Social Science & Medicine 75 (2012) 2250e2257

Contents lists available at SciVerse ScienceDirect

Social Science & Medicine

journal homepage: www.elsevier.com/locate/socscimed

The impact of migratory separation from parents on the health of adolescents in

the Philippines
Chantal Smeekens a, *, Margaret S. Stroebe a, b, Georgios Abakoumkin c
a
Department of Clinical & Health Psychology, Utrecht University, P.O. Box 80140, 3508 TC Utrecht, The Netherlands
b
Department of Clinical Psychology & Experimental Psychopathology, University of Groningen, The Netherlands
c
Department of Early Childhood Education, University of Thessaly, Volos, Greece

a r t i c l e i n f o a b s t r a c t

Article history: In the Philippines, as in several other low and middle-income countries in the world, it is usual for
Available online 1 September 2012 parents to leave the country to work abroad in order to improve the situation of their children who
remain behind. However, stressful life events such as the separation from a parent are known to have
Keywords: a severe impact on physical and mental health of children. This study, conducted in 2008e2009, explored
Philippines health consequences of migratory separation for remaining-behind adolescent children, comparing them
Health
with those whose parents remained at home. Participants were 205 high school students from the
Adolescent
Philippines. It was found that adolescents with a parent abroad, particularly the mother, reported poorer
Health psychology
Family
physical health than those with both parents at home, while socioeconomic status did not have impact.
Labour migration The parenteabroad adolescents reported a high level of missing their parent(s) and felt emotionally
Cognitive stress model lonelier than the parent-at-home group. Emotional loneliness and stress due to parental absence were
Separation reactions associated with poorer health. Avoidant coping appeared to moderate the parental absence-health
relationship. Paradoxically, it seems that, although many parents work abroad to improve the lot of
their children, the latter suffer emotional stress and physical health detriments. While Lazarus and
Folkman’s (1984) cognitive stress model is generally applicable for migratory separation, cultural
aspects need attention, both in theoretical implementation and interpretation. Limitations and impli-
cations are further discussed.
Ó 2012 Elsevier Ltd. All rights reserved.

Introduction through migratory separation affects the health of children left

behind: are there negative health consequences following this type
Stressful life events are known to have severe health implica- of separation too?
tions for those experiencing them (Folkman, 2011; Lazarus & Migratory separation can be defined as the relocation abroad of
Folkman, 1984; Miller, Chen, & Cole, 2009). For adolescents, parents entering a foreign work force (cf. Pottinger, 2005). One
family issues are a main source of stress (Brodzinsky et al., 1992). focus of investigation in this latter area of research has been on the
Poverty, divorce of parents and the experience of loss or absence of people themselves who go abroad for long periods of time, away
a parent have frequently been identified as predictors of mental from their loved ones and homeland (e.g., De Castro, Gee,
health problems among children and adolescents (e.g., Grant et al., & Takeuchi, 2008; Lu, 2010). However, as indicated above, chil-
2006). Likewise, the extreme case of separation from a loved person dren too can suffer from various types of separation, and we need to
through his or her death has been found to impact strongly on both establish the impact on these young people of the absence of
mental and physical health of survivors (Miller et al., 2009; Stroebe, a parent living and working abroad. Indeed, being separated from
Schut, & Stroebe, 2007). In the context of such findings, it becomes a parent as a result of such labour migration is an increasing issue
important to investigate whether temporary loss of parents for young people in low- and middle-income countries in
contemporary society, and yet this topic has received far less
scientific attention than the other types of loss illustrated above.
* Corresponding author. Present address: The Primary Care Unit, Department of Little is known about whether negative consequences of separation
Public Health & Primary Care, University of Cambridge, Strangeways Research undermine the potential advantages due to parental employment
Laboratory, Worts Causeway, Cambridge CB1 8RN, United Kingdom. Tel.: þ44 1223
and financial improvement, or what factors may be associated with
742089.
E-mail addresses: [email protected] (C. Smeekens), [email protected] high risk of health detriments for young people left behind. The aim
(M.S. Stroebe), [email protected] (G. Abakoumkin). of the current study was therefore to investigate whether

0277-9536/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.socscimed.2012.08.025
 C. Smeekens et al. / Social Science & Medicine 75 (2012) 2250e2257 2251

adolescents in the Philippines whose parents work overseas differ indeed negative mental and physical health outcomes among left-
in health from adolescents whose parents are both living with behind children of parents who have gone to another country or
them, focussing on under-researched physical health. We also try to area to work. Detrimental effects on mental health have been better
identify those who might be at particular risk of poor adaptation to documented than those on physical health. For example, studies of
separation. within-country migration in China (Liu, Li, & Ge, 2009; Liu, Sun,
In the following sections, we first describe the situation in the Zhang, Wang, & Gua, 2010; Zhao et al., 2009) examined the
Philippines, indicating why it is useful to conduct a study of mental health impact of parental labour migration to urban areas in
migratory separation in this country. Then we summarize the China on children remaining behind in rural areas. Children with
evidence that is available on the health consequences of this type of a migrated parent reported poorer mental health than their coun-
loss. To understand the impact of migratory separation more terparts with both parents at home. The results of a matched
precisely, it is important to consider possible mediators and control-group study conducted in Jamaica suggested a relationship
moderators (“risk factors”) of its link with ill-health. A theoretical between the separation of migrated parents and their 9e10 year
model derived from cognitive stress theory (Folkman, 2011; old children and children’s experience of stress, grief and loss
Lazarus & Folkman, 1984) provides a framework for systematic (Pottinger, 2005). It would seem that migratory separation of
examination of such variables. Therefore, this model and its a parent has an impact across a range of different mental health
implications for our study are then described, which leads to the consequences.
formulation of hypotheses. Regarding physical health, the well-established relationship
between stressors and health outcomes (Folkman, 2011) would
Migratory separation in the Philippines suggest negative outcomes for those with parents abroad. However,
findings have been conflicting. Left-behind children from parents
There are good reasons to study the impact of migratory separa- who migrated within China reported poorer health-related quality
tion on young people in the Philippines. The Philippines is one of the of life (Jia, Shi, Cao, Delancey, & Tian, 2010), were more overweight,
largest labour migrant-sending countries in the world. It is estimated had lower height and ate less-healthy diets than those with parents
that a quarter of under-aged people in this country are living sepa- at home (Gao et al., 2010). However, Nguyen, Yeoh, and Toyota
rated from their parents because the latter are working as Overseas (2006) reviewed studies on migratory separation in Asia and re-
Filipino Workers (OFW) (Parreñas, 2006; Reyes, 2008; Scalabrini ported contradictory findings; sometimes the migratory separation
Migration Center, 2004). In the Philippines, a relatively poor of a family member led to poorer physical health outcomes for
country, the main reason for parents to work abroad is to improve the those left behind, while in other studies it was associated with
economic situation of their remaining-behind families. But is this an better health outcomes. In a longitudinal study in rural Mexico
advantage or a disadvantage? On the one hand, financial remittances migratory separation from a father was found to be related to child
could make education, health care and nutritious food more acces- illness (Schmeer, 2009), while Kanaiaupuni and Donato (1999)
sible and may help to improve the left-behind children’s health found that for young children in Mexico, financial remittances
situation (Bryant, 2005; Garcia Dungo, 2007; Scalabrini Migration compensated for maternal absence and provided access to better
Center, 2004; Yang, 2004). Most of the relevant evidence is based food and health. While providing useful leads, it is evident that
on cross-sectional surveys (although some of the studies are purely research has provided limited information so far on the mediators
descriptive). It has been reported that children in the Philippines with and/or moderators of the relationship between migratory separa-
a parent abroad perceived themselves as healthier (University of the tion and health.
Philippines, Tel Aviv University, & Kaibigan, unpublished manuscript)
and suffered less often from illnesses than other children (Asis, Intervening variables: a theoretical perspective
2006a). This could possibly be explained by a growth in socioeco-
nomic status through financial remittances, as socioeconomic status The picture that emerges from studies in the Philippines and
(SES) is positively associated with physical health (Cohen, Kaplan, & other countries regarding the consequences of parental migration
Salonen, 1999; Miller et al., 2009). is unclear. Socioeconomic growth may lead to better physical
On the other hand, concerns have also been expressed that this health while the psychological difficulties of migratory separation
‘culture of migration’ (Asis, 2006b; Yang, 2004) may not bring solely may cause poorer outcomes. In our view, a useful strategy to clarify
positive effects. Left-behind OFW children generally miss their the complex relationship between migration and health is to start
absent parent (Garcia Dungo, 2007; Scalabrini Migration Center, from a theoretical perspective which systematically describes the
2004; University of the Philippines et al., unpublished manuscript) relevant variables in adaptation to stressful life events. Cognitive
and emotional, psychosocial and behavioural problems have been stress theory (Folkman, 2011; Lazarus & Folkman, 1984) provides
reported (Garcia Dungo, 2007; Parreñas, 2002, 2006; Scalabrini a taxonomy to organize research and test relationships between the
Migration Center, 2004), in particular in cases of maternal absence occurrence of a stressor (in our case, separation), processes of
(Asis, 2006a; Battistella & Conaco, 1998; Garcia Dungo, 2007; coping and appraisal (e.g., how migratory separation is evaluated
Parreñas, 2006). At the same time, some families appear to have by adolescents), and outcome (most importantly here, physical
fewer problems (Battistella & Conaco, 1998). Furthermore, because health). Slavin, Rainer, McCreary, and Gowda (1991) developed an
problems may also occur among non-OFW children and adoles- adaptation of the stress process model with attention to culturally-
cents, it is important to include control groups of non-OFW. It relevant dimensions. Using this, it is possible to document how
remains crucial to establish whether the distress of parental absence cultural characteristics might play a role in the appraisal process
outweighs the advantages of income growth for young people left and health outcomes in the Philippine setting. Thus, Fig. 1 shows
behind in the Philippines (cf. Rossi, 2008). the stress model adjusted to migratory separation in the
Philippines.
Health consequences of migratory separation: evidence from other Key variables can be seen in Fig. 1. In addition to the definition of
countries the stressful event and outcomes that have featured already
above, the Figure includes three components that can be integrated
Evidence from other countries is based mainly on cross- in an empirical investigation, as intervening variables. These
sectional studies. These investigations suggest that there are include primary and secondary appraisal processes, and specific
2252 C. Smeekens et al. / Social Science & Medicine 75 (2012) 2250e2257

Potentially stressful event: Primary appraisal: Adaptional outcomes:

Parental overseas migration Socioeconomic status Physical health
______________________ Gender of parent abroad ______________________
______________________
Cultural context: Cultural context:
The Philippines is one of Cultural context: Secondary appraisal: Tendency to somatization in
the largest migrant-sending Family characteristics Emotional stress Asian cultures
countries in the world Gender roles Loneliness Views on health
Poverty in the Philippines Avoidant coping Lifestyle in the Philippines
______________________

Cultural context:
Coping style
Value of family and
togetherness

Fig. 1. The cognitive stress model adjusted to migratory separation in the Philippines.

situational/environmental factors. We have focused our investiga- between the stressful event of separation and the outcome of poor
tion on those variables that emerged from qualitative research in physical health.
the Philippines and the general scientific literature in this area as
most central. Primary appraisal refers to the assessment of the Aims of the current study
event: in our case, appraisal in terms of missing the parent is highly
relevant. Secondary appraisal has to do with assessment of one’s The purpose of the current study in the Philippines was to gain
ability to cope with the stressor (Lazarus & Folkman, 1984). This understanding of the consequences for adolescents of having either
process can reduce or strengthen the outcomes of stress, relating one or both parents working abroad (classified here as the Overseas
here to physical health. One particular coping style, namely, avoi- Filipino Worker group: OFW). In particular, given the paucity of
dant coping, has been strongly associated with negative physical well-controlled quantitative research and its potential importance
health outcomes (e.g., Ruchkin, Eisemann, & Hägglöf, 2000) also for for adjustment among young persons, the focus here was on the
adolescents (Kardum & Krapic, 2001). Avoidance-focused coping relationship of such parental absence with physical health, by
strategies could be particularly relevant in collectivistic cultures comparing the OFW group with adolescents whose parents were
with a less direct way of controlling stressful situations, as in the both living with them in the Philippines (non-OFW). Furthermore,
Philippines (Chun, Moos, & Cronkite, 2007). A distinction can be our investigation was designed to examine a number of key vari-
made between cognitive and behavioural avoidant coping, ables that we expected to mediate or moderate the relationship
whereby cognitive avoidance mainly involves strategies such as not between migratory separation and physical health. Thus, the
thinking about the problem (e.g., “Trying not to think of the absent following hypotheses were formulated on migratory separation
parent”), while behavioural avoidance involves behaving in and physical health: (1) We expected the OFW adolescents to have
a certain way that avoids the problem (e.g., “Getting away from poorer physical health compared to non-OFW group. (2) We expect
missing him/her by doing something distractive with friends”) two variables to have a role in the relationship between OFW status
(Brodzinsky et al., 1992). and physical health: namely (a) socioeconomic status would
Furthermore, in the situation of OFW adolescents, appraisals of moderate this relationship (with the poorest health outcomes for
loneliness might strongly influence outcomes. Children of Filipino the low SES OFW group and the best health for the high SES non-
migrants reported higher levels of loneliness (Asis, 2006a; OFW group); and (b) within the OFW group maternal absence
Battistella & Conaco, 1998). Weiss (1973) described loneliness as (OFW with a mother abroad or both parents abroad) would be
a response to the absence of some specific social provision. Family associated with poorer physical health outcomes as compared to
members in the Philippines provide alternative social support to paternal absence. Further hypotheses could be derived from
that of parents, which might lead to less loneliness (Reyes, 2008). cognitive stress theory as described above (Lazarus & Folkman,
However, this may not protect adolescents fully from feeling 1984): (3) It was expected that (a) missing the absent parent
lonely. Weiss’ (1973) division between loneliness of social isola- more and (b) perceiving the migratory separation as more stressful
tion (the lack of provision in social integration) and emotional would be associated with poorer physical health. (4) In addition,
loneliness (a lack of attachment) is of relevance. Emotional following the theory’s construct of secondary appraisal (Lazarus &
loneliness seems particularly meaningful for the current study. Folkman, 1984), it was hypothesized that (a) feelings of loneliness
Family and ‘togetherness’ are important in the Philippines (social and emotional loneliness) would mediate the relationship
(Medina, 2001) and may be available still to adolescents with between OFW status and physical health. Further, it was hypothe-
a parent abroad, so social loneliness might not be as common. sized that (b) avoidant coping would have a moderating effect on
There may be less compensation for emotional loneliness and this the relationship between OFW status and physical health; the
could impact Filipino adolescents’ physical health outcomes. higher score on avoidant coping, the stronger the relationship of
However, if experienced by children of migrant parents, both OFW status with poor physical health will be.
types of loneliness are expected to have an impact on health
outcomes. Method
Finally, a situational/environmental factor that may help unravel
the positive impact of financial gains versus negative consequences Participants
of separation is socioeconomic status. As discussed earlier, higher
socioeconomic status seems to protect against negative health Initially, 267 adolescents filled in the questionnaire. Adolescents
consequences. Therefore, it appeared reasonable to examine who had a parent who passed away, whose parents were divorced or
socioeconomic status as a possible moderator of the relationship had a parent worked in a different area in the Philippines were
 C. Smeekens et al. / Social Science & Medicine 75 (2012) 2250e2257 2253

excluded from the study (n ¼ 44). Also, those with poor knowledge Measures
of English (n ¼ 13) were excluded from the study (based on a ques-
tion about their fluency in reading and writing) as well as those with OFW category and background: Questions were asked about
unreliable questionnaires (n ¼ 5) (e.g., lots of missing data). overseas labour migration of parents and the adolescents’ back-
Participants were 205 Filipino adolescents aged 13e18 ground (i.e., ethnic group, age, gender), to use as control variables.
(M ¼ 14.58, SD ¼ 1.04) years, with 65 males and 139 females Physical health: Physical health was measured with the Physical
(one participant did not indicate gender). One hundred and eight of Health Questionnaire (PHQ; Schat, Kelloway, & Desmarais, 2005).
them formed the OFW group and had one or both of their parents The scale consists of fourteen items that assess somatic symptoms
working abroad. Fifty nine participants in the OFW group had experienced in the last few weeks, measured on a 7-point Likert
a mother, 40 had a father, and 9 had both parents working abroad. scale (1 ¼ Not at all, 7 ¼ All the time), a ¼ .72.
The control group (non-OFW) consisted of 97 adolescents whose Socioeconomic status: School type (public or private) was taken
parents lived with them. Of the adolescents with a parent abroad, as an indicator for socioeconomic status. In the Philippines, private
73 attended a private school and 35 a public school. The schools schools are considered to be better (Battistella & Conaco, 1998), and
were located in the rural municipality of Isabela; a northern financial resources are often used to send children to more
province of the Philippines. Among the control group these expensive private schools (Bryant, 2005). In the rural Isabela
numbers were 72 and 25 respectively. Participants came from province, socioeconomic status (SES) is not accurately measurable
different ethnic groups, with the majority being Ibanag, but also by income, as many inhabitants live from their own agricultural
including Tagalog and Illocano. products. Moreover, most adolescents do not know the income of
their parents. Furthermore, a control measure showed that
Procedure adolescents attending the private school had significantly more
(M ¼ 6.15, SD ¼ 1.97) appliances (1e10) at home than adolescents
Data collection took place late 2008 to early 2009. The study was attending a public school (M ¼ 4.23, SD ¼ 2.05), t(202) ¼ 1.92,
conducted in accordance with general ethical principles of Utrecht p < .001.
University, The Netherlands. In addition, adjustments concerning Maternal absence: The OFW group was asked about the gender
participants’ consent as well as permissions to conduct the study of the parent(s) abroad, forming a group with only the father
were necessary in order to adapt to the local situation in the abroad and a group with only the mother and both parents abroad.
Philippines. The study was approved by the Cagayan Valley Pro- Missing parents: A single-item measure was chosen to obtain an
gramme on Environment and Development (CVPED) of Isabela overall idea of the extent that adolescents missed their absent
State University e Cabagan. Permission to conduct the research was parent. As it was expected that the adolescents generally miss their
obtained from the mayor of the municipality, local authorities and parent to some degree and that social desirability could have an
the school principals. Finally, guidance counsellors of the respective impact, there were 4 ‘Yes’ answer categories (‘Yes, a lot’, ‘Yes’, ‘Yes,
schools were asked for their cooperation. The guidance counsellors, a bit’, ‘Yes, a little bit’), and ‘No’.
who volunteered to cooperate with the researcher, approached the Perceived stress (due to missing the parent): The Homesickness
adolescents and asked them whether they would be willing to Questionnaire (HQ) (Archer, Ireland, Amos, Broad, & Currid, 1998)
participate in a study on “your health, your family life and your was adapted to the situation of migratory separation, to measure
habits”. It was made clear to them both by the guidance counsellor stress due to missing the parent. The scale measures features
and later by the researcher that they were free to refuse if they did associated with separation and loss and consists of 16 items (e.g., “It
not wish to participate. upsets me if I am unable to contact my parent who is working in
Concerning the practice followed with regard to participants’ another country”). Answers were given on a 5-point Likert scale
consent: we would like to mention that it is not customary in the (1 ¼ Not at all, 5 ¼ Always). The scale had a Cronbach’s a of .67 after
Philippines to ask for (written) parental consent. Moreover, many deletion of one item.
of the adolescents’ parents e or guardians e live and work far Loneliness: Loneliness was measured by four items (adopted
away, and it is also to be expected that some parents may not be from Stroebe, Stroebe, Abakoumkin, & Schut, 1996), using a 5-point
able to read and write. Therefore, the procedure was adopted Likert scale (1 ¼ Never, 5 ¼ Very often). Two items measured
whereby adolescents themselves, with the aid of guidance coun- Emotional Loneliness (e.g., “I often feel lonely even when I am with
sellors, decided whether to participate. Guidance counsellors are other people”, a ¼ .44) and two items measured Social Loneliness
at the schools for the adolescents’ interests, being “confidantes” (e.g., “I have a really nice set of friends”, a ¼ .75).
for them rather than authority figures. In line with this role, they Avoidant coping: Avoidant coping was assessed with the Coping
acted in the interest of the adolescents and they did not put any Scale for Children and Youth (CSCY; Brodzinsky et al., 1992). The
pressure on them to participate. Overall, this procedure was fol- scale consists of 29 items that measure coping styles that children
lowed to both preserve participants’ rights and to conform to local and youth may use when having a problem. In the current study
customs. only Cognitive Avoidance (11 items) and Behavioural Avoidance (6
The individuals who agreed to participate were invited to take items) were used, since avoidant coping is particularly related to
a seat in a classroom that was reserved for the study. There, the negative health outcomes (Ruchkin et al., 2000). A 5-point Likert
informed consent and short instructions for filling in the ques- scale was used (1 ¼ Never, 5 ¼ Always). A sample item of Cognitive
tionnaire were given. Both were explained verbally in English and Avoidance is “I put the problem out of my mind” (a ¼ .60). A sample
Ibanag and could be read in the English introduction letter that was item of Behavioural Avoidance is “I tried to get away from the
attached to the questionnaire. It was chosen to administer the problem for a while by doing other things” (a ¼ .82).
questionnaire in English. In the area, various different languages are
spoken, but students are not necessarily able to understand these Results
languages in written form, as school materials are written in
English. In fact, English is one of the official languages in the First, we examined whether gender and age of the adolescents
Philippines and it is understood well by most. Participants were were related to the variables physical health and migratory sepa-
given the opportunity to ask for verbal translations in their own ration of a parent. Age correlated with physical health (r ¼ .22,
language (Tagalog, Ibanag or Illocano). p < .01) and therefore all analyses with the variable physical health
2254 C. Smeekens et al. / Social Science & Medicine 75 (2012) 2250e2257

were controlled for age. Controlling for age did not change the the (secondary appraisal) variables social and emotional loneliness
pattern of the results, unless mentioned otherwise. Physical health mediated the relationship between the absence of a parent and
was not significantly associated with gender, t(202) ¼ 1.82, ns; physical health. First, the mediator (loneliness) was regressed on
hence, data were collapsed across gender for all analyses. Table 1 the independent variable (OFW status). Second, the dependent
presents means and standard deviations of the two groups for all variable (physical health) was regressed on the independent vari-
measures. able (OFW status). And third, the dependent variable was regressed
on the independent variable controlling for the mediator. All vari-
Parental migration and physical health ables were standardized for the analyses.
With regard to the first set, OFW did not predict social loneliness,
A two-way between groups analysis of variance (ANOVA) was b ¼ .13, R2 ¼ .02, F(1,190) ¼ 3.46, ns. Therefore, further regression
conducted to test whether separation from a parent working over- analyses on social loneliness were omitted. Then, a regression
seas (OFW) and socioeconomic status (SES; private school or public analysis was performed and showed that OFW predicted emotional
school) were associated with the adolescents’ physical health. The loneliness, b ¼ .17, R2 ¼ .03, F(1,193) ¼ 5.83, p < .05, with adolescents
variable OFW had a main effect on physical health, F(1,199) ¼ 9.08, with a parent abroad feeling more emotionally lonely than those
p < .01, with adolescents with a parent working overseas reporting with both parents at home. The next regression analysis showed that
poorer physical health (M ¼ 5.09, SD ¼ .78) than adolescents with OFW status, as already reported above, predicted physical health,
both parents at home (M ¼ 5.43, SD ¼ .63). The main effect of SES b ¼ .23, R2 ¼ .05, F(1,201) ¼ 11.30, p < .01. In the last step, when
was not significant, F(1,199) ¼ 2.55, ns. No interaction effect between entering OFW status into the equation while controlling for
the variables was found either, F(1,199) < 1; therefore, SES did not emotional loneliness, the mediator was significant, b ¼ .22,
moderate the relationship between OFW status and physical health. t(191) ¼ 3.18, p < .01, and the relationship between OFW and
Thus, our main hypothesis (1) could be confirmed, as adolescents physical health remained significant as well, b ¼ .20,
with a parent overseas (OFW) reported poorer health than their non- t(191) ¼ 2.86, p < 01; physical health was predicted by both OFW
OFW counterparts. However, Hypothesis 2a regarding the moder- status as well as emotional loneliness, R2 ¼ .10, F(2,191) ¼ 11.07,
ator role of SES was not confirmed. p < .001. This means that emotional loneliness did not substantially
We then divided the OFW adolescents into a group whose affect the relationship between OFW and physical health and was
mother was working abroad (‘mother absent’) and a group for therefore not found to be a mediator. This did not confirm Hypoth-
whom only the father was working abroad (‘only father absent’), esis 4a.
and conducted a one-way ANOVA of OFW status (non-OFW vs. Finally, Hypothesis 4b was examined, namely, that cognitive and
OFW mother absent vs. OFW father absent) on the physical health behavioural avoidant coping would moderate the relationship
scores. This analysis yielded a significant result, F(2,200) ¼ 5.78, between the absence of a parent and physical health. Two regres-
p < .01. As indicated by post-hoc comparisons (Bonferroni), this sion analyses were conducted. In a first step of each regression,
result was due to the difference between the ‘mother absence’ physical health was regressed on the independent variable and the
group (M ¼ 5.07, SD ¼ .75) and the non-OFW group (M ¼ 5.43, moderator (OFW status and the respective coping variable). In
SD ¼ .63), which confirms Hypothesis 2b. a second step, the interaction term was entered into the equation.
All variables were standardized for the analyses.
First, the moderator role of cognitive coping was examined. Both
Cognitive stress theory: testing for mediation and moderation
OFW status, b ¼ .23, t(198) ¼ 3.41, p < .01, as well as cognitive
coping, b ¼ .17, t(198) ¼ 2.54, p < .05, predicted physical health,
We analysed whether adolescents were missing their absent
R2 ¼ .08, F(2,198) ¼ 8.69, p < .01. When the interaction term was
parent, and whether missing the parent and stress due to missing
entered in the second step, both predictors remained significant,
the absent parent was associated with physical health outcomes. It
while their interaction only approached significance, b ¼ .12,
appeared that the majority of the OFW adolescents reported
t(197) ¼ 1.73, p ¼ .09. Nevertheless, this trend was in the pre-
missing their absent parent very much (M ¼ 4.59, SD ¼ 1.01). Due to
dicted direction. As depicted in Fig. 2, panel (a), the relationship
this strongly skewed distribution, the 1-item measure was omitted
between OFW status and physical health tended to be stronger
from further analyses and Hypothesis 3a could not be confirmed.
among participants with higher cognitive coping scores: for those
Then, we examined whether stress due to missing the absent
who applied avoidant cognitive coping to a higher degree, the
parent (adapted Homesickness Questionnaire (HQ); Archer et al.,
absence of a parent appeared to affect their health more than for
1998) would be associated with physical health outcomes. This
those who did not engage so much in avoidant coping.
was indeed the case (r ¼ .31, n ¼ 96, p < .01): More stress expe-
Second, the moderator role of behavioural coping was exam-
rienced by participants was associated with poorer physical health;
ined. Again, both OFW status, b ¼ .20, t(195) ¼ 2.89, p < .01, as
this supports Hypothesis 3b.
well as behavioural coping, b ¼ .21, t(195) ¼ 3.02, p < .01, pre-
Two sets of regression analyses following the procedure sug-
dicted physical health, R2 ¼ .09, F(2,195) ¼ 9.90, p < .001. This time,
gested by Baron and Kenny (1986) were performed to test whether
the interaction was significant, b ¼ .15, t(194) ¼ 2.20, p < .05,
while both predictors remained significant as well, R2 ¼ .11,
Table 1 F(3,194) ¼ 8.35, p < .001. As can be seen in Fig. 2, panel
Means and standard deviations of measures by OFW status. (b), behavioural coping exhibited a similar moderation pattern
Non-OWF OFW as cognitive coping: The more participants applied this
coping strategy, the more did their health suffer when a parent was
M SD M SD
absent.
Physical health** 5.43 .63 5.09 .78
Finally, it should be noted that the above picture did not change
Emotional loneliness* 2.35 .82 2.65 .89
Social loneliness 2.17 .98 2.44 1.05 considerably when controlling for age. The only notable change was
Cognitive avoidant coping 2.90 .60 2.84 .67 that the interaction between OFW status and cognitive coping
Behavioural avoidant coping 2.73 .58 2.87 .61 became significant. Overall then, hypothesis 4b was supported:
Note: Group means comparisons according to t-tests. High avoidant coping made the relationship between OFW status
*p < .05; **p < .01. and physical health stronger as compared to low avoidant coping.
 C. Smeekens et al. / Social Science & Medicine 75 (2012) 2250e2257 2255

Fig. 2. Physical health by OFW status and avoidant coping. Panel (a) cognitive avoidant coping; and panel (b) behavioural avoidant coping.

Discussion social loneliness, possibly because extended families in the

Philippines are close, emotional loneliness was a predictor of
The primary aim of the current study was to explore the physical health outcomes.
relationship between parental migratory separation and physical Following cognitive stress theory (Lazarus & Folkman, 1984), it
health of the remaining-behind adolescent children. Confirming was expected that an avoidant coping style would moderate the
our main hypothesis, adolescents with a parent abroad reported relationship between OFW status and physical health. This was
poorer physical health compared to adolescents with both parents confirmed in our analysis; behavioural avoidant coping had
living with them at home. The difference was still significant when a negative impact on physical health, especially when a parent is
age was controlled for. A major explanation for the current finding working abroad. Cognitive avoidant coping was also a significant
lies within cognitive stress theory (Lazarus & Folkman, 1984). moderator after controlling for age. Interventions directed at
However, although the emotional stress of missing a parent led to altering (avoidant) coping strategies which have been used to deal
poorer physical health among the OFW adolescents, it did not fully with the stress of parental absence might be beneficial to left-
explain the missing-parent effect. Still, there is evidence that behind adolescents.
Asians tend to somatize psychosocial or emotional problems and There were some limitations to our study. A main challenge
this could explain the relation between emotional stress and was to adapt the questionnaire to the adolescents’ situation in the
health (Chun et al., 2007; Wong, Tran, Kim, Van Horn Kerne, & Philippines. The questionnaire was not in the participants’ first
Calfa, 2010). language, which could have led to interpretation issues. An
Additional hypotheses were explored to investigate the rela- example is the interpretation of the answer categories on the
tionship between migratory separation and physical health. Likert scales. In Filipino language there is only one word for ‘often’
Socioeconomic status was expected to improve and lead to better and ‘always’. During pilot testing it became evident that the
physical health (Scalabrini Migration Center, 2004; University of adolescents understood that the answer possibilities were in
the Philippines et al., unpublished manuscript; Yang, 2004). ascending order, from never (1) to always (5) and no problems
However, here socioeconomic status did not relate to the health related to this scale emerged. Nevertheless, cross-cultural differ-
outcomes in adolescents. How could these contradictory findings ences in interpretation of scales and questions could still remain,
between the current and previous studies be explained? First, which makes comparisons with studies from other countries
socioeconomic status may relate more to actual bodily outcomes difficult. Another example is the statement “I decided to stay away
than to scores on the Physical Health Questionnaire, if these scores from people and be by myself”, from the Behavioural Avoidant
reflect somatization of psychological stress rather than actual Coping-subscale. This situation could be more problematic than in
biomedical health states. Second, transnational families may not a western culture, because in the Philippines it is perceived as
always spend their remittances wisely in terms of health; they normal to be together with others often (Medina, 2001). The
might rather buy luxury goods such as cigarettes and soft drinks. limitations in the questionnaire, mainly involving cultural and
In the Philippines, so-called western diseases and problems such language issues, suggest the need for further studies to adapt
as obesity have emerged, in particular among people from higher existing scales and to develop questionnaires from Filipino
income groups (Dahly, Gordon-Larsen, Popkin, Kaufman, & Adair, cultural perspective.
2010). Third, the operationalization of socioeconomic status by Another issue concerns two measures that were used, namely,
using school type might have been rather weak (we return to this assessment of socioeconomic status (SES) by means of school type,
later). We conclude, therefore, that improvement in socioeco- and of the extent that participants missed their parents using
nomic status, due to having a parent working overseas, was not a single-item measure. Regarding SES, a direct income measure
sufficient to compensate for other lacks such as the emotional may seem more appropriate. However, indirect measures like
burden resulting from the absence of a parent. school type might often be applicable or even preferable in some
As hypothesized, the adolescents whose mother worked over- countries or regions, like the Isabela province, where direct
seas reported significantly poorer health outcomes compared to the assessment of income is highly problematic (see also Piovesan,
group with both parents at home. Previous studies from the Padua, Ardenghi, Mendes, & Bonini, 2011). Moreover, in support
Philippines provide support for the view that mothers are impor- of the current measure, school type was related to number of
tant sources of support for physical and emotional health (Medina, appliances at home, a common measure of family assets. The
2001), whose absence could be emotionally stressful for left-behind assessment of missing the parent through a single-item measure
children (Asis, 2006a; University of the Philippines et al., makes it less reliable. The study investigated the experience of
unpublished manuscript). Although there was no difference in migration for a specific segment of the population, particularly with
2256 C. Smeekens et al. / Social Science & Medicine 75 (2012) 2250e2257

respect to socioeconomic status. Filipinos working overseas and Bryant, J. (2005). Children of international migrants in Indonesia, Thailand, and the
Philippines: A review of evidence and policies. Innocenti working paper no. 2005-
adolescents attending high schools are generally not from the
05. Florence: UNICEF Innocenti Research Centre.
poorest families. In the Philippines there are large differences Carlton, B. S., Goebert, D. A., Miyamoto, R. H., Andrade, N. N., Hishinuma, E. S.,
between poor and rich. Among the poorest, socioeconomic Makini, G. K., Jr., et al. (2006). Resilience, family adversity and well-being
improvement might predict greater changes in physical health among Hawaiian and non-Hawaiian adolescents. International Journal of Social
Psychiatry, 52(4), 291e308.
outcomes. Therefore, changes in socioeconomic status might be of Chun, C. A., Moos, R. H., & Cronkite, R. C. (2007). Culture: a fundamental context for
greater significance for the health of adolescents from the poorest the stress and coping paradigm. In P. T. P. Wong, & L. C. J. Wong (Eds.), Handbook
families than for those in our sample. of multicultural perspectives on stress and coping (pp. 29e53). New York: Springer.
Cohen, S., Kaplan, G. A., & Salonen, J. T. (1999). The role of psychological charac-
Although the interpretation that the absence of a parent leads to teristics in the relation between socioeconomic status and perceived health.
poorer physical health of the remaining-behind adolescent chil- Journal of Applied Social Psychology, 29(3), 445e468.
dren seems highly plausible, caution is needed: It is possible that Dahly, D. L., Gordon-Larsen, P., Popkin, B. M., Kaufman, J. S., & Adair, L. S. (2010).
Associations between multiple indicators of socioeconomic status and obesity
the causal relation between migratory separation and physical in young adult Filipinos vary by gender, urbanicity, and indicator used. Journal
health outcomes is the other way around. Poor physical health of of Nutrition, 140(2), 366e370.
children could be a reason for parents to go abroad rather than De Castro, A. B., Gee, G. C., & Takeuchi, D. T. (2008). Job-related stress and chronic
health conditions among Filipino immigrants. Journal of Immigrant Minority
a consequence of migration. Health, 10(6), 551e558.
The current patterns of findings, as well as limitations of this Folkman, S. (2011). The Oxford handbook of stress, health and coping. New York:
investigation, call for a longitudinal follow-up study among left- Oxford University Press.
Gao, Y., Li, L. P., Kim, J. H., Congdon, N., Lau, J., & Griffith, S. (2010). Impact of
behind adolescents, starting before parents’ departure. This
migration on adolescent health and health behaviors in rural southern China.
would give valuable information about the relationship between BMC Public Health, 10(56).
migratory separation and physical health; in particular about Garcia Dungo, N. (2007). The Filipino family left behind: caught in the ‘gains and pains’
causality and changes over time. An extension to other health of labour migration. In Symposium on overseas migration. Symposium conducted at
College of Home Economics. Diliman, Quezon City: University of the Philippines.
variables such as psychosomatic health, health risk behaviour and Grant, K. E., Compas, B. E., Thurm, A. E., McMahon, S. D., Gipson, P. Y.,
nutritional status would be valuable, besides extending the study to Campbell, A. J., et al. (2006). Stressors and child and adolescent psychopa-
include the parents working abroad and remaining behind. A thology: evidence of moderating and mediating effects. Clinical Psychology
Review, 26(3), 257e283.
further step would be to examine factors that may contribute to Jia, Z., Shi, L., Cao, Y., Delancey, J., & Tian, W. (2010). Health-related quality of life of
better physical health and well-being outcomes, such as resilience “left-behind children”: a cross-sectional survey in rural China. Quality of Life
of the left-behind adolescents and family support (e.g., Carlton Research, 19(6), 775e780.
Kanaiaupuni, S., & Donato, K. (1999). Migradollars and mortality: the effects of
et al., 2006), aspects that seem likely to be strong features within migration on infant survival in Mexico. Demography, 36(3), 339e353.
Filipino society. Kardum, I., & Krapic, N. (2001). Personality traits, stressful life events, and
In conclusion, the current study indicated that labour migration coping styles in early adolescence. Personality and Individual Differences,
30(3), 503e515.
and consequently, migratory separation from parents, is not always Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal and coping. New York: Springer.
associated with better physical health outcomes for their left- Liu, L. J., Sun, X., Zhang, C. L., Wang, Y., & Gua, Q. (2010). A survey in rural China of
behind adolescent children. Paradoxically, this parental absence parent-absence through migrant working: the impact on their children’s self-
concept and loneliness. BMC Public Health, 10(32).
was related to poorer physical health among the left-behind
Liu, Z., Li, X., & Ge, X. (2009). Left too early: the effects of age at separation from
adolescents in our study. The adolescents also reported missing parents on Chinese rural children’s symptoms of anxiety and depression.
their parents very much and in our investigation, higher levels of American Journal of Public Health, 99(11), 2049e2054.
emotional stress among them were associated with poorer physical Lu, Y. (2010). Ruraleurban migration and health: evidence from longitudinal data in
Indonesia. Social Science & Medicine, 70(3), 412e419.
health. The findings provide some insights to the ambiguous Medina, B. T. G. (2001). The Filipino family. Diliman, Quezon City: The University of
character of labour migration and the difficulties that it brings for the Philippines Press.
left-behind adolescent children in the Philippines. Miller, G., Chen, E., & Cole, S. W. (2009). Health psychology: developing biologically
plausible models linking the social world and physical health. Annual Review of
Psychology, 60, 501e524.
Acknowledgements Nguyen, L., Yeoh, B. S. A., & Toyota, M. (2006). Migration and the well-being of the
‘left-behind’ in Asia. Asian Population Studies, 2(1), 37e44.
Parreñas, R. S. (2002). Human sacrifices: what happens when women migrate and
We gratefully acknowledge the help of the following in the leave families behind? The Women’s Review of Books, 19(5), 16.
Philippines: Prof. C. Conaco (University of the Philippines e Dili- Parreñas, R. S. (2006). Children of global migrationIn Transnational families and
man), the CVPED (Cagayan Valley Programme on Environment and gendered woes, . Quezon City: Ateneo de Manila University Press.
Piovesan, C., Padua, M. C., Ardenghi, T. M., Mendes, F. M., & Bonini, G. C. (2011). Can type
Development), the participating high school students, and Maricel of school be used as an alternative indicator of socioeconomic status in dental
Angolluan (interpreter). Also, we thank the anonymous reviewers carried studies? A cross-sectional study. BMC Medical Research Methodology, 11(37).
for their thoughtful comments. Pottinger, A. M. (2005). Children’s experience of loss by parental migration in inner-
city Jamaica. American Journal of Orthopsychiatry, 75, 485e496.
Reyes, M. M. (2008). Migration and Filipino children left behind: a literature review.
References Retrieved from. http://www.unicef.org/philippines/mediacentre_10202.html
Last accessed on 18.02.12.
Archer, J., Ireland, J., Amos, S. L., Broad, H., & Currid, L. (1998). Derivation of Rossi, A. (2008). The impact of migration on children in developing countries.
a homesickness scale. British Journal of Psychology, 89(2), 205e221. Migrant youth & children of migrants in a globalized world. In Synthesis paper
Asis, M. M. B. (2006a). Living with migration: experiences of left-behind children in for Bellagio conference, Italy. Retrieved from http://crcw.princeton.edu/
the Philippines. Asian Population Studies, 2(1), 45e67. migration/papers.asp. Last accessed on 18.02.12.
Asis, M. M. B. (2006b). The Philippines’ culture of migration. Retrieved from Ruchkin, V. V., Eisemann, M., & Hägglöf, B. (2000). Coping styles and psychosomatic
http://www.migrationinformation.org/USFocus/display.cfm?ID¼364 Last problems: are they related? Psychopathology, 33(5), 235e239.
accessed on 18.02.12. Scalabrini Migration Center. (2004). Hearts apart: Migration in the eyes of Filipino
Baron, R. M., & Kenny, D. A. (1986). The moderatoremediator variable children. Manila: Planetboks.
distinction in social psychological research: conceptual, strategic and Schat, A. C. H., Kelloway, E. K., & Desmarais, S. (2005). The physical health ques-
statistical considerations. Journal of Personality and Social Psychology, 51(6), tionnaire (PHQ): construct validation of a self-report scale of somatic symp-
1173e1182. toms. Journal of Occupational Health Psychology, 10(4), 363e381.
Battistella, M., & Conaco, G. C. (1998). The impact of labour migration on the chil- Schmeer, K. (2009). Father absence due to migration and child illness in rural
dren left behind: a study of elementary school children in the Philippines. Mexico. Social Science & Medicine, 69(8), 1281e1286.
Journal of Social Issues in Southeast Asia, 13(2), 220e241. Slavin, L. A., Rainer, K. R., McCreary, M. L., & Gowda, K. K. (1991). Toward a multicultural
Brodzinsky, D. M., Elias, M. J., Steiger, C., Simon, J., Gill, M., & Hitt, J. C. (1992). Coping model of the stress process. Journal of Counseling & Development, 70(1), 156e163.
scale for children and youth: scale development and validation. Journal of Stroebe, M., Schut, H., & Stroebe, W. (2007). Health consequences of bereavement:
Applied Developmental Psychology, 13(2), 195e214. a review. The Lancet, 370(6903), 1960e1973.
 C. Smeekens et al. / Social Science & Medicine 75 (2012) 2250e2257 2257

Stroebe, W., Stroebe, M., Abakoumkin, G., & Schut, H. (1996). The role of loneliness Wong, Y. J., Tran, K. K., Kim, S. H., Van Horn Kerne, V., & Calfa, N. A. (2010). Asian
and social support in adjustment to loss: a test of attachment versus stress Americans’ lay beliefs about depression and professional help seeking. Journal
theory. Journal of Personality and Social Psychology, 70(6), 1241e1249. of Clinical Psychology, 66(3), 217e332.
University of the Philippines, Tel Aviv University, & Kaibigan. The study on the Yang, D. (2004). How remittances help migrant families. Retrieved from.
consequences of international contract labor migration of Filipino parents on their http://www.migrationinformation.org/USfocus/display.cfm?ID¼270 Last
children. Final scientific report to the NetherlandseIsrael development research accessed on 18.02.12.
programme, Unpublished manuscript. Zhao, K. F., Su, H., He, L., Wu, J. L., Chen, M. C., & Ye, D. Q. (2009). Self-concept and
Weiss, R. S. (1973). Loneliness: The experience of emotional and social isolation. mental health status of ‘stay-at-home’ children in rural China. Acta Paediatrica,
Cambridge, MA: MIT Press. 98(9), 1483e1486.