306-321_ISP_095117.indd

Background: The emigration of Mexicans to the USA has increased in the last
decades, and little is known about the effect of this on the mental health of those
who stay behind.
Aims: To  evaluate  the  association  of  emigration  of  husband  and  depressive
symptoms (DS) among women who stay in Mexico. We also tested the hypothesis
that  the  husband’s  migration  would  increase  the  woman’s  autonomy,  which  in
turn would decrease DS.
Methods: A survey was conducted in a rural area in Mexico. Participants (n = 418)
were  selected  through  probabilistic  sampling  in  three  stages:  localities,  house-
holds  and  individuals.  DS were  evaluated  using the  Centre  for Epidemiological
Studies-Depression (CES-D) scale.
Results: Having a partner in the USA was associated with higher odds of scoring
above the cut-off point in CES-D (OR 3.77, 95% CI 1.92–7.43). Economic auto-
nomy was also associated with DS (OR 1.45, 95% CI 1.04–2.02).
Conclusion: Migration of husband was associated with DS among women. The
construct of autonomy and its operational deﬁnition should be further explored.

Key words: depression, migration, personal autonomy, women, Mexico, rural

INTRODUCTION

International migration in search of labour has increased in recent years, prompted by economic
differences between countries. People from less-developed economies migrate – either temporarily
or permanently – to those countries where labour force is needed, thus bringing a diversity of
public health consequences. Emigration from Mexico to the USA has been increasing in the last
decades. In 2004, according to Mexico’s National Population Council (CONAPO, web page), 10.2
million Mexicans were permanently established in the USA.

The effect of migration on mental health has been studied in a number of countries and for

diverse contexts. Schizophrenia and common mental disorders are reported to be highly prevalent
among migrants (Bhugra, 2004). But a less explored area has been that of the consequences of
emigration on the mental health of those who stay in the communities of origin (Salgado de Snyder
et al

., 2007). The migration of the father has been reported to result in psychological difﬁculties

Singapore and Washington DC) www.sagepublications.com Vol 55(4): 306–321 DOI: 10.1177/0020764008095117

by Malgorzata Czyzewska on September 9, 2010

isp.sagepub.com

Downloaded from

BOJORQUEZ ET AL.: DEPRESSIVE SYMPTOMS AMONG WOMEN FROM A MEXICAN RURAL AREA

307

for children and teenagers (Aguilera-Guzman et al., 2004; Pottinger, 2005). On the other hand,
the effect of migration to the USA on drug use in rural communities in Mexico was found to be
mediated by social cohesion and cultural identity, thus showing that the impact of this phenomenon
depends on a number of social circumstances (Wagner et al., 2002). Qualitative interviews in Ireland
with people who had experienced the emigration of a close friend or relative showed responses
of sadness and grief (Sheehan, 1996). Women who stay behind experience feelings of sorrow and
uncertainty about the chances of eventual family reuniﬁcation (McGuire & Martin, 2007).

Massive Mexican migration to the USA began during the 1940s, and was predominantly male for

some decades. This pattern has been changing and every day more women cross the border either
to rejoin their families or in search of work (CONAPO, web page). Nevertheless, in the case of
established couples it is still common for women to stay in Mexico, taking charge of house and
children; and, for extended periods of time, these women keep in touch with the absent men
only by telephone. Women in this situation experience their partner’s migration as stressing life
events (Salgado de Snyder, 1996). And as life events have consistently been associated with the
onset of depressive symptoms (de Graaf et al., 2002; Kendler et al., 2002; Olstad et al., 2001),
an association could be expected between partner’s migration and depressive symptoms (DS)
among these women.

It has also been suggested that male migration allows an increase in women’s autonomy.

Women who stay in local communities while their husband migrates are reported to experience an
increase in decision-making power and in their participation in collective action in the family as
well as in the community (Chant, 1992; González de la Rocha, 1989; Hondagneu-Sotelo, 1994;
Zárate,  2000). As  lack  of  autonomy  could  be  a  gender-related  risk  factor  for  DS  (Patel  et al.,
2006;  Silverstein  &  Perlick,  1995; WHO,  2000),  an  increase  in  female  autonomy  following
the partner’s migration might lessen its negative effects on mental health among those women
who stay behind.

Another social factor related to DS is social support, which has been suggested as a buffer for the

impact of stressing life events (Olstad et al., 2001; Paykel, 1994; Thoits, 1995; Wade & Kendler,
2000). Also, socio-economic and education levels have repeatedly shown an inverse relation to
DS (Kahn et al., 2000; Medina-Mora et al., 2005; Skapinakis et al., 2006; Weich et al., 2001).
Age and number of children have also been proposed as related to DS, although this relationship is
less clear (Brown & Harris, 1978; Piccinelli & Wilkinson, 2000).

In this study we tested these associations with a conceptual frame schematized in Figure 1. This

follows on Cervantes and Castro (1985), according to whom the effect of stressors is lessened
by compensating and mediating factors, both internal and external. The ﬁrst hypothesis in this
study was that partner’s international migration (a stressing life event) would be associated with
more DS among women staying in the communities of origin. The second hypothesis was that
women with a migrant partner would beneﬁt from more autonomy, which in turn would be related
to less DS.

METHOD

The information was collected as part of a wider project to study the associations between
international migration and health in a poor rural area in Mexico (Proyecto Mixteca). A cross-
sectional study was conducted in three municipalities with high scores in the marginality

by Malgorzata Czyzewska on September 9, 2010

isp.sagepub.com

Downloaded from

308

INTERNATIONAL JOURNAL

SOCIAL PSYCHIATRY 55(4)

index of CONAPO and in the migratory-intensity index by the same institution (CONAPO, 2002).
The three municipalities were in the Mixteca Baja – an area located in the states of Guerrero, Oaxaca
and Puebla. All these municipalities were essentially rural, with all but one of their localities having
fewer than 2,500 inhabitants, according to the 2000 census (INEGI, web page).

Sampling design

The sample needed for the Proyecto Mixteca was calculated as 708 households. The sampling design
included stratiﬁcation on municipalities, with a sample size for each municipality proportional to
its population. Within municipalities, localities were selected with a probability of selection pro-
portional to the expected number of households, giving localities with more households a higher
probability of selection. In this way, a list of 21 localities was obtained. Within the localities, a
systematic sampling by blocks – and households within blocks – was conducted.

In each household one woman was selected for interview. Following the inclusion criteria of

Proyecto Mixteca

, women had to be of reproductive age (15–49 years) and having delivered at least

one live-born child. When more than one woman in the household was eligible for interviewing,
the selection was made on the basis of the closest birth date. If the selected woman was not present,
one more visit was made to the house. If the second visit failed to achieve an interview – or when
the woman did not wish to participate – it was coded as ‘No Response’ and no other woman was
selected in the household.

All participants were informed of the study objectives and the voluntary nature of their partici-

pation was remarked. A signed informed-consent form was obtained before conducting the inter-
views. The project was approved by the Ethics Committee of the National Institute of Public
Health, Mexico.

Instruments

A series of questions regarding the characteristics of the premises and the family were asked in
each household. The respondent of these questions could or could not be the woman of reproductive

Figure 1. Conceptual framework

Based upon Brown & Harris, 1978; Cervantes & Castro, 1985

by Malgorzata Czyzewska on September 9, 2010

isp.sagepub.com

Downloaded from

BOJORQUEZ ET AL.: DEPRESSIVE SYMPTOMS AMONG WOMEN FROM A MEXICAN RURAL AREA

309

age surveyed. In 33% of the households, the same woman answered the Household Questionnaire
and the Women’s Questionnaire.

Interviews were conducted by previously trained persons, with at least a junior high school edu-

cational level. Periodical supervision was conducted in the ﬁeld by the research team. The capture
of information from paper to electronic format was supervised through double-capture of a sample
of questionnaires. The instruments for the evaluation of the main variables are described below.

Depressive symptoms

The scale used to measure DS was the Centre for Epidemiological Studies – Depression (CES-D)
(Radloff, 1977), previously validated in Mexico for women in a rural area (Salgado de Snyder &
Maldonado, 1994). CES-D includes 20 items representing six mood and somatic aspects of de-
pression: depressed mood; guilt and worthlessness; helplessness and hopelessness; psychomotor
retardation; loss of appetite; and sleep disturbance. Response options correspond to the number of
days of the previous week that each symptom was present. The possible range of scores is 0 to 60,
with a higher score corresponding to a greater number and frequency of DS. CES-D has shown
both concurrent validity with other scales for the measurement of DS and its ability to discrim-
inate between general population and patients with a diagnosis of depression. It is also sensitive to
improvement of symptoms in depressed patients (Weissman et al., 1977), and reliable according to
test-retest (Bowling, 1995). The suggested cut-off point for clinically signiﬁcant DS is 16 (Weissman
et al

., 1977). The instrument is not intended to provide a diagnosis of depression, but to detect the

presence and intensity of component symptoms of depression. The statistical analysis considered
the score in CES-D both as a numerical variable (total score) and as a binary variable (score equal
or over 16 considered as clinically signiﬁcant DS).

Partner’s migration

Participants were asked about their partner’s habitual place of residence and their place of residence
at the time of the interview. Other questions included frequency and means of contact with partners
in the USA, and if remittances were received. Place of residence was categorized as: living in the
locality; living in other place within Mexico; living in the USA but currently visiting the locality;
living in the USA and staying there at the moment of the interview. A variable of ‘Recent Migration’
was computed as a partner’s ﬁrst departure for the USA occurring during the previous year, when
the partner was presently in the USA. Information to compute this variable was available for
only 68% of women with a migrant partner, so it was not included in the ﬁnal statistical models.
As a sensitivity analysis, models including this variable were also tested. The crude association
between recent migration and DS is also reported.

Autonomy

Autonomy has been deﬁned as the capacity of self-determination after a process of self-reﬂection
(Friedman, 2003). A series of dimensions of autonomy have been proposed as relevant for women:
knowledge (having the information required to choose and act); decision-making power; move-
ment (the capacity to move in and interact with the world outside home and immediate family);
emotional autonomy; economic autonomy (employment and property rights); political or social
autonomy (the capacity to engage in collective action or take part in collective decision making
beyond the private sphere); and corporeal autonomy (the capacity to control one’s own health
and sexuality) (Dyson & Morre, referred to in Pallitto & O’Campo, 2005; Govindasamy &
Malhotra, 1996; Jejeebhoy, 1995). Questions included in the questionnaire assessed autonomy in

by Malgorzata Czyzewska on September 9, 2010

isp.sagepub.com

Downloaded from

310

INTERNATIONAL JOURNAL

SOCIAL PSYCHIATRY 55(4)

four aspects. First, decision-making power was assessed as the participation of women in making
everyday decisions such as how much to spend on food, and major decisions such as buying a
new piece of land or making house improvements. Second, movement autonomy was assessed by
asking women if they had to ask for someone’s permission in order to do activities that ranged from
visiting friends or family members, to looking for a job. Third, political autonomy was evaluated
through the report of participation in organized collective activities, such as neighborhood or other
local organizations, parental or religious groups, and explicitly political organizations. Fourth,
economic autonomy was evaluated through asking women if they had property rights over land or
house, if they received a salary, or if they had been a recipient of credit by a bank or governmental
agency. The questions were based on those in the Encuesta Nacional sobre la Dinámica de las
Relaciones en los Hogares 2003

(National Survey of Household Relationships 2003) (Casique,

2004), and also on work reported by Bloom et al. (2001) and Govindasamy and Malhotra (1996).
Individual items were re-codiﬁed so that a higher score represented more autonomy, and combined
in indexes through robust principal components analysis (PCA) suitable for categorical data
(Kolenikov & Angeles, 2004) with the help of SPSS 13.0 software (SPSS Inc).

Social support

Participants were asked if they did or did not receive moral support, spiritual support, com-
panionship, and alimentary or other material support, and if it came from friends, neighbours or
family. Binary variables were computed for both perceived emotional support (moral, spiritual or
companionship support) and material support (either alimentary or material).

Socio-economic level (SEL)

An index was computed, through robust PCA, from information on housing characteristics (material
of roof and ﬂoors) and services (drinking-water source, electricity), number of people per room,
appliances, cattle and chicken, possession of cultivable land, and vehicle ownership. A higher
score in the index represented a higher SEL.

Educational level

The highest educational level reached was categorized in the following way: without formal edu-
cation, primary school, junior high school, and high school and over.

Statistical analysis

A bivariate analysis was conducted for the association between each independent variable and DS
as a continuous measure. A logistic regression model was adjusted, with clinically signiﬁcant DS
(score at or over 16 in CES-D) as a dependent variable. Goodness of ﬁt was evaluated through
the Hosmer-Lemeshow test (Hosmer & Lemeshow, 1989). In order to account for the sampling
design, the standard errors were adjusted with the Taylor linearized variance estimation using the
statistical software Stata 8.0 (Stata Corporation, 2003).

RESULTS

Out of 708 households to be surveyed, information was obtained in 698 (98.6% response rate); and
a woman was interviewed in 468 households. Of these 468 women, 418 (89%) were married or

by Malgorzata Czyzewska on September 9, 2010

isp.sagepub.com

Downloaded from

BOJORQUEZ ET AL.: DEPRESSIVE SYMPTOMS AMONG WOMEN FROM A MEXICAN RURAL AREA

311

had a permanent partner. Since the purpose of this study was to analyze the association of partner’s
migration and women’s DS, the sample for this report was composed of 418 women (Table 1).

Depressive symptoms

Only 311 of the 418 women answered all the items of CES-D. Their mean score was 11 (SD = 8.8),
with a range of 0–51. The prevalence of DS above the cut-off point was 23.2%. Women who did not
answer all the items were different to those with complete sets of answers: they were less educated,
had a lower SEL score, and had less autonomy and more children.

Partner’s migration

The partners of 142 (34%) of the participants were international migrants. Of them, 48% kept in
touch with their wife at least once a week, and 49% every two to four weeks. The partners of 2% of
the women never communicated with them. Telephone was the means of communication for 98% of
the participants. Remittances were sent by 98% of the partners. At the time of the interview, 27%
of the migrant partners were visiting in the locality. Information on the year of the partner’s ﬁrst
and last migrations was available for 70 women. Of them, 13 were considered as recent migration
(ﬁrst departure in the year previous to the survey, partner still in the USA).

Autonomy

A gender division was observed in autonomy: when compared to men, women show lower decision-
making power regarding major purchasing decisions. Movement autonomy was also limited, as
most participants had to ask for permission before visiting relatives or friends, attending a party or
looking for a job. The minority of them owned property, or had a personal income to support their
economic autonomy (Table 2).

Against the hypothesis, partner’s migration was not associated with scores in the index of

decision-making power, collective participation (political autonomy), or economic autonomy.

Table 1

Characteristics of respondents in three Mixteca-Baja municipalities (n = 418)

Total

Women with migrant

partner (n = 142)

Women with non-migrant

partner (n = 274)

Age (years) (mean, SD)

34.8 (8.3)

32.9 (8.4)

35.8 (8.1)

Number of children (mean, SD)

4.0 (2.4)

3.8 (2.3)

4.1 (2.4)

Educational level (%)
  No formal education
  Primary school
  Junior high school
  High school and over

15.9
65.3
13.0

5.8

14.3
68.6
11.4

5.7

16.1
64.1
13.9

5.9

Employed (%)

22.5

16.9

24.9

Socio-economic level (%)*
  Low
  Middle
  High

26.8
30.7
42.5

36.4
35.2
28.4

Two women had missing information on partner’s migratory status

* Tercile in the distribution of the SEL index.

by Malgorzata Czyzewska on September 9, 2010

isp.sagepub.com

Downloaded from

312

INTERNATIONAL JOURNAL

SOCIAL PSYCHIATRY 55(4)

Table 2

Responses to autonomy items

Total

Women with

migrant partner

(n = 142)

Women with non-

migrant partner

(n = 274)

Decision-making power
Who decides…

n (%)

how much to spend on food?
  You
  Your partner
  Both

164

218

(40)
(7)
(53)

(65)
(3)
(32)

70
25

164

(27)
(10)
(63)

how much to spend on medical attention or medicines?
  You
  Your partner
  Both

106

248

(26)
(14)
(60)

(61)
(4)
(35)

30
48

180

(12)
(19)
(70)

whether to improve the house?
  You
  Your partner
  Both

113
249

(10)
(28)
(62)

19
26
53

(19)
(27)
(54)

14
67

172

(6)
(26)
(68)

whether to buy new land?
  You
  Your partner
  Both

275

(2)
(26)
(71)

28
66

(2)
(29)
(69)

181

(1)
(23)
(76)

whether to buy a vehicle for transport or work?
  You
  Your partner
  Both

102
273

(1)
(27)
(72)

35
59

(1)
(37)
(62)

184

(1)
(22)
(78)

what to plant in a ﬁeld?
  You
  Your partner
  Both

151
208

(3)
(41)
(56)

42
43

(8)
(46)
(47)

(1)
(41)
(55)

when to buy clothes for yourself?
  You
  Your partner
  Both

200

174

(48)
(10)
(42)

(71)
(6)
(24)

101

125

(39)
(13)
(48)

Movement autonomy
To attend a party
  You don’t do it
  You have to ask for permission
  You don’t have to ask for permission

263

(13)
(67)
(20)

18
53
27

(18)
(54)
(27)

183

(9)
(75)
(16)

To visit family or friends
  You don’t do it
  You have to ask for permission
  You don’t have to ask for permission

256
119

(6)
(65)
(30)

57
34

(9)
(57)
(34)

177

(4)
(69)
(27)

To have a job or look for one
  You don’t do it
  You have to ask for permission
  You don’t have to ask for permission

263

(16)
(69)
(16)

14
60
19

(15)
(64)
(20)

172

(16)
(70)
(14)

(Continued)

by Malgorzata Czyzewska on September 9, 2010

isp.sagepub.com

Downloaded from

BOJORQUEZ ET AL.: DEPRESSIVE SYMPTOMS AMONG WOMEN FROM A MEXICAN RURAL AREA

313

Women whose partner was in the USA were high scorers in movement autonomy, with a median
of –0.58 (inter-quartile range (IQR) -0.69–1.27); those whose partner was in the locality had a
median score of -0.69 (IQR -0.69–0.35); and those whose partner was in other place within Mexico
had a median of 0.45 (IQR -0.69–0.88); while those whose partner had migrated but was currently
in the locality had a median of -0.69 (IQR -0.69–0.44) (p = 0.005, Kruskall-Wallis test).

Education was not associated with decision-making power. However, a higher educational

level was associated with more movement autonomy (p < 0.001) and economic autonomy
(p < 0.001). On the contrary, a higher educational level was associated with less collective
participation (p = 0.011, Kruskall-Wallis test).

Socio-economic level

The variables that explained more variance, aggregated in the ﬁrst component for the SEL index,
were the house roof and ﬂoor materials, and having appliances such as a refrigerator, a stove or
a washing machine. Households containing a woman who had a migrant partner had a better
SEL (p = 0.0001, Mann-Whitney Test). Women in households with a higher SEL also showed
higher scores in decision-making power (Spearman’s r = 0.09, p = 0.032), economic autonomy
(Spearman’s r = 0.16, p = 0.001), and movement autonomy (Spearman’s r = 0.26, p < 0.001).

Variables associated with depressive symptoms

More DS were observed among women with a migrant partner (Table 3). Out of the 13 women
with a recently migrated partner, eight (62%) had CES-D scores above the cut-off point (not shown
in table). As for the autonomy indexes, CES-D score was associated only with decision-making
power (

ρ = –0.12, p = 0.039). Receiving emotional social support was associated with lower

CES-D scores. There was a tendency for DS to be less at higher education levels. No association
was observed with age or number of children.

The ﬁnal logistic regression model included the following independent variables: partner’s

migration, decision-making power, economic autonomy, emotional social support, SEL, and
education. The interaction of each autonomy index and partner’s migration did not reach statis-
tical signiﬁcance and was not included in the ﬁnal model. The same happened for the interaction

Total

Women with

migrant partner

(n = 142)

Women with non-

migrant partner

(n = 274)

Political autonomy
Takes part in…
  school parent’s meetings
  church meetings
  community meetings
  meetings of political associations

306
132

77
30

(74)
(32)
(19)
(7)

82
30
26
11

(80)
(29)
(25)
(11)

185

88
43
16

(72)
(34)
(17)
(6)

Economic autonomy
  Has received credit
  Has a paid employment
  The family’s properties are registered in her name

(10)
(23)
(15)

13
18
19

(14)
(17)
(20)

15
64
30

(6)
(25)
(12)

Table 2 (Continued)

by Malgorzata Czyzewska on September 9, 2010

isp.sagepub.com

Downloaded from

314

INTERNATIONAL JOURNAL

SOCIAL PSYCHIATRY 55(4)

between social support and partner’s migration. The interaction observed between SEL and partner’s
migration was included (Table 4).

According to the model, having a partner who had migrated to the USA was associated with

higher odds for DS, even among those whose partner was visiting in the locality and after controlling
for other social factors. Decision-making power was not associated with DS, after controlling for
the other variables in the model; while economic autonomy was directly associated with those
symptoms. Emotional social support was inversely associated with DS. A higher education was
associated with less DS. A negative-sign interaction was observed between SEL and partner’s
international migration, showing that among women whose partner was a migrant a better SEL
was associated with less DS.

DISCUSSION

According to the results, almost one quarter of surveyed women had experienced signiﬁcant DS
during the previous week (a score of 16 or more in CES-D). This proportion is similar to those
found in rural women in other third world countries (Bolton et al., 2004; Patel & Kleinman, 2003),
and in women of low SEL in the USA (Kahn et al., 2000), using the same instrument and cut-off
point. In this way, the prevalence observed is concordant with a population in conditions of social
and economical disadvantage.

Table 3

Social factors and their relationship with depressive symptoms

Variable

CES-D score

(mean, SD)

Variable

CES-D score

(Spearman’s r)

Partner’s migratory condition
  Living in the community
  Somewhere else in Mexico
  Migrant, visiting community
  Living in the USA

9.9 (8.3)

10.7 (7.9)
12.5 (8.5)
13.2 (10.1)

= 0.030

Score in autonomy indexes
Decision-making power
Economic autonomy
Movement autonomy
Political autonomy

= –0.12

= 0.07

= 0.04

= –0.02

= 0.039

= 0.248

= 0.469

= 0.683

Emotional social support
Receives support
Does not receive support

9.7 (8.2)

12.2 (9.3)

= 0.013

Score in socio-economic
level index

= –0.07 p = 0.203

Material social support
Receives support
Does not receive support

11.5 (9.1)
10.8 (8.8)

= 0.613

Age

= 0.03 p = 0.649

Educational level
  No formal education
  Primary school
  Junior high school
  High school and over

14.0 (8.0)
10.3 (8.1)
11.2 (11.5)
10.7 (9.8)

= 0.001

Number of children

= 0.08 p = 0.183

ANOVA

test

by Malgorzata Czyzewska on September 9, 2010

isp.sagepub.com

Downloaded from

BOJORQUEZ ET AL.: DEPRESSIVE SYMPTOMS AMONG WOMEN FROM A MEXICAN RURAL AREA

315

In this sample, the expected association between partner’s migration and DS was conﬁrmed. In

comparison to the women whose partner lived in the locality, those women whose partner was in the
USA or visiting in the locality had more DS. Among the former, DS could be partly a result of the
stressing situation of the partner’s migration. Among the latter, it is possible that family relationships
and arrangements that had evolved to cope with the man’s departure were challenged and reversed
when he came back. Also, women in this situation would know that their partners were bound to
leave again relatively soon. This process could also result in a stressful situation. It is worthwhile
to notice that many of the women whose partner was a recent migrant had CES-D scores above the
cut-off point, showing that the effect of migration over DS might be stronger in the ﬁrst stages of
the migratory process. In this study the effect of migration on DS in those who stay was demon-
strated, as it has been shown in previous research (Salgado & Maldonado, 1992). Nevertheless, the
results also signal the need for further analyzing of the speciﬁc conditions of the migratory process
that are more strongly associated to DS and other common mental-health problems.

As for the relationship between partner’s migration and female autonomy, the hypothesis that

men’s migration would increase the autonomy of the women who stayed behind was not clearly
conﬁrmed. In the sample, women whose partner was in the USA had more movement autonomy,
but their scores in the other three dimensions were similar to those of women whose partner lived
in the locality. It is possible that male migration in this area has not brought forth an increase in
decision-making power, economic autonomy or political participation for women. This subject
has been discussed by Ariza (2000). Contrary to the observations of previous authors such as

Table 4

Logistic regression model for depressive symptoms (n = 297)*

95% IC

p value

Partner
  Living in the community
  Somewhere else in Mexico
  Migrant, visiting
  Living in the USA

1
0.86
8.33
3.77

0.06,
2.61,
1.92,

11.75
26.58
7.43

0.907
0.001
0.001

Decision-making power

0.90

0.67, 1.21

0.482

Economic autonomy

1.45

1.04, 2.02

0.029

Receiving emotional social support

0.36

0.21, 0.64

0.002

Socioeconomic level (score in index)

1.33

0.87, 2.04

0.177

Educational level
  No formal education
  Primary school
  Junior high school
  High school and over

1
0.33
0.26
0.45

0.12,
0.07,
0.16,

0.92
1.01
1.23

0.035
0.052
0.112

Interaction
  Somewhere else in Mexico/SEL**
  Migrant, visiting/SEL
  Living in the USA/SEL

2.45
0.25
0.48

0.15,
0.10,
0.23,

39.44
0.62
1.03

0.504
0.005
0.059

> F = 0.079

* Dependent variable: CES-D score

≥ 16

** SEL: socio-economic level

by Malgorzata Czyzewska on September 9, 2010

isp.sagepub.com

Downloaded from

316

INTERNATIONAL JOURNAL

SOCIAL PSYCHIATRY 55(4)

D’Aubeterre (1995) and Chant (1992), who took migration to be an empowering force in the lives
of women, this author suggests that male migration increases the burden of responsibilities carried
by women, but only in a temporary manner. When men come back the responsibilities return to
them, without a permanent change in gender equity. If this were the case in our study, the observed
increase in movement autonomy for women with migrant partners would not necessarily mean
more equitable gender relations, a subject worthy of further study. Furthermore, the experience
of increased movement autonomy during the partner’s absence, followed by a decrease at the
husband’s return, might also help to explain the high prevalence of DS among women whose partner
was a migrant temporarily visiting the familiy.

Only one of the four dimensions of autonomy considered in this study (decision-making power)

was associated with DS in the crude analysis. The negative association observed was in accord with
the hypothesis of autonomy as a protective factor. In the multivariate analysis, the association lost
statistical signiﬁcance but maintained its direction. Economic autonomy, contrary to expectation,
was positively associated with DS in the multivariate analysis. Neither movement nor political
autonomy showed an association with DS. In this way, the expected clear-cut relationship between
higher autonomy and less DS was not apparent in the sample. A number of reasons for this, and
the need for further studies to clarify the question, are suggested below.

The inverse association of decision-making power with DS is in accord with the ﬁndings of

Chen et al. in the USA (Chen et al., 2005). Those authors studied the contextual effect of autonomy
over DS at the state level and at the individual level. They found that women who lived in states
with more female autonomy had lower average scores in CES-D. Also, in a survey of women in
India, less decision-making power was associated with a higher prevalence of common mental
disorders (Patel et al., 2006). In this sense, our results were as expected.

However, in our study more economic autonomy was associated with a higher CES-D score.

Contrary to this, the classic work by Brown & Harris (1978) stated that not being employed was
a  risk  factor  for  DS.  However,  as  pointed  out  by  Govindasamy  &  Malhotra  (1996),  for  some
women in the developing world, having a paid job is not a matter of choice but of necessity, and
might not be experienced as a liberating condition. Women in our sample might feel that having
to work outside of the home is the result of a lack of support, an observation already made in the
context of Mexico–USA migration (Salgado & Maldonado, 1993). The double burden of work and
household  tasks  could  also  explain  the  direction  of  this  association  (Kuehner,  2003;  Piccinelli
& Wilkinson, 2000).

The ﬁndings of this study are an invitation to discuss the autonomy construct and its oper-

ationalization among women in a context of social disadvantage. The deﬁnition of autonomy included
in the theoretical frame (Friedman, 2003) is part of a conception of individual self-determination
and consequent action as a fundamental good. Autonomy in this view is considered essential for
well-being, in as much as it allows the individual to choose and act in a way that is in accord
with her best interests (Nussbaum, 2002). However, it has been discussed before that, in certain
settings, individual decision-making power and freedom of movement might not be the main assets
that allow a person to follow her chosen course of action or to achieve her more cherished goals.
In a more traditional social milieu, the strength of social networks might be more determinant
of an individual’s action capacity (Govindasamy & Malhotra, 1996). So, if we consider that the
positive effect of autonomy over mental health is the result of an autonomous person being able to
achieve her ends, maybe social ties would constitute a more relevant dimension to consider in the
context of these rural women.

by Malgorzata Czyzewska on September 9, 2010

isp.sagepub.com

Downloaded from

BOJORQUEZ ET AL.: DEPRESSIVE SYMPTOMS AMONG WOMEN FROM A MEXICAN RURAL AREA

317

Also, the meaning of the situations taken as indicators of autonomy ought to be different in

different cultural contexts (Casique, 2001). According to Brown (1998) and Cervantes & Castro
(1985), the effect of provoking factors over DS depends on the meaning people assign to them. The
effect of a certain condition, taken as an indicator of autonomy by the researcher, might depend on
the woman’s perception of it as a liberating circumstance or, on the contrary, as an inappropriate
situation. In this sense, if having a job is experienced as a burden, or as inappropriate according to
gender expectations for women, this indicator of economic autonomy would also be an indicator
of a less protected situation that might be putting the woman at greater stress.

Following this, in further studies it would be convenient to explore other aspects such as inter-

personal relationships and the experience of collectivity, and their association with DS. Deﬁnition
and operationalization of constructs are not minor issues, and the ﬁndings of this study show
the need of further discussion on both aspects in the relationship between autonomy and DS. A
recommendation in this sense is that different aspects of autonomy could be important for each
speciﬁc problem; so that, for empirical research, speciﬁc aspects should be addressed (Govindasamy
& Malhotra, 1996). Also, the meaning of each of those aspects in a speciﬁc social and cultural
context should be investigated.

It has been proposed that social support buffers the effect of stressing life events over DS (Olstad

et al

., 2001; Paykel, 1994; Stansfeld et al., 1998; Thoits, 1995; Wade & Kendler, 2000). We did

not observe such an interaction in our study, maybe because the support we evaluated was not
speciﬁcally directed to the difﬁculties that accompany migration. In our assessment of perceived
social support the independent association we observed between support and DS could be a reverse
causality phenomena, where depressed women might report receiving less emotional support.

After controlling for SEL, an association was observed between education and DS. As other

studies conducted in developing countries have found (Araya et al., 2003; Husain et al., 2004), SEL
did not show an independent association with DS. It is possible that, in our sample, SEL did not
vary enough for associations to be observed. On the other hand, the observed interaction of SEL
and partner’s migration showed that a better SEL mitigates the effect of migration on DS.

An interesting ﬁnding of this study is the difference between women with complete and incom-

plete sets of responses to the CES-D items. On average, women with incomplete responses were
less educated, had a lower SEL and less movement autonomy, and had more children. Showing
these characteristics, and according to our theoretical framework, these women, as a group,
should be more affected by DS. Leaving them outside of the analysis might underestimate the pre-
valence of DS in our sample.

A weakness of this study is the cross-sectional design, which precludes the estimation of

causal relationships. However, a reverse causality phenomenon – with women’s DS inﬂuencing
the partner’s migration – is not likely. An additional difﬁculty arises regarding the temporality of
CES-D. This instrument asks about DS during the previous week, preventing the evaluation of the
temporal relation of symptoms and partner’s departure. Thus, we cannot know if DS appeared right
after the partner’s departure, if they were already present, or if they increased or diminished after
the study. A possible incidence-prevalence bias (Szklo & Nieto, 2003) could also have occurred,
in which CES-D high scorers would be those women with a tendency to chronicity of DS, related
to unmeasured factors; while women who presented acute DS after a partner’s departure were not
detected in our study. Even with these considerations, partner’s migration could still represent a
risk factor with an effect on susceptible persons. The observation of more clinically signiﬁcant
DS among women with a recently migrated partner supports the proposed association.

by Malgorzata Czyzewska on September 9, 2010

isp.sagepub.com

Downloaded from

318

INTERNATIONAL JOURNAL

SOCIAL PSYCHIATRY 55(4)

A number of issues raised in this study merit further investigation through different methods.

The use of qualitative research techniques, such as in-depth interviews, focus groups or life stories,
would help clarify how speciﬁc aspects of the migratory process, such as family characteristics,
the woman’s social position, and others, mediate the impact of the husband’s migration. Also,
the meaning and subjective experience associated with taking care of a number of activities and
decisions should be explored, in order to understand the apparently paradoxical results on auto-
nomy indicators and DS.

According to the observed results, international migration of the partner is associated with more

DS among women who stay in Mexico. This relationship should be considered when evaluating
the general effects of international migration, and in the development of policies. The effect of
migration on the health of those who stay in the original communities is a ﬁeld in need of further
investigation. Representative samples of wider populations and longitudinal designs will allow
us to understand the effects of migration on the mental health of those who stay behind.

ACKNOWLEDGMENTS

Proyecto Mixteca

was supported by the Wellcome Trust, grant number 042789MF. The ﬁrst author

was supported by the National Council for Science and Technology (CONACYT) during this
research project. The authors want to thank the people of Tlapa, Guerrero, Huajuapan, Oaxaca
and Acatlan, Puebla, for their participation in this study.

NOTE

A previous version of this paper was presented at the Latin American Studies Association (LASA) 2007 International
Congress.

REFERENCES

Aguilera-Guzman, R.M., Salgado de Snyder, V.N., Romero, M. & Medina-Mora, M.E. (2004) Paternal absence and

international migration: stressors and compensators associated with the mental health of Mexican teenagers of
rural origin. Adolescence, 39, 711–723.

Araya, R., Lewis, G., Rojas, G. & Fritsch, R. (2003) Education and income: which is more important for mental

health? Journal of Epidemiology and Community Health, 57, 501–505.

Ariza, M. (2000) Género y migración femenina: dimensiones analíticas y desafíos metodológicos. In Migración y

Relaciones de Género en México

(eds D. Barrera & C. Oehmichen). Mexico: GIMTRAP/IIA-UNAM.

Bhugra, D. (2004) Migration and mental health. Acta Psychiatrica Scandinavica, 109, 243–258.
Bloom, S., Wypij, D. & Das Gupta, M. (2001) Dimensions of women’s autonomy and the inﬂuence on maternal

health care utilization in a North Indian city. Demography, 38, 67–78.

Bolton, P., Wilk, C.M. & Ndogoni, L. (2004) Assessment of depression prevalence in rural Uganda using symptom

and function criteria. Social Psychiatry and Psychiatric Epidemiology, 39, 442–447.

Bowling, A. (1995) Measuring Disease. A Review of Disease-Speciﬁc Quality of Life Measurement Scales. Buckingham:

Open University Press.

Brown, G.W. (1998) Genetic and population perspectives on life events and depression. Social Psychiatry and

Psychiatric Epidemiology

, 33, 363–372.

Brown, G.W. & Harris, T. (1978) Social Origins of Depression: A Study of Psychiatric Disorder in Women.

New York: The Free Press.

by Malgorzata Czyzewska on September 9, 2010

isp.sagepub.com

Downloaded from

BOJORQUEZ ET AL.: DEPRESSIVE SYMPTOMS AMONG WOMEN FROM A MEXICAN RURAL AREA

319

Casique, I. (2001) What Difference Does It Make? Women’s Autonomy and Power and Use of Contraception in Mexico.

XXIV IUSSP General Population Conference, Salvador, Bahia, August 2001.

Casique, I. (2004) Indices de empoderamiento femenino y su relación con la violencia de género. In Violencia de

Género en las Parejas. Resultados de la Encuesta Nacional sobre la Dinámica de las Relaciones en los Hogares
2003

(ed Instituto Nacional de las Mujeres). Mexico: Inmujeres.

Cervantes, R. & Castro, F. (1985) Stress, coping and Mexican American mental health: A systematic review.

Hispanic Journal of Behavioral Sciences

, 7, 1–73.

Chant, S. (1992) Towards a framework for the analysis of gender-selective migration. In Gender and Migration in

Developing Countries

(ed S. Chant). London: Belhaven Press.

Chen, Y.Y., Subramanian, S.V., Acevedo-García, D. & Kawachi, I. (2005) Women’s status and depressive

symptoms: A multilevel analysis. Social Science and Medicine, 60, 49–60.

Consejo Nacional de Población (CONAPO) (2002) Índices de Intensidad Migratoria México–Estados Unidos, 2000.

Mexico: CONAPO.

Consejo Nacional de Población (CONAPO) http://www.conapo.gob.mx/mig_int/03.htm.
D’Aubeterre, M.E. (1995) Tiempos de espera: emigración masculina, ciclo doméstico y situación de las mujeres en

San Miguel Acuexcomac, Puebla. In Relaciones de Género y Transformaciones Agrarias (eds S. Gonzalez &
V. Salles). México: El Colegio de México.

Friedman, M. (2003) Autonomy, Gender, Politics. Oxford: Oxford University Press.
González de la Rocha, M. (1989) El Poder de la Ausencia: Mujeres y Migración en una Comunidad de los Altos

de Jalisco

. XI Coloquio Las Realidades Regionales de la Crisis Nacional, 25–27 October 1989. Zamora,

Michoacán: El Colegio de Michoacán.

Govindasamy, P. & Malhotra, A. (1996) Women’s position and family planning in Egypt. Studies in Family Planning,

27, 328–340.

de Graaf, R., Bijl, R.V., Ravelli, A., Smit, F. & Vollebergh, A.M. (2002) Predictors of ﬁrst incidence of DSM-III-

R psychiatric disorders in the general population: ﬁndings from the Netherlands Mental Health Survey and
Incidence Study. Acta Psychiatrica Scandinavica, 106, 303–313.

Hondagneu-Sotelo, P. (1994) Gendered Transitions. Mexican Experiences of Immigration. Berkeley: University of

California Press.

Hosmer, D.W. & Lemeshow, S. (1989) Applied Logistic Regression. New York: Wiley.
Husain, N., Gater, R., Tomenson, B. & Creed, F. (2004) Social factors associated with chronic depression among

a population-based sample of women in rural Pakistan. Social Psychiatry and Psychiatric Epidemiology, 39,
618–624.

Instituto Nacional de Estadística, Geografía e Informática (INEGI). Sistema Municipal de Base de Datos (SIMBAD).

Retrieved from http://www.inegi.gob.mx/prod_serv/contenidos/espanol/simbad/default.asp?c=73.

Jejeebhoy, S.J. (1995) Women’s Education, Autonomy, and Reproductive Behaviour: Experience from Developing

Countries

. Oxford: Clarendon Press.

Kahn, R.S., Wise, P.H., Kennedy, B.P. & Kawachi, I. (2000) State income inequality, household income, and maternal

mental and physical health: Cross-sectional national survey. British Medical Journal, 321, 1311–1315.

Kendler, K.S., Gardner, C.O. & Prescott, C.A. (2002) Toward a comprehensive developmental model for major

depression in women. American Journal of Psychiatry, 159, 1133–1145.

Kolenikov, S. & Angeles, G. (2004). The Use of Discrete Data in PCA: Theory, Simulations, and Applications to

Socio-Economic Indices. Work Document WP-04–85

. North Carolina: Chapel Hill University.

Kuehner, C. (2003) Gender differences in unipolar depression: An update of epidemiological ﬁndings and possible

explanations. Acta Psychiatrica Scandinavica, 108, 163–174.

McGuire, S. & Martin, K. (2007) Fractured migrant families: Paradoxes of hope and devastation. Family & Com-

munity Health

, 30, 178–188.

Medina-Mora, M.E., Borges, G., Lara, C., Benjet, C., Blanco, J., Fleiz, C., Villatoro, J., Rojas, E. & Zambrano, J.

(2005) Prevalence, service use, and demographic correlates of 12-month DSM-IV psychiatric disorders in
Mexico: Results from the Mexican National Co-morbidity Survey. Psychological Medicine, 35, 1–11.

Nussbaum, M. (2002) Las Mujeres y el Desarrollo Humano. Barcelona: Herder.
Olstad, R., Sexton, H. & Søgaard, A.J. (2001) The Finnmark Study. A prospective population study of the social

support buffer hypothesis, speciﬁc stressors and mental distress. Social Psychiatry and Psychiatric Epidemiology,
36, 582–589.

by Malgorzata Czyzewska on September 9, 2010

isp.sagepub.com

Downloaded from

320

INTERNATIONAL JOURNAL

SOCIAL PSYCHIATRY 55(4)

Pallitto, C.C. & O’Campo, P. (2005) Community level effects of gender inequality on intimate partner violence

and unintended pregnancy in Colombia: Testing the feminist perspective. Social Science and Medicine, 60,
2205–2216.

Patel, V. & Kleinman, A. (2003) Poverty and common mental disorders in developing countries. Bulletin of the

World Health Organization

, 81, 609–615.

Patel, V., Kirkwood, B.R., Pednekar, S., Pereira, B., Barros, P., Fernades, J., Datta, J., Pai, R., Weiss, H. & Mabey,

D. (2006) Gender disadvantage and reproductive health risk factors for common mental disorders in women.
A community survey in India. Archives of General Psychiatry, 63, 404–413.

Paykel, E.S. (1994) Life events, social support and depression. Acta Psyhciatrica Scandinavica, 377 (Suppl),

50–58.

Piccinelli, M. & Wilkinson, G. (2000) Gender differences in depression. Critical review. British Journal of

Psychiatry

, 177, 486–492.

Pottinger, A.M. (2005) Children’s experience of loss by parental migration in inner-city Jamaica. American Journal

of Orthopsychiatry

, 75, 485–496.

Radloff, L. (1977) The CES-D Scale: A self-report depression scale for research in the general population. Applied

Psychological Measure

, 1, 385–401.

Salgado de Snyder, N. (1996) Problemas psicosociales de la migración internacional. Salud Mental, 19 (Suppl),

53–59.

Salgado de Snyder, V.N., Bojorquez, I., Gonzalez, T. & Infante, C. (2007) Migración México-Estados Unidos:

Consecuencias Para la Salud.

Cuernavaca: Instituto Nacional de Salud Pública.

Salgado de Snyder, V.N. & Maldonado, M. (1992) Respuestas de enfrentamiento e indicadores de salud mental en

esposas de emigrantes a los Estados Unidos. Salud Mental, 15, 28–35.

Salgado de Snyder, V.N. & Maldonado, M. (1993) Funcionamiento psicosocial en esposas de emigrantes mexicanos

a los Estados Unidos. Revista Latinoamericana de Psicología, 25, 167–180.

Salgado de Snyder, V.N. & Maldonado, M. (1994) Características psicométricas de la Escala de Depresión del

Centro de Estudios Epidemiológicos en mujeres mexicanas adultas de áreas rurales. Salud Pública de México,
36, 200–209.

Sheehan, E.P. (1996) The social consequences of emigration from Ireland for those who stay. Irish Journal of

Psychology

, 17, 13–22.

Silverstein, B. & Perlick, D. (1995) The Cost of Competence: Why Inequality Causes Depression, Eating Disorders

and Illness in Women

. New York: Oxford University Press.

Skapinakis, P., Weich, S., Lewis, G., Singleton, N. & Araya, R. (2006) Socio-economic position and common mental

disorders. Longitudinal study in the general population in the UK. British Journal of Psychiatry, 189, 109–117.

SPSS 13.0. CATPCA. Created by DTSS, Leiden University.
Stansfeld, S.A., Fuhrer, R. & Shipley, M.J. (1998) Types of social support as predictors of psychiatric morbidity in a

cohort of British civil servants (Whitehall II Study). Psychological Medicine, 28, 881–892.

Stata Corporation (2003) Stata/SE 8.0 for Windows. Texas: College Station.
Szklo, M. & Nieto, J. (2003) Epidemiología Intermedia. Conceptos y Aplicaciones. Madrid: Diaz de Santos.
Thoits, P.A. (1995) Stress, coping and social support processes: Where are we? What next? Journal of Health and

Social Behavior

, 35, 53–79.

Wade, T.D. & Kendler, K.S. (2000) Absence of interactions between social support and stressful life events in

the prediction of major depression and depressive symptomatology in women. Psychological Medicine, 30,
965–974.

Wagner, F., Diaz, D.B., Lopez, A.L., Collado, M.E. & Aldaz, E. (2002) Social cohesion, cultural identity, and drug

use in Mexican rural communities. Substance Use & Misuse, 37, 715–747.

Weich, S., Lewis, G. & Jenkins, S.P. (2001) Income inequality and the prevalence of common mental disorders in

Britain. British Journal of Psychiatry, 178, 222–227.

Weissman, M.M., Sholonskas, D., Pottenger, M., Prusoff, B.A. & Locke, B.Z. (1977) Assessing depressive symptoms

in ﬁve psychiatric populations: A validation study. American Journal of Epidemiology, 106, 203–214.

World Health Organization (2000) Women’s Mental Health. An Evidence-Based Review. Geneva: World Health

Organization.

Zárate, M. (2000) Participación política, migración y mujer en Michoacán. In Migración y relaciones de género en

México

(eds D. Barrera & C. Oehmichen). Mexico: GIMTRAP/ IIA-UNAM.

by Malgorzata Czyzewska on September 9, 2010

isp.sagepub.com

Downloaded from