptive statistics of the paid caregivers
and the control group, classified according to sex and education.
In support of Hypothesis 1, the MANCOVA results for the va-
riable being a caregiver [Wilks (10, 280) = .814, p ≤ .000, η2
= .19] confirm that not working as a caregiver has a significant
effect on health variables, in this case with a high effect size.
The MANCOVA results for the variable being a caregiver
(Ta b l e 2 ) show statistically significant differences in the num-
ber of ailments, self-esteem, work satisfaction, family satisfac-
tion, and self-perceived physical health. Caregivers present a
greater amount of physical ailments, indicate less self-esteem
and family satisfaction and assess their own physical health as
worse than their matched control counterparts. However, care-
givers are more satisfied with their work than are non-caregiv-
ers. The size of these differences is moderate for all variables.
With respect to Hypotheses 2 and 3, the results from the
MANCOVA indicate that no statistically significant results
were found for sex [Wilks (10, 280) = .968, p ≤ .506], educa-
tional level [Wilks (10.280) = .950, p ≤ .149], or the being a
caregiver* sex interaction [Wilks (10.280) = .939, p ≤ .059].
Although Wilks’ lambda for sex and educational level does
not have significance below the specified level, in the between-
subjects effects test, a significant effect of the variable sex is
found on the number of ailments [F (1, 289) = 4.030, p ≤ .046,
η2 = .01] and anxiety [F (1, 289) = 3.753, p ≤ .054, η2 = .01].
The estimated values of the comparison indicate that women,
whether working as caregivers or not, suffer a greater number
of ailments and present greater anxiety than males. Educational
level has a significant effect on work satisfaction [F (1, 289) =
3.775, p ≤ .053, η2 = .01] and number of visits to the doctor [F
(1, 289) = 5.494, p ≤ .020, η2 = .01]. The subgroup with a
higher educational level shows greater work satisfaction, and
the group with a lower educational level shows a greater num-
ber of visits to the doctor. The effect size of these differences is
small.
However, the results from the MANCOVA indicate a level of
significance below the specified level (.05) for the caregiver*
52

I. CUÉLLAR-FLORES ET AL.
Table 1.
Descriptive statistics for non-caregivers and caregivers.
Non-caregivers (n = 165) M (SD) Caregivers (n = 165) M (SD)
Men (n = 30) Women (n = 135) Men (n = 30) Women (n = 135)
Health
Var.
Second Ed.
(n = 3)
Higher Ed.
(n = 27)
Second Ed.
(n = 72)
Higher Ed.
(n = 63)
Second Ed.
(n = 3)
Higher Ed.
(n = 27)
Second Ed.
(n = 72)
Higher Ed.
(n = 63)
1. .0 (.0) .4 (.75) .4 (1.2) .4 (1.1) 1.0 (1.7) .1 (.3) .7 (1.1) .4 (.8)
2. 13.6 (2.3) 13.1 (2.8) 16.1 (2.9) 15.5 (3.2) 17 (2.6) 15.6 (9.0 ) 18.5 (4.1) 16.5 (2.9)
3. 3.7 (3.1) 4.5 (2.3) 4.4 (2.1) 4.1 (2.1) 3.3 (1.2) 3.8 (2.0) 4.2 (2.4) 4.3 (2.1)
4. 18.3 (7.6) 15.7 (9.6) 21.8 (11.6) 22.7 (11.6) 21 (6.6) 17.5 (12.5)24.3 (13.1) 23.9 (9.5)
5. 15.7 (.6) 15.5 (2.4) 15.5 (2.5) 15.2 (2.0) 10.3 (6.4) 13.2 (8.4) 10.4 (4.4) 10.7 (6.9)
6. 3 (3.5) 8.2 (1.6) 7.1 (1.6) 6.8 (2.3) 10.0 (.0) 8 (1.4) 7.8 (2.1) 8.2 (1.7)
7. 8.7 (1.2) 8.5 (1.2) 8.4 (1.6) 8.1 (1.8) 8.7 (1.2) 7.3 (1.6) 7.4 (1.6) 7.7 (1.2)
8. 7.3 (1.6) 8.5 (0.8) 7.9 (1.3) 7.7 (1.4) 8.3 (1.2) 7.8 (1.2) 7.6 (1.5) 8.0 (1.1)
9. 2.3 (2.5) 1.8 (2.3) 5.1 (9.8) 3.5 (4.5) 12.3 (2.5) 1.4 (1.5) 3.0 (2.7) 3.2 (3.3)
10. 8.7 (1.2) 8.0 (1.3) 8.0 (1.3) 7.9 (1.3) 6.7 (1.2) 7.8 (1.3) 6.9 (1.7) 7.7 (1.7)
Note: Health variables: 1 = Chronic disease, 2 = Ailment, 3 = Substance consumption, 4 = Anxiety, 5 = Self-esteem, 6 = Work satisfaction, 7 = Family satisfaction, 8 =
Overall satisfaction, 9 = Visits to the doctor, 10 = Self-perceived physical health. Abbreviations: Second Ed, Secondary education; Higher Ed, Higher Education.
Table 2.
Multivariate analyses of covariance: non-caregivers versus caregivers.
Health variables Non-caregivers (n = 165) M (SD) Caregivers (n = 165) M (SD) F (330)ª η2
1. Chronic .42 (1.02) .47 (.87) 1.17 .004
2. Ailments 15.33 (3.18) 17.07 (4.90) 6.45* .022
3. Substance consumption 4.30 (2.13) 4.30 (2.23) .25 .001
4. Anxiety1 21.10 (11.28) 22.82 (11.72) .64 .002
5. Self-esteem 15.38 (2.28) 11.12 (11.26) 17.59*** .057
6. Work satisfaction 7.07 (2.05) 8.13 (1.83) 26.94*** .085
7. Family satisfaction 8.32 (1.62) 7.34 (1.64) 3.77* .013
8. Overall satisfaction 7.90 (1.30) 7.79 (1.32) .09 .000
9. Visits to the physician 3.90 (7.18) 2.98 (3.09)a 1.94 .007
10. Self perceived physical health 8.01 (1.28) 7.46 (1.60) 7.861** .026
Note: *p < .05, **p < .01, ***p = .001; ªMANCOVA (test of Between-Subjects Effects; factor: being caregiver); ¹n = 134.
educational level interaction [Wilks (10, 280) = .912, p ≤ .003,
η2 = .06] and for the interaction of the three factors caregiver*
sex* educational level [Wilks (10, 280) = .895, p ≤ .000, η2
= .10]. In addition, the MANCOVA shows a significant effect
for the covariate age [Wilks (10, 289) = .935, p ≤ .041, η2
= .06]. The effect size of these differences is moderate.
Multivariate Analyses of Covariance for the caregiver* edu-
cational level interaction (Ta ble 3 ) show a statistically signifi-
cant, although small, effect on work satisfaction. Post hoc
comparisons show that male and female caregivers with secon-
dary education present greater work satisfaction than their col-
leagues with higher education. However, in the group of non-

I. CUÉLLAR-FLORES ET AL.
Table 3.
Multivariate analyses of covariance: being caregiver* educational level interaction.
Non-caregivers (n = 165) M (SD) Caregivers (n = 165) M (SD)
Health variables
Sec. Ed. (n = 75) Higher Ed. (n = 90) Sec. Ed. (n = 75) Higher Ed. (n = 90)
F (330)ª η2
1. Chronic .4 (1.1) .4 (1.0) .7 (1.1) .3 (.7) 2.99 .010
2. Ailments 16.0 (2.9) 14.8 (3.3) 18.4 (4.0 ) 16.2 (5.6) .35 .001
3. Substance consumption 4.4 (2.1) 4.2 (2.1) 4.2 (2.4) 4.2 (2.1) .00 .000
4. Anxiety1 21.7 (11.4) 20.6 (11.2) 24.1 (12.8) 22.0 (11.0) .05 .000
5. Self-esteem 15.5 (2.5) 15.3 (2.1) 10.4 (4.4) 11.5 (7.4) .75 .003
6. Work satisfaction 6.9 (1.9) 7.2 (2.2) 8.0 (2.1) 8.1 (1.6) 14.84*** .049
7. Family satisfaction 8.4 (1.6) 8.2 (1.6) 7.5 (1.6) 7.5 (1.4) .29 .001
8. Overall satisfaction 7.9 (1.3) 7.9 (1.3) 7.6 (1.4) 8.0 (1.2) .73 .003
9. Visits to the physician 5.0 (9.6) 3.0 (4.0) 3.5 (3.4) 2.6 (3.0) 2.80 .010
10. Self perceived physical health 8.1 (1.3) 8.0 (1.3) 6.9 (1.7) 7.7 (1.6) 3.59 .012
Note: *p < .05, **p < .01, ***p = .001; ªMANCOVA (test of Between-Subjects Effects; factor: being caregiver* educational level interaction); ¹n = 134.
caregivers, greater work satisfaction was found in those with a
higher educational level.
Although the Wilks’ lambda for caregiver* sex is not signifi-
cant, the between-subjects effects test shows a statistically sig-
nificant, although small, effect for the interaction caregiver* sex
on work satisfaction [F (1, 289) = 7.542, p ≤ .006, η2 = .03] and
number of visits to the doctor [F (1, 289) = 5.568, p ≤ .019, η2
= .02]. In non-caregivers, men show greater work satisfac- tion
than women, whereas men and women in the caregiver group
have similar levels of satisfaction. Concerning the num- ber of
visits to the doctor, male caregivers visit the doctor more often
than do male non-caregivers; in contrast, female caregiv- ers go
to the doctor less often than non-caregiving females do.
Regarding the caregiver* sex* educational level interaction, a
statistically significant effect on work satisfaction was ob-
tained [F (1, 289) = 21.295, p < .000, η2 = .07], with a moderate
effect size. This interaction also has a significant effect on
overall satisfaction [F (1, 289) = 3.963, p ≤ .047, η2 = .01] and
number of visits to the doctor [F (1, 289) = 5.855, p ≤ .016, η2
= .02], but the effect size is low.
The male caregivers with secondary education had the high-
est work satisfaction, followed by male non-caregivers with a
higher educational level, and by female caregivers with a higher
educational level. Female caregivers with secondary education
were next in rank, followed by female non-caregivers with
secondary education, female non-caregivers with a higher edu-
cational level and, finally, male non-caregivers with secondary
education. Regarding overall satisfaction, the ranking from
highest to lowest is as follows: male non-caregivers with a
higher educational level, male caregivers with secondary edu-
cation, female caregivers with a higher educational level, fe-
male non-caregivers with secondary education, male caregivers
with a higher educational level, female non-caregivers with a
higher educational level, female caregivers with secondary edu-
cation and, finally, male non-caregivers with secondary educa-
tion. Male caregivers with secondary education visited the doc-
tor most often, followed by female non-caregivers with se-
condary education. These groups were followed by female non-
caregivers with a higher educational level, female caregivers
with a higher educational level, female caregivers with secon-
dary education, male non-caregivers with secondary education,
male non-caregivers with a higher educational level and male
caregivers with a higher educational level.
Discussion
The results support the hypothesis that paid caregiving, re-
gardless of the level of formal training, the characteristics of the
situation, and the kind of people cared for, has a consistent
negative effect on health, especially physical health. In our
study, the caregiver group presented poorer self-perceived
physical health and a greater number of ailments than the con-
trol group. This finding is in agreement with the results of other
studies (Eriksen, 2003a, 2003b, 2006; McNeely, 2005), but it
contradicts the results from recent studies such as Winwood et
al. (2006a), indicating a greater impact of psychological stress
than physical work strain in nursing.
The caregiver group also presented lower self-esteem, which
has been considered a predictor of health problems (Stinson et
al., 2008). No differences were found concerning the level of
anxiety, which contradicts previous findings reporting a high
frequency of distress and burn-out symptoms in health person-
nel (Bakker et al., 2000; Mealer et al., 2009). This finding
might be explained by the fact that the type of measurements
used in these studies differs from the measurements used in the
present work (i.e., previous studies focused on general distress
and burn-out symptoms). No differences were found in overall
satisfaction, but the caregivers in our study felt more satisfied
with their work and less satisfied with their family life than the
control group.
These differences in work and family satisfaction could be
related to how care is perceived when it is performed in a pro-
fessional context. Some studies have shown that the perception
of caregiving as an emotionally and psychologically rewarding
experience is based less on the characteristics of the assistance
provided and more on the self-perceived role and responsibility
54

I. CUÉLLAR-FLORES ET AL.
and the meaning attributed to the role of the caregiver (Green-
berg & Morris, 1994; Schwartz, 2003). Nevertheless, and de-
spite the higher levels of work satisfaction manifested by the
caregivers in this study, the data indicate that caregivers have
worse physical health and lower self-esteem, which means that
work satisfaction does not seem to offer them protection from
health issues.
Our results do not support the hypothesis, based on previous
research (Wall et al., 1997), that female caregivers have worse
health than do men (whether caregivers or not) and female
non-caregivers. The fact of being a woman or a man was only a
determinant for modulating the differences in work satisfaction
and visits to the doctor between caregivers and non-caregivers,
but not in the sense established by the hypothesis. The results of
this study indicate that women, in general, present a greater
number of ailments and greater anxiety than men, although the
effect size is very small. However, female caregivers do not
present poorer health than the other three groups. In fact, fe-
male caregivers show greater work satisfaction than male and
female non-caregivers and have fewer visits to the doctor than
female non-caregivers.
It is possible that women, who are used to providing care in
the family setting, feel greater acknowledgment and, thus,
greater satisfaction in the professional caregiving context. Nev-
ertheless, it is important to bear in mind that the number of men
in this study was small. Therefore, potential differences or dif-
ferential effects on health due to being a woman or a man might
not have been detected. However, several studies have shown
that when some social and demographic variables are matched
in men and women, the differences in physical and psychologi-
cal health due to sex disappear (Bartley et al., 1998; Sánchez-
López et al., 2008) or the size of the differences decreases
(Pinqüart & Sörensen, 2006).
The third hypothesis was not confirmed. This finding sug-
gests that caregivers with secondary education (for example,
caregiving assistants) have poorer health than non-caregivers
(regardless of their educational level) and caregivers with a
higher educational level (physicians, nurses, etc.). Differences
were only found in one of the variables, and the results indicate
that caregiving professionals with a lower educational level
have greater work satisfaction than professionals with higher
education, whereas the opposite results were found in the non-
caregiver group. According to the results obtained, being a
caregiver is more relevant for explaining health characteristics
than the specific job category in the caregiving sector. It is pos-
sible that controlling for social and demographic variables (by
matching participants) in this study had an influence on the
results, whereas studies that found differences regarding pro-
fessional categories (Foxall et al., 1990; Hardy et al., 1997)
may not have taken the effect of these variables into account.
Nevertheless, when the interaction between the three factors
(being a caregiver, sex and educational level) is analysed, the
situation is different because differences were found in three of
the health variables under study. Therefore, it seems necessary
to include both variables (educational level and sex) to better
understand the characteristics of care-related professions.
Study Limitations and Implications
We note several limitations of the current study. This re-
search should be replicated in the future with populations of
different nationalities to continue to verify its generalisability.
Another limitation of this study is the use of self-reported mea-
sures. It is necessary to utilise objective health indicators to
complete the investigation of caregiving and health. Although
the proportion of females to males in this study reflects the
real-life situation (INE, 2007b), the presence of such a small
number of men is a limitation. A sample with a greater number
of paid caregivers, especially men, is required for an in-depth
study to confirm the combined modulating effect of sex and
educational level (or professional category) on the results.
Being a caregiver per se is more important than the profes-
sional category for health and satisfaction. Based on the data
obtained in this study, it would be interesting to design studies
of social and organisational characteristics and other moderat-
ing factors associated with different occupational groups and to
test whether these characteristics are responsible for the differ-
ences between professionals in health and satisfaction.
Other variables that were not included in this work, such as
the level of depression, classic burn-out measurements and
objective measures, should be part of future research on the
health of caregivers.
Conclusion
Paid caregiving, regardless of the level of formal training, the
characteristics of the situation, and the kind of people cared for,
is related to a negative effect on physical health. The caregivers
presented lower self-esteem as well as greater satisfaction with
their work and less satisfaction with their family life.
Caregivers, whether nursing, medical staff, physiotherapy,
psychology, or occupational therapy staff, perform a crucial
role in the functioning of social and health service institutions.
Therefore, it is even more necessary that they receive attention
in research focused on health. The results of this study show
that both caregivers and those for whom they care have a need
for care and attention to secure the welfare of both groups and
the smooth operation of services. This knowledge is potentially
useful in health programs and socio-health policies, and it re-
veals the need to systematically introduce the study of caregiv-
ers to health-related investigations and interventions to facili-
tate the diagnosis and prevention of poor health.
REFERENCES
Bakker, A. B., Killmer, C. H., Siegrist, J., & Schaufeli, W. B. (2000).
Effort-reward imbalance and burnout among nurses. Journal of Ad-
vanced Nursing, 31, 884-891. doi:10.1046/j.1365-2648.2000.01361.x
Baños, R., & Guillén, V. (2000). Psychometric characteristics in normal
and social phobic samples for a Spanish version of the Rosemberg
self-esteem scale. Psyc hol ogi cal Reports, 87, 269-274.
Bartley, M., Davey Smith, G., & Blaine, D. (1998). Beyond the black
report. Sociology of Health a nd Illness, 20, 563-577.
doi:10.1111/1467-9566.00119
Bernardi, M., Catania, G., & Marceca, F. (2005). The world of nursing
burnout. A literature review. Professioni Infermieristiche, 58, 75-79.
BLS, Bureau of Labor Statistics (2003, 2006). Work-related injury and
illness statistics by industry. www.bls.gov
Celentano, D. D., & Johnson, J. V. (1987). Stress in health care workers.
Occupational Medicine, 2, 593-608.
Chang, E., Daly, J., Hancock, K., Bidewell, J., Johnson, A., Lambert,
V., & Lambert, C. (2006). The relationships among workplace stress-
sors, coping methods, demographic characteristics and health in Aus-
tralian nurses. Journal of Proffesional Nursing, 22, 30-38.
doi:10.1016/j.profnurs.2005.12.002
Clarke, S., Rockett, J., Sloane, D., & Aiken, L. (2002). Organizational

I. CUÉLLAR-FLORES ET AL.
56
climate, staffing, and safety equipment as predictors of needlestick
injuries and near-misses in hospital nurses. American Journal of In-
fection Control, 30, 207-216. doi:10.1067/mic.2002.123392
Díaz-Morales, J. F., & Sánchez-López, M. P. (2002). Connections
between personality types and self-perceived satisfaction in several
life areas in University students. Psichotema, 14, 100-105.
Durán, M. A. (2000). New division of labour in health care. Política y
Sociedad, 35, 9-30.
Edwards, D., & Burnard, P. (2003). A systematic review of stress and
stress management interventions for mental health nurses. Journal of
Advanced Nursing, 42, 169-200.
doi:10.1046/j.1365-2648.2003.02600.x
Eriksen, W. (2003a). The prevalence of musculoskeletal pain in Nor-
wegian nurses’ aides. International Archives of Occupational and
Environmental Health, 76, 625-630. doi:10.1007/s00420-003-0453-6
Eriksen, W. (2003b). Service sector and perceived social support at
work in Norwegian nurses’ aides. International Archives of Occupa-
tional and Environmental He al th , 76, 549-552.
doi:10.1007/s00420-003-0447-4
Eriksen, W., Bruusgaard, D., & Knardahl, S. (2003). Work factors as
predictors of sickness absence: A three month prospective study of
nurses’ aides. Occupational and Environmental Medicine, 60, 271-
278. doi:10.1136/oem.60.4.271
Eriksen, W. (2006). Work factors as predictors of persistent fatigue: A
prospective study of nurses’ aides. Occupational and Environmental
Medicine, 63, 428-434. doi:10.1136/oem.2005.019729
Eriksen, W., Bjorvatn, B., Bruusgaard, D., & Knardahl, S. (2008). Work
factors as predictors of poor sleep in nurses’ aides. International Ar-
chives of Occupational and E nvi r onmental Health, 81, 301-310.
doi:10.1007/s00420-007-0214-z
Foxall, M. J., Zimmerman, L., Standley, R., & Bene-Captain, B. (1990).
A comparison of frequency and sources of nursing job stress per-
ceived by intensive care, hospice and medical-surgical nurses. Jour-
nal of Advanced Nursing, 15, 577-584.
doi:10.1111/j.1365-2648.1990.tb01857.x
Greenberg, M., & Morris, N. (1994) Engrossment: The newborn’s im-
pact upon the father. American Journal of Orthopsychiatry, 44, 520-
531. doi:10.1111/j.1939-0025.1974.tb00906.x
Hansen, S. (2003). Mental health issues associated with cardiovascular
disease in women. Psychiatric clinics of North America, 26, 693-712.
doi:10.1016/S0193-953X(03)00037-6
Hardy, G. E., Shapiro, D. A., & Borrill, C. S. (1997) Fatigue in the
workforce of national health service trusts: Levels of symptomatol-
ogy and links with minor psychiatric disorder, demographic, occupa-
tional and work role factors. Journal of Psychosomatic Research, 43,
83-92. doi:10.1016/S0022-3999(97)00019-6
Hochwälder, J. (2008). A longitudinal study of the relationship between
empowerment and burnout among registered and assistant nurses.
Work: A Journal of Prevention, Assessment and Rehabilitation, 30,
343-352.
INE Instituto Nacional de Estadística (2001). Population and housing
census. www.ine.es
INE Instituto Nacional de Estadística (2007). Working population sur-
vey. Registered health professionals. www.ine.es
INE Instituto Nacional de Estadística (2008). Survey on disabilities,
self-autonomy, & dependence situations. www.ine.es
Lambert, V., Lambert, C., Itano, J., Inouye, J., Kim, S., Kuniviktikul,
W., Sitthimongkol, Y., Pongthavornkamol, K., Gasemgitvattana, S.,
& Ito, M. (2004) Cross-cultural comparison of workplace stressors,
ways of coping and demographic characteristics as predictors of
physical and mental health among hospital nurses in Japan, Thailand,
South Korea and the USA (Hawaii). International Journal of Nurs-
ing Studies, 41, 671-684. doi:10.1016/j.ijnurstu.2004.02.003
Limiñana, R., Corbalán, F. J., & Sánchez-López, M. P. (2009). Think-
ing styles and coping when caring for a child with severe spina bifida.
Journal of Developmental and Physical Disabilities, 19, 125-134.
Martínez-Sánchez, F., Cano-Vindel, A., Castillo, J. C., Sánchez, J. A.,
Ortiz, B., & Gordillo, E. (1995). A reduced anxiety scale based on
the anxiety situations and responses inventory. Anales de psicología,
11, 97-104.
Mealer, M., Burnham, E. L., Goode, C. J., Rothbaum, B., & Moss, M.
(2009). The prevalence and impact of post traumatic stress disorder
and burnout syndrome in nurses. Depression and Anxiety, 26, 1118-
1126. doi:10.1002/da.20631
McNeely, E. (2005). The consequences of job stress for nurses’ health:
Time for a check-up. Nursing Outlook, 53, 291-299.
doi:10.1016/j.outlook.2005.10.001
Mickus, M., Luz, C., & Hogan, A. (2004). Voices from the front: Re-
cruitment and retention of direct care workers in long term care
across Michigan. Lansing, MI: Michigan State University.
Pinqüart, M., & Sörensen, S. (2006). Gender differences in caregiver
stressors, social resources, and health: An updated meta-analysis.
Journals of Gerontology, Series B: Psychological Sciences and So-
cial Sciences, 61B, 33-45. doi:10.1093/geronb/61.1.P33
Rella, S., Winwood, P., & Lushington, K. (2009). When does nursing
burnout begin? An investigation of the fatigue experience of Austra-
lian nursing students. Journ al of Nu rsing Management, 17, 886-897.
doi:10.1111/j.1365-2834.2008.00883.x
Revicki, D. A., & May, H. J. (1989). Organizational characteristics.
occupational stress, and mental health in nurses. Behavioral Medi-
cine, 15, 30-36. doi:10.1080/08964289.1989.9935149
Sánchez-López, M. P., Aparicio-García, M. E., & Dresch, V. (2006).
Anxiety, self-esteem and self-perceived wellbeing: Differences be-
tween men and women. Psicothema, 18, 584-590.
Sánchez-López, M. P., López-García, J. J., Dresch, V., & Corbalán, J.
(2008) Sociodemographic, psychological and health-related factors
associated with poor mental health in Spanish women and men in
midlife. Women, & Health, 48 , 445-465.
doi:10.1080/03630240802575096
Schwartz, Ch. (2003) Parent of children with chronic disabilities: The
gratification of caregiving. Families in Society, 84, 576-584.
doi:10.1606/1044-3894.143
Stinson, D. A., Logel, C., Zanna, M. P., Holmes, J. G., Cameron, J. J.,
Wood, J. V., & Spencer, S. J. (2008) The cost of lower self-esteem:
Testing a self- and social-bonds model of health. Journal of Person-
ality and Social Psychology, 94, 412-428.
doi:10.1037/0022-3514.94.3.412
Wall, T. D., Bolden, R. I., Borrill, C. S., Carter, A. J., Golya, D. A.,
Hardy, G. E., Haynes, C. E., Rick, J. E., Shapiro, D. A., & West, M.
A. (1997). Minor psychiatric disorder in NHS trust staff: Occupa-
tional and gender differences. British Journal of Psychiatry, 171,
519-523. doi:10.1192/bjp.171.6.519
Winwood, P., Lushington, K., & Winefield, A. H. (2006a) Disentan-
gling the effects of psychological and physical work demands on
sleep and recovery within a sample of Australian nurses. Journal of
Advanced Nursing, 56, 679-689.
doi:10.1111/j.1365-2648.2006.04055.x
Winwood, P. C., Lushington, K., & Winefield, A. H. (2006b). The con-
tribution of age, domestic responsibilities and shiftwork to work-re-
lated fatigue and recovery within a sample of full-time Australian
nurses. Journal of Advanced Nursing, 56, 438-449.
doi:10.1111/j.1365-2648.2006.04011.x