close

Вход

Забыли?

вход по аккаунту

?

Seasonal changes in household food insecurity and symptoms of anxiety and depression.

код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 135:225–232 (2008)
Seasonal Changes in Household Food Insecurity
and Symptoms of Anxiety and Depression
Craig Hadley1* and Crystal L. Patil2
1
2
Department of Anthropology, Emory University, Atlanta, GA 30322
Department of Anthropology, University of Illinois at Chicago, Chicago, IL 60607-7139
KEY WORDS
mental health; well-being; security; burden; Tanzania; East Africa
ABSTRACT
There is growing awareness that common mental health disorders are key contributors to the
burden of disease in developing countries. Studies examining the correlates of mental health have primarily
been carried out in urban settings and focused on the
burden rapid economic change places on individuals. In
these settings, poverty and low education are consistent
predictors of anxiety and depressive symptoms. We
argue here that these variables are proxies for insecurity, and that a more general model of symptoms of
depression and anxiety should focus on locally salient
forms of insecurity. Building on previous work in a seasonal subsistence setting, we identify food insecurity as
a potent source of insecurity in a rural African setting,
and then test whether seasonal changes in food insecurity are correlated with concomitant changes in a mea-
sure of symptoms of anxiety and depression among 173
caretakers. Results indicate that food insecurity is a
strong predictor of symptoms of anxiety and depression
(P < 0.0001), that changes in food insecurity across the
seasons predict changes in symptoms of anxiety and
depression (P < 0.0001), and that this is robust to the
inclusion of covariates for material assets and household
production. These results hold for individuals in both
ethnic groups studied (Pimbwe and Sukuma); however,
at the group level the burden falls disproportionately on
Pimbwe. The results add to the growing literature on
the causes of population level differences in mental
health disorders and suggest new research avenues and
strategies to link mental health disorders with variation
in physical and biosocial outcomes. Am J Phys Anthropol
135:225–232, 2008. V 2007 Wiley-Liss, Inc.
There is increasing recognition that common mental
health disorders are a major contributor to the global
burden of disease and that the impact of this burden is
not confined to developed countries. Worldwide estimates
of the burden of common mental disorders place common
mental disorders alongside many physical health disorders (Murray and Lopez, 1997; WHO, 2001). Further,
evidence suggests that women are disproportionably
affected by anxiety and mood disorders (Murray and
Lopez, 1997). The recognition of these burdens signals a
shift in international public health and anthropological
thinking away from a focus on infectious diseases and
diseases of undernutrition toward a broader aim at identifying the causes of poor health, mental health included
(Patel et al., 2006). Recent reviews of the relationship
between poverty and common mental health disorders in
developing country settings have highlighted the association between broad indicators of socioeconomic status,
gender, and educational attainment using cross-sectional
studies. More nuanced studies suggest that the underlying feature that links these indicators with common
mental health outcomes is insecurity, which is consistent
with the assumption that unpredictability and uncertainty are predictors of the onset and reoccurrence of
anxiety and depressive disorders (Patel et al., 1999;
WHO, 2001). In many seasonal and subsistence economies throughout the world, insecure access to food is a,
if not the, dominant form of insecurity (Ferro-Luzzi,
1990; Tomkins, 1993; Ulijaszek and Strickland, 1993).
The objectives of this article are, first, to lay out a general ecological argument that the contributors to poor
mental health outcomes are likely to be localized and
context-specific, and therefore knowledge of the local
ethnography of an area is critical to capture key domains
of burden, and, second, to assess the relationship
between food insecurity and a measure of anxiety and
depression among female caretakers in a subsistence
economy setting in rural Tanzania.
The biosocial model generally invoked to examine the
causes and consequences of mental heath outcomes
views mental health as the product of biological, psychological, and social factors (WHO, 2001). Mood fluctuations, or for instance depressive symptoms, are envisioned to vary across a population with some individuals
experiencing a greater severity of symptoms than others.
Shocks to the system or the perception of imminent
shocks to the system, such as violence or job loss, are
predicted to increase the number of symptoms experienced, and in turn, move individuals across diagnostic
thresholds where they subsequently become classified as
experiencing a mental health disorder. In line with this
model and the predicted causative agents, a review of
studies carried out in developing countries (mostly
transitional economies) by Patel and Kleinman (2003)
C 2007
V
WILEY-LISS, INC.
C
Grant sponsor: National Science Foundation; Grant number:
0412210. Grant sponsor: Department of Anthropology, Washington
University in St. Louis.
*Correspondence to: Craig Hadley, Department of Anthropology,
Emory University, 207 Anthropology Building, 1557 Dickey Drive,
Atlanta, GA 30322, USA. E-mail: chadley@emory.edu
Received 15 March 2007; accepted 27 August 2007
DOI 10.1002/ajpa.20724
Published online 28 November 2007 in Wiley InterScience
(www.interscience.wiley.com).
226
C. HADLEY AND C.L. PATIL
identified 10 studies that provided clear evidence of an
association between measures of poverty and common
mental health disorders. In particular, they identified
low income, poor education, and female gender as significant predictors of common mental health disorders. In a
subsequent review, the predictive power of low education
was particularly highlighted, ‘‘especially in developing
countries that are facing a growing lack of security for
employees as economies are reformed’’ (Patel et al.,
2006). From this framework it has also been hypothesized that one’s ability to cope with insecurity is buffered
by material and social resources (Cohen et al., 2000;
Patel et al., 2006).
Although a majority of the research conducted on mental health in developing countries identifies low socioeconomic status as a strong correlate of mood disorders it
should be noted that these studies have also largely
taken place in urban or peri-urban settings within the
larger context of rapid economic change. The focus on
rapidly urbanizing context is of course warranted given
the rapid transition occurring within many countries in
sub-Saharan Africa. Tanzania for instance has seen the
size of the urban population climb from 6% in 1967 to
23% in 2002 (Macro, 2005); nevertheless, nearly 80% of
the population lives in rural areas. Health theories from
medical anthropology link health outcomes and risk factors with the specific ecology within which an individual
resides (McElroy and Townsend, 2004; Trostle, 2005;
Scrimshaw, 2006). The risk factors identified in studies
of urban settings therefore may be specific to those particular socioecological settings; risk factors are likely to
be highly context specific. Risk factors like education, for
example, may be salient in economies where opportunities are often structured along educational lines, but
may be of limited importance as a causative agent in
areas where this is not the case. In these areas education is unlikely to have the same meaning (and consequences) as it does in settings with skills-based labor
economies. In the absence of such economies and opportunities, low education is unlikely to be associated with
mental health because low education does not necessarily create differential opportunities or buffer against
insecurity and predictability in the food or income supply as it can in urban settings.
In rural subsistence economies, a more meaningful
source of insecurity than education and job security may
be the availability of food (Pike, 2004; FAO, 2005; Pike
and Patil, 2006; Pike and Williams, 2006). Food insecurity, the limited or uncertain availability of nutritionally
adequate and safe foods, or limited, or uncertain ability
to acquire acceptable foods in socially acceptable ways,
may be particularly important in seasonal subsistence
economies that are dominated by distinct peaks in food
shortfalls often referred to as ‘‘hunger seasons’’ (Richards, 1939). These seasonal periods of food shortage
have been associated with a broad array of adverse physical health outcomes including increased mortality
among children born during these periods, slower
growth, greater incidence of diarrhea and malarial morbidity (Tomkins, 1993), and a range of coping strategies
including migration, trading labor for food, and begging
and borrowing (Dirks, 1980; Shipton, 1990). The conceptual framework guiding the current study links insecurity to the expression of symptoms of common mental
health disorders, and we theorize that insecurity must
be defined locally and that it must be meaningful to the
actors occupying the study area. If food insecurity is a
locally important manifestation of insecurity then the
occurrence of food insecurity is assumed to generate symptoms of anxiety and depression as people struggle to feed
themselves, and their families, and to achieve well-being.
Consistent with this model, our previous ethnographic
and survey work among two ethnic groups living in
rural Tanzania strongly suggests that food insecurity
was a highly salient and locally appropriate source of
insecurity and was commonly mentioned by respondents
and informants as a source of considerable distress,
much more so than lack of education or limited ownership of material assets (Hadley, 2005b; Hadley et al.,
2007). In our previous work, which utilized quantitative
and qualitative methods, we noted what appeared to be
an increase in symptoms of anxiety during the wet season, the period when food insecurity was most problematic. In other studies, we have attempted to measure the
physical health consequences of this seasonal burden,
and the social and economic predictors of food insecurity.
On the basis of this previous survey and ethnographic
data, we hypothesized that the unpredictability and
uncertainty that defines food insecurity would predispose
individuals in food insecure households to a higher prevalence of common mental health disorders (Hadley and
Patil, 2006). Based on this hypothesis, we predicted and
subsequently identified a strong association between an
experientially based measure of food insecurity and a
measure of anxiety and depression in a multi ethnic
group study. In the study populations that we surveyed,
respondents living in food insecure households scored
significantly higher on a measure of anxiety and depression. Like many studies examining the relationship
between measures of poverty and mental health, our
earlier work was cross-sectional. This study design
invites critiques of reverse causality: perhaps those with
many symptoms of anxiety and depression farmed less
and therefore became food insecure. Only a longitudinal
study design can counter this valid criticism (for a US
example see Siefert et al., 2004).
In the current study, we examine the extent to which
a measure of insecurity is associated with a measure of
anxiety and depression among individuals living in a
highly seasonal subsistence setting. We do this in a longitudinal fashion that exploits seasonal changes in food
availability and therefore combats a central weakness of
earlier studies, including our own. Dry (food secure) and
wet (food insecure) season data are used to test for correlated changes in food insecurity and maternal anxiety
and depression. We also move forward by controlling for
a greater number of covariates to assess whether material wealth is driving the food insecurity–anxiety and
depression relationship. Based on the hypothesis that
insecurity leads to symptoms of anxiety and depression,
and on our previous results and ethnographic knowledge
of the area, we made two predictions: (1) changes in food
insecurity across the seasons would be associated with
concomitant changes in a measure of anxiety and
depression, and that (2) the measure of anxiety and
depression would be associated with food insecurity independent of other locally appropriate measures of wealth
and socioeconomic status.
STUDY COMMUNITIES AND METHODS
This study took place in rural Western Tanzania
among agropastoralists (Sukuma) and horticulturalists
American Journal of Physical Anthropology—DOI 10.1002/ajpa
FOOD INSECURITY AND ANXIETY AND DEPRESSION
(Pimbwe) who occupy a series of villages near the border
of a large national park. As is detailed elsewhere
(Hadley, 2005b; Hadley et al., 2007), the study area is
marked by a highly seasonal environment, subsistence
agriculture, and limited health and economic infrastructure. Within this setting Sukuma and Pimbwe live in
close proximity and interact on a daily basis but the two
groups differ across a number of dimensions. Sukuma
herd cattle and farm rice and corn and typically enjoy
greater levels of household production and consumption
than do their Pimbwe neighbors. Pimbwe households are
smaller than Sukuma households and Pimbwe generally
farm smaller plots of land without the aid of cattle. The
seasonal nature of the economy, limited access to purchased foods, and the poverty in the area create a yearly
increase in the proportion of households that are food
insecure.
During the food insecure wet season (December–
March), but very rarely in the dry (June–October), individuals were repeatedly observed to be struggling, as
evidenced by persons seeking loans from others, selling
their labor for food, and lamenting that their children
were ‘‘going to bed hungry.’’ A common response to food
shortages among families was ‘‘kuhemea’’ which means
to sell or trade one’s labor for food; this was particularly
true for Pimbwe families. Unfortunately, the search for
food often has the unintended result of reducing the
amount of time the food insecure family can allocate to
their own farming. Among the most adversely affected
households, individuals woke each morning and began a
daylong search for affordable food or to exchange their
labor for food. In the face of uncertain access to food,
household members were clearly concerned about their
day-to-day food supply and undertook a number of strategies to make their food last longer. These strategies
included consuming meals less frequently throughout
the day, consuming smaller meals, and consuming foods
that were generally viewed as ‘‘hunger’’ foods. Households were also observed eating maize without removing
the outer shell through pounding or machine—in this
way no maize was lost during the milling process, which
led to a greater amount of food. Households were also
observed consuming wild plants and fish, which are
abundant for a short part of the wet season. As many of
the villagers note in formal and informal interviews, the
single growing season means that if the crop yield is
poor or the rains fail to come on time then they are
bound to experience periods of food insecurity. In the dry
period of the year, especially in the immediate postharvest season, however, food is abundant and household
food stores were routinely observed to be filled, meals
were consumed throughout the day, and children were
observed leaving food behind. Although individuals from
both Pimbwe and Sukuma households were observed to
experience or express concern over food insecurity, as a
group Pimbwe appear to suffer disproportionately from
seasonal food insecurity.
To investigate the impact that these seasonal differences in food insecurity had on symptoms of anxiety and
depression, 173 female caretakers were surveyed in the
dry (June–July 2005) and wet seasons (December–January 2006). These women represent a subset of a random
sample of 211 women who provided answers to the questionnaire in the dry season. The loss is due to attrition
and to the fact that one location of the study area was
not revisited because of difficulties securing transportation during the wet season, underscoring the difficulties
227
that arise during the wet season. For the dry and wet
season round, trained interviewers interviewed women
face-to-face using techniques previously employed in the
study area (Hadley, 2005a,b; Hadley and Patil, 2006).
Sociodemographics
Questionnaires were used to collect standard sociodemographic information such as age, marital status,
education, and information on household composition,
production, and assets. Because of the nature of educational achievement in the study area, education was
recoded into no education and more than one year of
education. Survey rounds occurred in June–July and December–January, representing the dry and wet seasons,
respectively.
Food insecurity
Food insecurity was measured using a 15-item scale
based loosely on the USDA’s food insecurity instrument
(Bickel et al., 2000). The instrument is an experientially
based instrument, which means it attempts to measure
the phenomena of interest directly, food insecurity,
rather than proxy indicators, which may or may not
actually tap the core construct. The instrument has high
face and content validity: it is predictive of household
wealth, dietary intake, and engagement in common coping strategies and has high internal consistency. Crosscultural studies show that the core elements of this
instrument tap into the experience of food insecurity and
show that the measure is valid in a range of settings
(Coates et al., 2006a,b; Frongillo and Nanama, 2006;
Melgar-Quinonez et al., 2006; Webb et al., 2006). The
items include questions about the frequency of meals,
reductions in meals, skipping meals because of hunger,
and caretakers were asked whether they have experienced each item in the previous 3 months. This instrument is referred to hereafter as the food insecurity
instrument (FSI) and was asked in both dry and wet
seasons rounds of data collection. Affirmative responses
were coded as 1, nonaffirmative were coded as 0. A caretaker’s food insecurity score was constructed by summing across all items with a maximum possible score of
15, representing the most severe food insecurity and a
minimum of 0, representing the most food secure. Seasonal changes in food security were calculated by subtracting dry season values from wet season values.
Measuring symptoms of anxiety and depression
The Hopkins Symptom Checklist (HSCL) is a well
validated inventory of anxiety and depressive symptoms
that consists of 25 items. The instrument has been used
extensively in vulnerable populations, such as refugees,
as well as war survivors and in a range of cross cultural
contexts (Mollica et al., 2004) including Tanzania (Kaaya
et al., 2002; Pike and Patil, 2006) and the present field
site (Hadley and Patil, 2006). The instrument includes
25 symptoms of depression and anxiety and respondents
are asked to identify whether they have experienced
each of the items as well the severity of the symptom on
a 4-point Likert scale in the 2 weeks before the survey.
These include statements such as ‘‘In the past week how
bothered have you been by your heart pounding or racing?’’ and ‘‘In the past week how bothered were you by
feeling low in energy, slowed down?’’ We excluded one
question on sexual desire because it was not culturally
American Journal of Physical Anthropology—DOI 10.1002/ajpa
228
C. HADLEY AND C.L. PATIL
appropriate to ask respondents. Thus, scores for the
HSCL were calculated by summing the respondent’s
total responses; a respondent’s score could range
between 24 (not endorsing any of the items) to a high of
96 (endorsing all items at their highest level); note that
this is lower than the usual score because of the exclusion of the one item. The HSCL was asked in both the
dry and the wet season and dry season values were subtracted from wet season values to measure change in
symptoms of anxiety and depression. A respondent’s
score on the HSCL can be compared against a cut-off to
make an initial assessment of whether she is positive for
high depression and anxiety. Rather than using the cutoff we used a continuous version of the HSCL for two
reasons. First, the cut-off has not been clinically determined in this setting and therefore may be misleading.
Second, and more importantly, even subclinical levels of
depressive and anxiety symptomology are associated
with impairment, elevated risk of major depression, and
may influence mother-child interactions (Sohr-Preston
and Scaramella, 2006).
Socioeconomic status
Several additional measures were only collected in the
wet season so that we could examine the possible influence of other measures of socioeconomic status on symptoms of anxiety and depression. Respondents were asked
about their household’s production of corn (measured in
bags produced), the number of different crop types that
they had planted, and the number of domestic animals
they owned at the time of interview. The number of animals owned was highly skewed and therefore was categorized into a three level ordinal variable representing
no animal holdings, low animal holdings, and high animal holdings. In addition to production, material wealth
was measured by asking whether the household owned
six different assets (bike, radio, watch, mosquito net,
plow, tin roof). These material goods were then summed
together to create an asset score. Note that the wealth
data, the production data, and data on livestock holdings
were collected only in the wet season round.
Statistical analyses
Paired t-tests were used to assess changes in the FSI
and the HSCL across seasons of measurement. Spearmans correlation was used to examine the association
between the wet season FSI and the wet season HSCL.
Our primary analyses employed ordinary least squares
regression to assess the association between changes in
the FSI and changes in HSCL while controlling for the
following measures: respondent ethnicity, material
wealth, household production, animal ownership, and
maternal and household characteristics. Logistic regression was used to examine the predictors of whether a
household was successfully followed up or not, to assess
whether attrition led to systematic bias in the sample.
The 0.05 value was selected as the criterion for statistical significance.
RESULTS
Data were available for 173 of the 211 caretakers who
originally provided complete information from the dry
season round. To examine whether the women who were
not followed differed systematically from those women
who were, we ran a logistic regression model predicting
TABLE 1. Selected sample characteristics (n 5 173)
Mother’s age (yr)
Any education? % yes
Material assets (items)
Animals
Household size
Corn produced (bags)
Number of crops planted
Household food
insecurity—wet season
HSCL score—wet season
HSCL score—dry season
Average dry season
HSCL [ 1.75, % yes
Average wet season
HSCL [ 1.75, % yes
Sukuma
Pimbwe
P
28
22%
2.2
52
7.8
12.9
3.9
2.5 (2.9)
30
77%
1.7
6
5.5
4.0
2.2
5.1 (4.4)
0.19
\0.0001
\0.0001
0.0017
\0.0001
\0.0001
\0.0001
\0.0001
31.9 (7.3)
36.9 (10.6)
22.2%
37.7 (9.7)
34.0 (9.5)
17.8%
\0.0001
0.06
0.46
10%
31.7%
0.0004
follow-up in the wet season round. Measures of food
insecurity, symptoms of anxiety and depression in the
dry season, corn production, child weight for age at baseline, caregiver’s weight at baseline, and ethnicity were
entered into this model. Only the ethnicity variable
emerged as a significant predictor of whether a household was followed up or not: Sukuma were less likely to
be followed up (P 5 0.013). Removing Pimbwe from the
sample, the same model revealed no significant predictors of wet season follow up, suggesting that Sukuma
who were not followed up did not differ systematically
from those who were successfully followed up.
Characteristics of the final study sample are shown in
Table 1. The respondents in the sample were on average
29 years of age and there was no difference in age
between women in each ethnic group. Sukuma women
were far less likely to have been to school (22%) than
Pimbwe women (77%). Sukuma women also reported
owning more material goods, more animals, living in
larger households, and producing more corn. Sukuma
households also had more diverse crop portfolios as evidenced by the significantly greater number of different
types of crops planted (Table 1).
In the full sample, mean values did not differ for
symptoms of anxiety and depression (HSCL) in the dry
and wet seasons. Mean values for dry and wet season
were 35.3 (10.1) and 34.9 (9.1), respectively (P [ 0.05).
Further analysis showed that this simple dry-wet comparison concealed important heterogeneity. In the dry
season, Sukuma women scored marginally higher than
Pimbwe on the HSCL (P 5 0.06; we have reported this
difference previously although the exact value differs
somewhat because of sample attrition; Hadley and Patil,
2006). In contrast, in the wet season, Pimbwe women
scored significantly higher than did Sukuma women on
the measure of anxiety and depression. The mean value
for the Pimbwe women was 37.7 (9.7) and for the
Sukuma women it was 31.9 (7.3; P \ 0.0001). Paired
tests within the Pimbwe ethnic group showed a mean
increase in HSCL across the seasons (P \ 0.0001)
whereas Sukuma women’s mean HSCL score declined (P
\ 0.0001).
In the dry season survey, the average food insecurity
score among both groups combined was 2.2 (3.1) but
increased to 3.9 (4.0) in the wet season survey; this difference was statistically significant (P \ 0.0001). As previously reported (Hadley and Patil, 2006), there was no
difference in the dry season food insecurity score
American Journal of Physical Anthropology—DOI 10.1002/ajpa
229
FOOD INSECURITY AND ANXIETY AND DEPRESSION
TABLE 2. Regression model predicting changes in the number
of symptoms of anxiety and depression (n 5 173, F 5 9.89,
r2 5 34%, P < 0.0001)
Fig. 1. Interseason changes in food insecurity and maternal
anxiety and depression (n 5 173, r 5 0.38, P < 0.0001). Positive
values on the x-axis indicate increasing food insecurity across the
study period.
between the Sukuma and Pimbwe. Sukuma respondents
had an average food insecurity score of 1.97 (2.9)
whereas Pimbwe respondents of 2.38 (3.2; P 5 0.39). In
contrast, during the wet season survey period significant
differences in mean food insecurity values emerged. In
the wet season, Sukuma households reported a mean
food insecurity value of 2.5 (2.9) and Pimbwe households
reported a mean value of 5.1 (4.4); this ethnic difference
was statistically significant (P \ 0.0001). Paired tests
show that Sukuma households experienced no overall
change in their food insecurity status (P 5 0.21) while
Pimbwe respondents showed a significant increase indicating higher levels of food insecurity (P \ 0.0001).
In the full sample, wet season HSCL scores were associated with wet season FSI scores (r 5 0.44, P \ 0.0001).
This association maintained in each ethnic group
(Pimbwe: r 5 0.35, P 5 0.0004; Sukuma: r 5 0.44, P \
0.0001) suggesting that at the individual-level, increased
food insecurity was associated with increased symptoms.
Next changes in food insecurity across the dry and wet
seasons were regressed on changes in the HSCL across
seasons. This exercise revealed a highly significant association: women who reported experiencing increasingly
severe food insecurity in the wet season (relative to their
dry season food insecurity value) also reported increased
severity of symptoms of anxiety and depression (r 5
0.38, P \ 0.0001; Fig. 1). Separate models for each ethnic group also revealed positive and statistically significant coefficients for change in food insecurity and
changes in symptoms of anxiety and depression (P \
0.01 for food insecurity change coefficient for women of
both ethnic groups). This shows that as food insecurity
increased or decreased so did symptoms of anxiety and
depression and this was true for women in both study
communities.
Covariates representing wet season household asset
ownership and production along with measures of mater-
Source
Beta (SE)
F
Sig.
Intercept
Change in food insecurity score
Ethnicity—Pimbwe
Ethnicity—Sukuma (ref)
Household size
Mother’s age (years)
Any education? (% yes)
Material assets (items)
Corn produced (bags)
Number of crops planted
4.13 (4.58)
0.79 (0.21)
6.79 (2.32)
–
0.5 (0.28)
20.23 (0.13)
22.14 (1.89)
0.61 (0.35)
0.86 (0.59)
21.55 (0.63)
4.58
0.21
2.32
–
0.28
0.13
1.89
0.35
0.59
0.63
0.368
\0.001
0.004
–
0.077
0.074
0.259
0.080
0.143
0.015
nal age and education were entered into a model predicting changes in the anxiety and depression score. The
hypothesized predictor variable, change in food insecurity, was also entered into this model. Results show that
the association between changes in household food insecurity and symptoms of anxiety and depression maintained statistical significance (P \ 0.0001). As shown in
Table 2, the measure of asset ownership was not associated with changes in symptoms of anxiety and depression, while one of the measures of food production was
(number of crops planted, P 5 0.01), but the strength of
the association between anxiety and depression symptoms and food insecurity remained virtually unchanged
after the inclusion of these covariates. Household size
and the caretaker’s education level were also not associated with changes in maternal anxiety and depression
symptoms but Pimbwe ethnicity was associated with
greater number of symptoms. An additional test of
whether increasing food insecurity was associated with
increases in symptoms of anxiety and depression used a
dichotomous variable signaling whether the household’s
food insecurity status increased across study rounds. In
this model, which used the same covariates, an increase
in food insecurity was positively associated with changes
in anxiety and depressive symptoms (P \ 0.01). In none
of the models was the interaction between ethnic group
and changes in food insecurity significant (P [ 0.05).
Finally, a cross-sectional analysis using the maternal
HSCL in the wet season produced nearly identical
results (results not shown).
DISCUSSION
To the best of our knowledge, this article is the first
longitudinal study reporting an association between seasonal changes in food insecurity and seasonal changes in
a measure of anxiety and depression in a rural African
setting. The key contribution of this study is the finding
that changes in a household’s food insecurity status are
associated with concomitant changes in a mother’s reporting of symptoms of anxiety and depression. The
strengths of the study lie in the sample which included
randomly selected women from two ethnic groups, as
well as the longitudinal nature of the study design, and
the ethnographic work that underlies the theoretical
framework and study design. We exploited our understanding of the ethnography and ecology to make predictions about the changing relationship between symptoms
of common mental health disorders and limited access to
American Journal of Physical Anthropology—DOI 10.1002/ajpa
230
C. HADLEY AND C.L. PATIL
food across seasons of measurement. This study contributes to large amount of work by physical anthropologists
focusing on the physical growth consequences of life in
seasonal environments by showing that maternal symptoms of anxiety and depression vary with seasonal food
insecurity.
This study extends previous work in several ways.
Existing studies linking symptoms of common mental
health disorders have been somewhat limited in that
they are generally cross-sectional in nature, which
makes teasing apart the direction of causality difficult;
in reality the arrow of causation is likely bi-directional
to some degree. Previous studies also focused primarily
on urban environments and on indicators of insecurity
that are closely linked to market-based economies. Our
study results show that insecurity in bioculturally important domains such as certain and secure access to
sufficient food is indeed a risk factor for the development
of symptoms of depression and anxiety. The results from
studies in developing countries illustrate the role of the
local biocultural environment in structuring risk factors
for common mental health disorders. By this we mean
that the patterns we identified in this setting are likely
unique because of (1) the availability and level of technology, (2) the rural subsistence nature of the study
area, and (3) the unpredictable and seasonal rainfall in
the area. Our argument would predict that the seasonal
nature of anxiety and depression symptoms observed
would disappear if alternative food sources were available and households had resources available to them to
acquire those alternatives during the food insecure wet
season. The argument would also predict that limited
education would become an increasingly salient risk
factor for symptoms of anxiety and depression as the
economy shifts towards one based on skills, and therefore rewards education. For better or worse, this latter
prediction may soon be testable as development schemes
and improved roads find their way into this rural area.
There are many limitations to this study as well, including attrition. A more challenging limitation is the
confounding that occurs between measures of wealth
and ethnicity. Sukuma as a group are wealthier than
Pimbwe. Statistically this makes controlling for wealth
in regression models difficult. Solutions to this problem
are both methodological and analytical. Methodologically
a study design that attended more directly to matching
characteristics of women might be beneficial. Analytical
techniques, such as propensity score matching may be
useful; unfortunately our study samples were too small
to reliably employ this technique.
The key finding of the study is that increases in food
insecurity were associated with increased symptoms of
anxiety and depression. This has interesting policy and
theoretical implications. Given some work in the USA
(Whitaker et al., 2006) and ethnographic work globally
(Dirks, 1990) this may not be too surprising. Yet, the
food insecurity literature has been dominated by a focus
on nutritional outcomes despite emerging recognition
that mental health outcomes are important sources of
disability. That food insecurity is also associated with
mental health outcomes is an important, policy-relevant
finding and may help broaden the evidence-base for the
health robbing impacts of food insecurity globally. The
other implication of our finding is although the food insecurity-mental health connection may hold at the individual level, population-level differences in food insecurity
mean that certain groups are likely more affected. In the
study here, women in both groups were similarly affected by increased food insecurity, but the burden of increased food insecurity fell disproportionately on Pimbwe
women. To the extent that subsistence and social systems structure the risk of food insecurity they will create
differential group-level outcomes.
The finding that material assets are not associated
with maternal anxiety and depression but food insecurity is suggests that it is food deprivation and not ownership of material goods that is responsible for some of the
variation observed in this sample. This may not be too
surprising given the role that women in these communities (and many others throughout sub-Saharan Africa)
play in food production, preparation, and feeding young
children (Quisumbing et al., 1995). Women may be particularly sensitive to fluctuations and unpredictability in
access to foods. These fluctuations may be augmented by
sociocultural norms that prioritize men’s access to high
quality foods (Messer, 1997). We might also hypothesize
that men’s anxiety and depression may be more closely
associated with relative inequality and material ownership, whereas women’s mental health disorders may be
more affected by periods of food insecurity and deprivation. Future research should test this hypothesis by collecting similar data for men and women, and examine
whether a shared set of covariates predicts both women’s
and men’s anxiety and depression.
Future studies will also do well to examine the extent
to which maternal anxiety and depression is predicted
by the presence or absence of kin networks and other
informal forms of social support, and how this might
vary across differing levels of socioeconomic position.
Examining correlates beyond wealth and education of
anxiety and depression will also be critical to understand
the ethnic difference in anxiety and depression that we
observed in these data during the wet season round of
measurement. Indeed, Patel et al. note in their review of
mental health correlates that in developing countries
limited social support may be a key contributor to common mental health disorders (Patel et al., 2006). Preliminary data from our own studies suggest social support is
associated with the levels of anxiety and depression
observed among study participants in rural Tanzania
(Patil and Hadley, 2006). Examining these noneconomic
correlates may provide insight into why the values of
anxiety and depression actually declined for many
Sukuma women. There are a number of hypotheses that
may explain this time-dependent reduction including
decreasing maternal workloads. For many Sukuma
households obtaining water from the river, which is 2
km or more from most Sukuma households, is an arduous and time-consuming task. As the rains come the
task of obtaining water eases as pools and puddles are
used rather than the river. For Pimbwe, the wet season
brings with it not only declining food supplies but a
large increase in workloads, as almost all farming is
done by hand and the small household sizes mean little
additional help with farm labor. Sukuma women do little
of their own planting and because Sukuma households
have access to cattle, most plowing is a task done by animals, not people. Sukuma also farm in cooperatives further reducing the energetic and time demands on individuals. These seasonally contingent and ethnic-group
specific work patterns may be responsible in turn for
shifting levels of anxiety and depression and exemplify
the intimate connections between overall well-being and
one’s biocultural milieu.
American Journal of Physical Anthropology—DOI 10.1002/ajpa
FOOD INSECURITY AND ANXIETY AND DEPRESSION
The finding that women’s anxiety and depression
symptomology is associated with their food security situation also has implications for children’s health and
nutritional status. Recognition that maternal health may
be critical for child health has been noted for sometime
(Winkvist, 1995) but it is only recently that there has
emerged a solid evidence-base in developing countries
identifying an association between maternal mental
health and children’s growth (Rahman et al., 2002; Patel
et al., 2004; Grantham-McGregor et al., 2007). However,
the mechanisms linking maternal mental health to children’s growth are not yet well understood (WHO, 2004).
It is also unlikely that children’s nutritional status is
singularly affected by maternal mental health; rather as
several recent studies have shown, physical growth and
social and emotional outcomes are likely to be closely
inter-related (Fernald and Grantham-McGregor, 2002).
Our previous work in these villages has revealed tremendous differences in the nutritional status between
Pimbwe and Sukuma children with Sukuma children
being significantly taller and heavier than Pimbwe children. The disparities in anxiety and depression symptoms may be in part responsible for these growth differences. Anthropologists may want to think more broadly
about the pyschosocial and neuropsychiatric causes of
physical growth outcomes. In particular, investigators
should examine the extent to which children’s nutritional status is associated with maternal anxiety and
depression and, in turn, the extent to which children’s
social and emotional outcomes are also affected.
In this article, we have attempted to make some
inroads into incorporating mental health into the study
of biological anthropology. We have argued that locally
salient measures of insecurity are likely to be highly
predictive of symptoms of anxiety and depression. We
showed that some of the variation in symptoms of anxiety and depression are related to the level of food insecurity and that changes in food insecurity are associated
with changes in symptoms of anxiety and depression.
This relationship holds when confronted with other covariates. The results also highlight the mechanisms
through which group-level differences in subsistence
strategies likely influence food insecurity, which in turn
can create group-level differences in health outcomes.
We anticipate building on these results to explore the
functional consequences of symptoms of anxiety and
depression in individuals and their networks.
LITERATURE CITED
Bickel G, Nord M, Price C, Hamilton W, Cook J. 2000. Guide to
measuring food insecurity. Alexandria, VA: Department of
Agriculture, Food, and Nutrition Service.
Coates J, Frongillo EA, Rogers BL, Webb P, Wild PE, Houser
RF. 2006. Commonalities in the experience of household food
insecurity across cultures: what are measures missing?
J Nutr 136:1438S–1448S.
Coates J, Wilde PE, Webb P, Rogers BL, Houser RF. 2006. Comparison of a qualitative and a quantitative approach to developing a household food insecurity scale for Bangladesh.
J Nutr 136:1420S–1430S.
Cohen S, Underwood L, Gottlieb B. 2000. Social support
measurement and intervention: a guide for health and social
scientists. Oxford: Oxford University Press.
Dirks R. 1980. Social responses during food shortages and famine. Curr Anthropol 21:21–44.
FAO. 2005. The state of food insecurity in the world. Rome: FAO.
231
Fernald LC, Grantham-McGregor SM. 2002. Growth retardation
is associated with changes in the stress response system and
behavior in school-aged Jamaican children. J Nutr 132:3674–3679.
Ferro-Luzzi A. 1990. Seasonality studies in three developing
countries: introduction and background. Eur J Clin Nutr 44
(Suppl 1):3–6.
Frongillo EA, Nanama S. 2006. Development and validation of
an experience-based measure of household food insecurity
within and across seasons in northern Burkina Faso. J Nutr
136:1409S–1419S.
Grantham-McGregor S, Cheung YB, Cueto S, Glewwe P, Richter
L, Strupp B. 2007. Developmental potential in the first 5
years for children in developing countries. Lancet 369:60–70.
Hadley C. 2005a. The costs and benefits of kin: kin networks
and children’s health among the Pimbwe of Tanzania. Hum
Nat 15:377–395.
Hadley C. 2005b. Ethnic expansions and between-group differences in children’s health: a case study from the Rukwa Valley, Tanzania. Am J Phys Anthropol 128:682–692.
Hadley C, Borgerhoff Mulder M, Fitzherbert E. 2007. Seasonal
food insecurity and perceived social support in rural Tanzania. Public Health Nutr 10:544–551.
Hadley C, Patil CL. 2006. Food insecurity in rural Tanzania is
associated with maternal anxiety and depression. Am J Hum
Biol 18:359–368.
Kaaya SF, Fawzi MC, Mbwambo JK, Lee B, Msamanga GI,
Fawzi W. 2002. Validity of the Hopkins Symptom Checklist-25
amongst HIV-positive pregnant women in Tanzania. Acta Psychiatr Scand 106:9–19.
Macro. 2005. Tanzania Demographic and Health Survey 2004–
5. Dar es Salaam, Tanzania: National Bureau of Statistics
and ORC Macro.
McElroy A, Townsend PK. 2004. Medical anthropology in ecological perspective. Boulder, CO: Westview Press.
Melgar-Quinonez HR, Zubieta AC, MkNelly B, Nteziyaremye A,
Gerardo MF, Dunford C. 2006. Household food insecurity and
food expenditure in Bolivia, Burkina Faso, and the Philippines. J Nutr 136:1431S–1437S.
Messer E. 1997. Intra-household allocation of food and health
care: current findings and understandings—introduction. Soc
Sci Med 44:1675–1684.
Mollica RF, McDonald L, Massagli M, Silove D. 2004. Measuring trauma, measuring torture. Cambridge: Harvard Program in Refugee Studies.
Murray CJ, Lopez AD. 1997. Global mortality, disability, and
the contribution of risk factors: Global Burden of Disease
Study. Lancet 349:1436–1442.
Patel V, Araya R, de Lima M, Ludermir A, Todd C. 1999.
Women, poverty and common mental disorders in four
restructuring societies. Soc Sci Med 49:1461–1471.
Patel V, Flisher A, Cohen A. 2006. Mental health. In: Merson
MH, Black RE, Mills AJ, editors. International public health:
diseases, programs, systems, and policies. Gaithersburg: Aspen.
p 355–392.
Patel V, Kleinman A. 2003. Poverty and common mental disorders in developing countries. Bull World Health Organ
81:609–615.
Patel V, Rahman A, Jacob KS, Hughes M. 2004. Effect of maternal mental health on infant growth in low income countries:
new evidence from South Asia. BMJ 328:820–823.
Patil CL, Hadley C. 2006. Wealth and social support among four
ethnic groups in Tanzania: linkages with depression and anxiety. American Anthropological Association Conference 2006,
San Jose, CA, Nov 19, 2006.
Pike IL. 2004. The biosocial consequences of life on the run: a
case study from Turkana District, Kenya. Hum Organ 63:
221–235.
Pike IL, Patil CL. 2006. Understanding women’s burdens: preliminary findings on psychosocial health among Datoga and
Iraqw women of northern Tanzania. Cult Med Psychiatry
30:299–330.
Pike IL, Williams SR. 2006. Incorporating psychosocial health
into biocultural models: preliminary findings from Turkana
women of Kenya. Am J Hum Biol 18:729–740.
American Journal of Physical Anthropology—DOI 10.1002/ajpa
232
C. HADLEY AND C.L. PATIL
Quisumbing A, Brown L, Feldstein H, Haddad L, Pena C. 1995.
Women: the key to food security. Washington, DC: IFPRI.
Rahman A, Harrington R, Bunn J. 2002. Can maternal depression increase infant risk of illness and growth impairment in
developing countries? Child Care Health Dev 28:51–56.
Richards AJ. 1939. Land, labour and diet in northern Rhodesia;
an economic study of the Bemba tribe. New York: Oxford University Press.
Scrimshaw S. 2006. Culture, behavior, and health. In: Merson
MH, Black RE, Mills AJ, editors. International public health:
diseases, programs, systems, and policies. Gaithersburg, MD:
Aspen. p 43–70.
Shipton P. 1990. African famines and food insecurity: anthropological perspectives. Annu Rev Anthropol 19:353–394.
Siefert K, Heflin CM, Corcoran ME, Williams DR. 2004. Food
insufficiency and physical and mental health in a longitudinal
survey of welfare recipients. J Health Soc Behav 45:171–
186.
Sohr-Preston SL, Scaramella LV. 2006. Implications of timing of
maternal depressive symptoms for early cognitive and language development. Clin Child Fam Psychol Rev 9:65–83.
Tomkins A. 1993. Environment, season and infection. In: Ulijaszek SJ, Strickland SS, editors. Seasonality and human eco-
logy: 35th Symposium Volume of the Society for the Study of
Human Biology. New York: Cambridge University Press. p
123–134.
Trostle JA. 2005. Epidemiology and culture. New York: Cambridge University Press.
Ulijaszek SJ, Strickland SS. 1993. Seasonality and human ecology: 35th Symposium Volume of the Society for the Study
of Human Biology. New York: Cambridge University Press.
Webb P, Coates J, Frongillo EA, Rogers BL, Swindale A, Bilinsky P. 2006. Measuring household food insecurity: why it’s so
important and yet so difficult to do. J Nutr 136:1404S–1408S.
WHO. 2001. The world health report. Mental health: new
understandings, new hope. Geneva: WHO.
WHO. 2004. The importance of caregiver-child interactions for
the survival and healthy development of young children: a
review. Geneva: WHO.
Winkvist A. 1995. Health and nutrition status of the caregiver:
effect on caregiving capacity. Food Nutr Bull 16:389–406.
Whitaker RC, Phillips SM, Orzol SM. 2006. Food insecurity and
the risks of depression and anxiety in mothers and behavior
problems in their preschool-aged children. Pediatrics 118:
e859–e868.
American Journal of Physical Anthropology—DOI 10.1002/ajpa
Документ
Категория
Без категории
Просмотров
2
Размер файла
145 Кб
Теги
anxiety, change, household, food, symptom, insecurity, depression, seasonal
1/--страниц
Пожаловаться на содержимое документа