Depression and suicide ideation in late adolescence and early adulthood are an outcome of child hunger

Journal of Affective Disorders 150 (2013) 123–129

Contents lists available at SciVerse ScienceDirect

Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad

Preliminary communication

Depression and suicide ideation in late adolescence and early adulthood are an outcome of child hunger Lynn McIntyre a,1,n, Jeanne V.A. Williams a, Dina H. Lavorato a, Scott Patten a,b,1 a b

Department of Community Health Sciences, Faculty of Medicine, University of Calgary, Calgary, AB, Canada T2N 4Z6 Hotchkiss Brain Institute, University of Calgary, Calgary, AB, Canada, T2N 4Z6

a r t i c l e i n f o

abstract

Article history: Received 23 October 2012 Accepted 6 November 2012 Available online 29 December 2012

Background: Child hunger represents an adverse experience that could contribute to mental health problems in later life. The objectives of this study were to: (1) examine the long-term effects of the reported experience of child hunger on late adolescence and young adult mental health outcomes; and (2) model the independent contribution of the child hunger experience to these long-term mental health outcomes in consideration of other experiences of child disadvantage. Methods: Using logistic regression, we analyzed data from the Canadian National Longitudinal Survey of Children and Youth covering 1994 through 2008/2009, with data on hunger and other exposures drawn from NLSCY Cycle 1 (1994) through Cycle 7 (2006/2007) and mental health data drawn from Cycle 8 (2008/2009). Our main mental health outcome was a composite measure of depression and suicidal ideation. Results: The prevalence of child hunger was 5.7% (95% CI 5.0–6.4). Child hunger was a robust predictor of depression and suicidal ideation [crude OR ¼ 2.9 (95% CI 1.4–5.8)] even after adjustment for potential confounding variables, OR ¼ 2.3 (95% CI 1.2–4.3). Limitations: A single question was used to assess child hunger, which itself is a rare extreme manifestation of food insecurity; thus, the spectrum of child food insecurity was not examined, and the rarity of hunger constrained statistical power. Conclusions: Child hunger appears to be a modifiable risk factor for depression and related suicide ideation in late adolescence and early adulthood, therefore prevention through the detection of such children and remedy of their circumstances may be an avenue to improve adult mental health. & 2012 Elsevier B.V. All rights reserved.

Keywords: Depression Suicide Youth Longitudinal Population studies

1. Introduction Food insecurity refers to the financial inability of households to access adequate food and is generally measured as moderate versus severe (Health Canada, 2007). Moderate food insecurity means that there is an indication that quality and/or quantity of food consumed have been compromised. Severe food insecurity means that there is an indication of reduced food intake and disrupted eating patterns. According to the 2007–2008 Canadian Community Health Survey (CCHS), the national prevalence of food insecurity is 8% for households overall, while the child-level of food insecurity is 5% (Statistics Canada, 2010a). Severe child food insecurity, analogous to the more emotive term ‘child hunger’,

n Correspondence to: Dept. of Community Health Sciences, Faculty of Medicine, University of Calgary, Teaching Research & Wellness (TRW) Building Room 3E14 (3rd Floor), Calgary, AB T2N 4Z6. Tel.: þ 1 403 220 8664; fax: þ1 403 270 7307. E-mail address: [email protected] (L. McIntyre). 1 Authors contributed equally to this study.

0165-0327/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jad.2012.11.029

occurs in about 2% of Canadian children (Statistics Canada, 2010a). Population-based studies have demonstrated that food insecurity affects specific vulnerable populations with well-described characteristics related to gender (women/mothers), age (children4 adults), household composition (lone parent-led), household income (inadequate, poverty level), housing (rented), ethnicity (aboriginal off-reserve), and main income source (social assistance) (Che and Chen, 2001; Health Canada, 2007; Rainville and Brink, 2001; Statistics Canada, 2010a). Thus, children affected by severe food insecurity likely live in conditions of general child adversity. There is good evidence from Canada and the United States that demonstrates that household food insecurity is associated, in cross-sectional studies, with a range of poor physical health outcomes among children (e.g., Alaimo et al., 2001; McIntyre et al., 2000; Broughton et al., 2005; Gundersen and Kreider, 2009; To et al., 2004). A sparser literature has shown that food insecurity in childhood has long-term negative impacts on physical health (Casey et al., 2010), including chronic conditions such as asthma,

124

L. McIntyre et al. / Journal of Affective Disorders 150 (2013) 123–129

even when controlling for confounding conditions of baseline health, preexisting chronic conditions and other household markers of disadvantage such as low income and rental housing (Kirkpatrick et al., 2010). Less is understood about the food insecurity- mental health relationship. Cross-sectional studies have demonstrated a relationship between child food insecurity and child and youth mental health problems. Weinreb et al. (2002) found that school-aged children with severe hunger had significantly higher anxiety and internalizing behaviors independent of other associations. Alaimo et al. (2001) examined dysthymia and suicidal ideation in food insufficient adolescents and found a significant independent association. A small prospective study from the United Kingdom found that food insecure children, followed for 2 to 5 years until age 12, had moderately higher levels of emotional problems relative to food secure children when adjusted for various household environmental and maternal variables (Belsky et al., 2010). Another two year follow up study of children 4 to 14 years in the United States found that children from food insecure households were significantly more likely to have internalizing and externalizing problems, independent of poverty status (Slopen et al., 2010). A burgeoning literature on food insecurity among women deals with the impact of mothers’ depressive symptomatology on adverse child outcomes such as problem behaviors (Whitaker et al., 2006) as well as her depression appearing to perpetuate the household living in a food insecure state (Casey et al., 2004; Melchior et al., 2009). Such studies add to well-established evidence that maternal depression is associated with a small but consistently increased risk of behavioral, emotional, and developmental problems in their children across the developmental age span (Goodman et al., 2010). Maternal depression is usually considered to contribute to these associations through poor parenting and parental interactions with their children (Goodman et al., 2010). From an epidemiological point of view, only a longer term prospective study can untangle the causal connections between child hunger and mental health of both the mother and the child. Whitaker et al.’s (2006) early life stress hypothesis would predict longer term mental health problems resulting from the childhood experience of food insecurity. They have called for studies to examine whether or not household food insecurity does increase a child’s susceptibility to later mental health problems. Slopen et al. (2010) also felt that their short-term follow up study of child disadvantage and behavioral problems implicated food insecurity as a novel risk factor for child mental well-being; suggesting that if causal, this might motivate prevention efforts. We recognized that the National Longitudinal Survey of Children and Youth (NLSCY) presented a unique opportunity to study the long term impact of child and youth hunger on mental health in late adolescence and young adulthood. This is because of the richness of the dataset which includes a consistent hunger measure over sixteen years of follow up and a variety of sociodemographic variables and diverse health outcomes. The objectives of the study were therefore to: (1) examine the long-term effects of the reported experience of child/youth hunger on late adolescence and young adult mental health outcomes; and (2) model the independent contribution of the child/youth hunger experience to these long-term mental health outcomes in consideration of other experiences of child disadvantage.

2. Methods The NLSCY was a long-term study conducted jointly by Statistics Canada and Human Resources and Skills Development

Canada (HRSDC), which collected detailed data on the health, education, social development, and well-being of a representative sample of Canadian children and youth on a biennial basis from 1994 until 2009. The survey included both crosssectional and longitudinal components. The longitudinal cohort is identified in Cycle 2 using a variable indicating that the record is for a longitudinal child. This cohort of children was aged 0 to 11 years in Cycle 1 (n ¼15,468); by Cycle 8 (2008/2009) they had reached 14 to 25 years. 68% of the original cohort (n ¼22,831) participated in Cycle 8, and 52% participated in every follow up.

2.1. Data and measures We analyzed data covering 1994 through 2008/2009, with data on hunger drawn from NLSCY Cycle 1 (C1, 1994) through Cycle 7 (C7, 2006/2007) and mental health outcome data drawn from Cycle 8 (C8, 2008/2009). Over the course of 8 cycles, both the Person Most Knowledgeable about the child (PMK) reports and youth self-reports were collected; where there were discrepancies, we privileged youth over PMK reports. The C8 data were entirely youth-reported rather than PMK-reported. The hunger question was left intact over all cycles: ‘‘Has [the child]/Have you ever experienced being hungry because the family has run out of food or money to buy food?’’ with response categories yes or no. The only exception was in C5 when the question was modified for youth 18–19 years by changing the time prior to ‘past six months’ from ‘ever’. Although the hunger experience timeframe is ‘ever’, previous analyses show variation in reporting across cycles in relation to changes in known risk factors for food insecurity (e.g., income and household composition) (McIntyre et al., 2001; Kirkpatrick et al., 2010), suggesting that the reference period is likely interpreted as the time elapsing since the last administration of the survey. A respondent was considered to have ever experienced hunger (‘ever hungry’) if hunger was reported for at least one cycle by the PMK from C1 to C7 or by the youth between C4 and C7. Sociodemographic information for all respondents was retrieved from all 8 cycles. Household sociodemographic covariates over C1–C7 could confound associations in the mental health outcome model; hence, their effects were smoothed over time. For example, we derived average household income (‘permanent income’), and mean number of adults and children in the household over the exposure period. Other exposures were distilled to ever/never exposures, e.g., ever lived with a lone mother. The key mental health outcome variable was depression, which was assessed in the NLSCY using an abbreviated version of the Center for Epidemiologic Studies Depression Rating Scale (CES-D). It was available for the full sample aged 14–25 years. A cut-off point of 21 was used to classify respondents into depressed or nondepressed categories. Because of high collinearity and to maximize sample size, this variable was combined with suicide ideation (yes/ no), in which the youth was asked whether they had seriously considered or attempted suicide. The dataset included other scales that were related to the mental health of youth and young adults, specifically General Self-image (self-esteem and youth experiences), and Emotional Quotient (measure of emotional intelligence). These were examined as secondary mental health outcomes to determine the specificity of the depression/suicide ideation composite outcome for late adolescents and young adults who experienced child/youth hunger. A brief description of these measures is found in Table 1. In the absence of a meaningful diagnostic threshold for General Selfimage and Emotional Quotient scores, the lower quartile was used in the analysis, indicating poor outcome.

L. McIntyre et al. / Journal of Affective Disorders 150 (2013) 123–129

125

Table 1 Description of scales and indicators used to assess mental health outcomes for youth and young adults, NLSCY Cycle 8. Source: Statistics Canada, 2010b. Name of measure

Age group Description (years)

Depression scale

16–25

Feelings and behaviors General self-image scale

14–19 22–23 14–19

Emotional quotient scales

22–23 14–17

Emotional quotient Inventory youth version (14–17) emotional quotient adult version (21–25)

Shorter version of the depression scale (CES-D), comprising 12 questions. Measures the occurrence and severity of symptoms associated with depression during the previous week. Thoughts of suicide (yes/no). Adapted from the 1992 British Columbia Adolescent Health Survey General self-image scale of the Marsh Self-Description Questionnaire. Measures youth’s overall self-esteem. Measures of emotional intelligence. Comprised of five major dimensions: intrapersonal, interpersonal, adaptability, stress management and general mood.

20–21 24–25

2.2. Statistical analyses We used the first 7 cycles as a means of defining the exposure in relation to the presence (or lack) of specific mental health outcomes at C8, using logistic regression. These models controlled for age and sex, whether the child ever lived without the biological mother, depression of the biological mother when she was the PMK for the survey cycle, ever lived with a lone mother, highest level of education attained by the PMK, whether the household always lived in rental accommodation, permanent (average) household income, number of people in the household, and whether the household ever lived in a rural area. In addition, we tested interactions in the models, e.g., child sex and PMK depression (as no interactions were found, the data are not reported). Analyses were conducted using STATA 11.0 data analysis and statistical software (Statacorp, College Station, TX). All analyses were weighted using longitudinal weights provided for this purpose by Statistics Canada, and standard error estimates were generated using a bootstrap replication process to account for the complex survey design of the NLSCY. All estimates were required to be subjected to Statistics Canada disclosure rules for small cell sizes. The study received ethical approval from the University of Calgary/Alberta Health Services (Calgary Zone) Conjoint Health Research Ethics Board.

3. Results Table 2 presents the mental health outcomes, sociodemographic characteristics, and baseline covariates of the longitudinal sample. Overall, the prevalence of child/youth experience of hunger was 5.7% (95% CI 5.0–6.4), including 5.2% (95% CI 4.3– 6.0) in male respondents and 6.3% (95% CI 5.3–7.3) in female respondents. The prevalence of the composite depression/suicide ideation variable at C8 was 2.4% (95% CI 2.0–2.8) overall, 1.4% (95% CI 0.9–1.9) in male respondents and significantly higher at 3.3% (95% CI 2.5–4.0) in female respondents. Table 3 presents odds ratios for depression/suicide ideation in late adolescence and young adulthood adjusted for each covariate. The unadjusted odds ratio for the association of the experience of child/youth hunger and depression/suicide ideation was 2.9 (95% CI 1.4–5.8). Other factors with similar strength of association were depression in the biological mother, ever living without one’s biological mother, and female gender. The righthand column of Table 3 presents odds ratios for hunger adjusted for each of these covariates. None of these adjustments substantially altered the association of child/youth hunger with

depression/suicide, indicating that the association was not due to confounding by these variables Table 4 presents two logistic regression models that examine the independent contribution of child/youth experience of hunger as a risk for depression/suicide ideation in late adolescence or young adulthood. The first model includes all covariates, and the reduced model includes only those that were statistically significant. The results from one model to the other are essentially unchanged and the effect of the experience of child/youth hunger is essentially unchanged after adjustment for these variables. In fact, child/hunger experience shared over a two-fold risk (OR¼2.5, 95% CI 1.3–4.9) with three other covariates, including depression in the biological mother, child ever lived without the biological mother, and female gender. In contrast, neither General Self-image (OR¼1.4, 95% CI 0.9– 2.1) (Table 5), nor Emotional Quotient in either those aged 14 to 17 (OR ¼1.7, 95% CI 0.7–4.4) or the older age group (20–21, 24–25 years) (OR ¼1.2, 95% CI 0.7–2.0) or combined (OR ¼1.2, 95% CI 0.7–2.0) (Table 6) were significantly associated with the experience of child/youth hunger. This did not change with adjustment for covariates, as shown in Tables 5 and 6.

4. Discussion Our results demonstrate that the experience of child/youth hunger is an independent risk factor for subsequent depression/ suicide ideation during late adolescence and young adulthood. Because of the longitudinal nature of the study, inferences on causation between hunger as an adverse childhood exposure and depression as a negative mental health outcome are suggested. In examining covariates that are predictive of depression/suicide ideation in late adolescence and young adulthood without consideration of hunger, it appears that hunger displaces many of the sociodemographic variables usually used to characterize child disadvantage, such as low household income and parental education, living in rental housing and with more household members, and draws attention to factors that might have more of a sociobiological basis. Rutter (2003) has highlighted issues related to social selection versus social causation that have been considered in studies of poverty and mental health. Even for children where social selection may be less of an issue than for adults, there remain concerns about genetic predisposition and parenting as direct effects on child mental health versus poverty’s larger social deprivations. The role of parenting and family functioning as mediators in the poverty-child-behavioral problem hypothesis also remains at issue. Our study included youth whose PMK was almost always the biological parent (usually the mother); we further adjusted

126

L. McIntyre et al. / Journal of Affective Disorders 150 (2013) 123–129

Table 2 Mental health outcomes and covariates, Cycle 8. Mental health outcomes

Age (yrs)

Depression/thoughts of suicide Yes No General self-image lower quartile

14–25

Proportion (%)

95% CI

2.4 97.6

2.0–2.8 97.2–98.0

34.2 65.8

32.3–36.1 63.9–67.7

27.8 72.2

25.0–30.5 69.5–75.0

23.8 76.2

21.4–26.2 73.8–78.6

5.7 94.3 19.3

5.0–6.4 93.6–95.0 19.3–19.4

14–19 22–23

Yes No Emotional Quotient 5 Factor Youth lower quartile

20–21 24–25

Yes No Self complete lower quartile Yes No Covariates Hunger ever Cycles 1–7 (PMK, youth, self-complete) Yes No Age at Cycle 8 (mean) Gender Male Female Permanent household income (mean)($) Number of individuals in household—average Number of individuals above household mean (44.3) Education of the PMK o High school graduation High school graduation Some post-secondary Completed trade or community college Bachelor/post-graduate degree Urban/rural Rural Urban Housing tenure: always rented Owned Always rented Ever lived without biological mother Ever lived with lone mother Depression of biological mother as PMK

14–17

51.2 48.8 68,105 4.3 43.1

50.8–51.7 48.3–49.2 68,735–69,475 4.3–4.3 41.5–44.7

15.4 18.8 28.4 21.0 16.3

14.2–16.6 17.7–20.0 27.0–29.9 19.7–22.3 15.0–17.6

19.0 81.0

18.2–19.9 80.1–81.8

88.7 11.3 26.9 26.3 7.5

87.6–89.8 10.2–12.4 25.6–28.3 25.0–27.7 6.7–8.3

Table 3 Risk factors for depression/suicide ideation in late adolescence and young adulthood: unadjusted odds ratios for hunger and covariates, and covariate-adjusted odds ratios for hunger. Odds ratios (unadjusted) (95% CI)

Hunger Covariates Sex (female) Age at Cycle 8 Household income per $10,000 Household number Household mean (44.3) Education of PMKn o High school graduate High school graduate Some post-secondary Completed trade/community college Bachelor/post-graduate degree Urban Housing tenure: always rented Ever lived without biological mother Ever lived with lone mother Depression of biological mother as PMK n

P-value

2.9(1.4–5.8)

0.003

2.3(1.5–3.6) 1.0(0.9–1.0) 0.9(0.9–1.0) 0.9(0.7–1.1) 0.9(0.6–1.4)

o 0.001 0.659 0.024 0.374 0.614

1.3(0.7–2.7) 1.4(0.7–3.0) 1.6(0.8–3.3) 1.0(0.5–2.1) Reference 1.1(0.7–1.7) 1.8(0.9–3.5) 1.9(1.3–2.8) 1.2(0.8–1.9) 2.3(1.3–4.1)

0.432 0.370 0.209 0.993 0.565 0.091 0.001 0.384 0.004

Odds ratio for child/youth hunger, adjusted for each covariate (95% CI)

P-value

2.8(1.4–5.6) 2.9(1.5–5.9) 2.6(1.3–5.6) 2.8(1.4–5.6) 2.9(1.4–5.8) 2.7(1.4–5.5)n

0.004 0.002 0.007 0.003 0.003 0.004

Reference 2.9(1.4–5.8) 2.7(1.4–5.2) 2.9(1.5–5.8) 2.8(1.4–5.6) 2.6(1.3–5.1)

0.003 0.004 0.003 0.003 0.006

Education dichotomized against bachelor degree.

for PMK (maternal) depression, thus controlling for in effect social selection which slightly attenuated but did not eliminate the

independent finding of child experience of hunger being socially causal for depression/suicide ideation.

L. McIntyre et al. / Journal of Affective Disorders 150 (2013) 123–129

127

Table 4 Child/youth hunger experience as a risk factor for depression/suicide ideation in late adolescence or young adulthoodn, all inclusive and reduced logistic regression models. Model 1 all inclusive

Hunger Sex (female) Age at Cycle 8 Household income per $10,000 Household number Household mean (4 4.3) Education of PMK ohigh school graduate high school graduate some post-secondary completed trade/community college bachelor/post-graduate degree Urban Housing tenure: always rented Ever lived without biological mother Ever lived with lone mother Depression of biological mother as PMK n

Model 2 reduced

Odds ratio (95% CI)

P-value

Odds ratio (95% CI)

P-value

2.3(1.2–4.3) 2.4(1.5–3.7) 1.0(0.9–1.0) 1.0(0.9–1.0) 0.9(0.6–1.4) 1.1(0.5–2.4)

0.01 o0.001 0.690 0.202 0.680 0.756

2.5(1.3–4.9) 2.4(1.5–3.7)

0.006 o0.001

0.9(0.4–2.0) 1.2(0.5–2.6) 1.4(0.6–2.8) 0.9(0.4–2.0) Reference 1.1(0.7–1.8) 1.4(0.7–2.8) 2.1(1.4–3.1) 0.9(0.5–1.5) 2.0(1.2–3.4)

0.851 0.679 0.425 0.829

2.1(1.4–3.1)

o0.001

2.1(1.2–3.6)

0.007

0.578 0.311 0.001 0.561 0.010

Note that outcome data are collected from youth and young adults 14–25 years.

Table 5 Child/youth hunger experience as a risk factor for low general self-image in late adolescence or young adulthoodn, all inclusive and reduced logistic regression models. Model 1 all inclusive

Hunger Sex (female) Age at Cycle 8 Household income per $10,000 Household number Household mean (4 4.3) Education of PMK o high school graduate high school graduate some post-secondary completed trade/community college bachelor/post-graduate degree Urban Housing tenure: always rented Ever lived without biological mother Ever lived with lone mother Depression of PMK n

Model 2 reduced

Odds Ratio (95% CI)

P-value

Odds ratio (95% CI)

P-value

1.4(0.9–2.1) 1.3(1.1–1.6) 0.9(0.9–0.9) 1.0(0.95–1.01) 1.0(0.8–1.1) 1.1(0.8–1.4)

0.124 0.004 o 0.001 0.189 0.826 0.635

1.4(1.0–2.1) 1.3(1.1–1.6) 0.9(0.9–0.9)

0.076 0.003 o 0.001

1.5(1.1–2.0)

0.015

1.0(0.7–1.5) 1.0(0.7–1.3) 1.1(0.8–1.4) 0.9(0.7–1.2) Reference 1.2(1.0–1.4) 1.1(0.7–1.7) 1.1(0.9–1.4) 0.9(0.8–1.2) 1.4(1.0–1.9)

0.856 0.762 0.723 0.575 0.048 0.681 0.393 0.600 0.025

Note that outcome data collected from youth and young adults 14–19 and 22–23 years, respectively.

Our regression model supports a partial genetic basis for depression/suicide ideation, in line with our finding of a history of depression in the biological mother, notwithstanding a wellcharacterized socioenvironmental effect of maternal depression on children (Whitaker et al., 2006; Goodman et al., 2010). We also suggest that our model points to the contribution of early childhood development adversity and later depression/suicide ideation through mechanisms such as attachment problems (Cassidy and Shaver, 2008) that might be seen in children who are removed from their biological mothers for some or all of their upbringing. The finding that depression is significantly more common in women than in men aligns with the population-level prevalence of the problem which appears to have a sociobiological basis (Nolen-Hoeksema et al., 1999; Wilhelm et al., 2008). This leaves the independent effect of the child/youth hunger experience which is as strong as the other factors and seemingly unique among them. One mechanism by which child/youth hunger might lead to depression in later life might be through nutritional deprivation, e.g., chronic iron deficiency anemia (Rangan et al., 1998) or folate deficiency (Alpert et al., 2000). There is absolutely no evidence to

suggest that in Canada the report of child hunger is associated with macronutrient malnutrition or serious micronutrient deficiencies (Glanville and McIntyre, 2006; Broughton et al., 2005; Mark et al., 2012); indeed parents, particularly mothers, take extraordinary measures to protect their children from food deprivation (McIntyre et al., 2003). On the other hand, it is clear that household food insecurity is a household stressor that is very much felt by the children (Connell et al., 2005; Fram et al., 2011). In recognition of this mechanism, Whitaker et al. (2006) have proposed an early life stress hypothesis that would predict longer term mental health problems resulting from the childhood experience of food insecurity. They acknowledge, as we have also shown (McIntyre et al., 2003), that mothers buffer their children from the nutritional deprivation of household food insecurity, but may not be able to buffer them against its psychosocial stress (Connell et al., 2005). They suggest that food insecurity stress might act through brain neural circuits involving corticotropinreleasing factor (citing Coplan et al., 1996) and point out that there is a distinct biological subtype of adult depression that is known to be causally associated with such neural expression (citing Nemeroff and Vale, 2005).

128

L. McIntyre et al. / Journal of Affective Disorders 150 (2013) 123–129

Table 6 Hunger experience as a risk factor for low emotional quotientn, all inclusive and reduced logistic regression models. Model 1 all inclusive

Hunger Sex (female) Age at Cycle 8 Household income per $10,000 Household number Household mean ( 44.3) Education of PMK o High school graduate High school graduate Some post-secondary Completed trade/community college Bachelor/post-graduate degree Urban Housing tenure: always rented Ever lived without biological mother Ever lived with lone mother Depression of PMK n

Model 2 reduced

Odds ratio (95% CI)

P-value

Odds ratio(95% CI)

P-value

1.2(0.7–2.0) 0.6(0.5–0.9) 0.9(0.9–1.0) 1.0(0.9–1.0) 0.9(0.7–1.1) 1.6(1.0–2.5)

0.523 0.003 0.061 0.793 0.192 0.045

1.3(0.8–2.1) 0.7(0.5–0.9)

0.339 0.003

1.3(0.8–2.4) 0.7(0.4–1.2) 0.7(0.4–1.2) 0.7(0.4–1.2) Reference 1.0(0.7–1.3) 0.9(0.5–1.7) 0.7(0.5–1.0) 1.2(0.8–1.7) 1.1(0.6–1.9)

0.304 0.227 0.177 0.216 0.192 0.842 0.860 0.355 0.863

Note that outcome data collected from youth and young adults 14–17 and 20–21, 24–25 years, respectively.

This area of inquiry is extremely topical. For example, Del Giudice et al. (2011) have recently published a review of what they call the Adaptive Calibration Model, an evolutionary– developmental theory of individual differences in the functioning of the stress response system. Their model is built on the theory of biological sensitivity to context, which has been the joint work of Boyce and Ellis (2005) as well as Michael Meaney’s experimental work and discussion of epigenetics (Meaney, 2010). Epigenetic mechanisms causing long-lasting alterations in stress-responsiveness following exposure to early-life stressors have been described in animal models. In mice, for example, early life adversity has been shown to alter the epigenome in ways that reduce expression of a glucocorticoid receptor gene promoter in the hippocampus, thereby producing a longstanding diminishment of glucocorticoid-mediated negative feedback inhibition of stress responses (Meaney et al., 2007; Szyf et al., 2008; Weaver et al., 2004). Consistent with these data, adults with a history of childhood adversity show increased reactivity in stressful circumstances (see reviews by Taylor et al., 2004; Taylor, 2010). Such stress sensitization may be important to the etiology of adult mood disorders (Hammen et al.,2000) as well as in linking adversity to other adverse health outcomes such as diabetes, coronary heart disease, and osteoporosis (Stetler and Miller, 2011). These methylation-based epigenetic changes can also be passed on transgenerationally (at least in mice), so may explain intergenerational transmission of adverse mental health outcomes (Franklin et al., 2010). In effect, the results of our study indicate that child hunger may be one source of stress during childhood that leads to permanent vulnerabilities across the lifespan. The specificity of associations observed in this study (a positive association with depression/suicide ideation, but not with General Self-image or Emotional Quotient) enhances their interpretability as an outcome of epigenetic (or other stressresponse calibration mechanisms) rather than as a non-specific negative effect on mental health.

5. Limitations The major limitation of this study is that a single indicator, rather than a validated multi-item instrument (Health Canada, 2007) is used to assess child/youth hunger, which itself is a rare

extreme manifestation of food insecurity, thus constraining statistical power. However, inadequate power cannot explain the highly significant association reported. Further, the study relies on self-reported measures for hunger and mental health outcomes. The composition of the sample does not reflect immigration in Canada and did not include a covariate related to ethnicity in the modeling.

6. Conclusion Our study provides novel evidence in support of taking action on households with children who fall below a threshold of normatively acceptable disadvantage who are in priority need for income supports (Milligan and Stabile, 2008). We found that the experience of child/youth hunger is a modifiable risk factor for depression and related suicide ideation in late adolescence and early adulthood. Therefore, the need for prevention through the detection of such children and remedy of their circumstances should be ever more convincing.

Conflict of interest The authors declare no conflicts of interest.

Role of funding source This study was funded through a CIHR Operating Grant (Number: SEC-117126).

Acknowledgements None.

References Alaimo, K., Olson, C.M., Frongillo Jr, E.A., 2001. Food insufficiency and American school-aged children’s cognitive, academic, and psychosocial development. Pediatrics 108, 44–53. Alpert, J.E., Mischoulon, D., Nirenberg, A.A., Fava, M., 2000. Nutrition and depression: focus on folate. Nutrition 16, 544–581. Belsky, D.W., Moffitt, T.E., Arseneault, L., Melchior, M., Caspi, A., 2010. Context and sequelae of food insecurity in children’s development. American Journal of Epidemiology 172, 809–818. Boyce, W.T., Ellis, B.J., 2005. Biological sensitivity to context: I. An evolutionary– developmental theory of the origins and functions of stress reactivity. Developmental Psychopathology 17, 271–301.

L. McIntyre et al. / Journal of Affective Disorders 150 (2013) 123–129

Broughton, M., Janssen, P., Hertzman, C., Innis, S., Frankish, C., 2005. Predictors and outcomes of household food insecurity among inner city families with preschool children in vancouver. Canadian Journal of Public Health 97, 214–216. Casey, P., Goolsby, S., Berkowitz, C., Frank, D., Cook, J., Cutts, D., Heeren, T., 2004. Maternal depression, changing public assistance, food security, and child health status. Pediatrics 113, 298–304. Casey, P.H., Ettinger de Cuba, S.A., Cook, J.T., Frank, D.A., 2010. Child hunger, food insecurity, and social policy. Archives of Pediatrics and Adolescent Medicine 164, 774–775. Cassidy, J., Shaver, P.R., 2008. Handbook of Attachment: Theory, Research and Clinical Applications. Guilford Press, New York. Che, J., Chen, J., 2001. Food insecurity in Canadian households. Health Reports 12, 11–22. Connell, C.L., Lofton, K.L., Yadrick, K., Rehner, T.A., 2005. Children’s experiences of food insecurity can assist in understanding its effect on their well-being. Journal of Nutrition 135, 1683–1690. Coplan, D., Andrews, M.W., Rosenblum, L.A., Owens, M.J., Friedman, S., Gorman, J.M., Nemeroff, C.B., 1996. Persistent elevations of cerebrospinal fluid concentrations of corticotropin-releasing factor in adult nonhuman primates exposed to early-life stressors: implications for the pathophysiology of mood and anxiety disorders. Proceedings of the National Academy of Sciences of the United States of America 93, 1619–1623. Del Giudice, M., Ellis, B.J., Shirtcliff, E.A., 2011. The adaptive calibration model of stress responsivity. Neuroscience and Biobehavioral Reviews 35, 1562–1592. Fram, M.S., Frongillo, E.A., Jones, S.J., Williams, R.C., Burke, M.P., DeLoach, K.P., Blake, C.E., 2011. Children are aware of food insecurity and take responsibility for managing foodresources. Journal of Nutrition 141, 1114–1119. Franklin, T.B., Russig, H., Weiss, I.C., Graff, J., Linder, N., Michalon, A., Mansuy, I.M., 2010. Epigenetic transmission of the impact of early stress across generations. Biological Psychiatry 68, 408–415. Glanville, N.T., McIntyre, L., 2006. Diet quality of Atlantic families headed by single mothers. Canadian Journal of Dietetic Practice and Research 67, 28–35. Goodman, S.H., Rouse, M.H., Connell, A.M., Broth, M.R., Hall, C.M., Heyward, D., 2010. Maternal depression and child psychopathology: a meta-analytic review. Clinical Child and Family Psychology Review 14, 1–27. Gundersen, C., Kreider, B., 2009. Bounding the effects of food insecurity on children’s health outcomes. Journal of Health Economics 28, 971–983. Hammen, C., Henry, R., Daley, S.E., 2000. Depression and sensitization to stressors among young women as a function of childhood adversity. Journal of Consulting and Clinical Psychology 68, 782–787. Health Canada, 2007. Canadian Community Health Survey Cycle 2.2, Nutrition (2004) -Income-Related Household Food Security in Canada. Available from: /http://www.hc-sc.gc.ca/fn-an/surveill/nutrition/commun/income_food_secsec_alim-eng.phpS (accessed October 22, 2012). Kirkpatrick, S.I., McIntyre, L., Potestio, M.L., 2010. Child hunger and long-term adverse consequences for health. Archives of Pediatric and Adolescent Medicine 164, 754–762. Mark, S., Lambert, M., O’Loughlin, J., Gray-Donald, K., 2012. Household income, food insecurity and nutrition in Canadian youth. Canadian Journal of Public Health 103, 94–99. McIntyre, L., Connor, S.K., Warren, J., 2000. Child hunger in Canada: Results of the 1994, National longitudinal survey of children and youth. Canadian Medical Association Journal 163, 961–965. McIntyre, L., Walsh, G., Connor, S.K., 2001. A follow-up study of child hunger in Canada Report no.: W-01-1-2E. Available from: /http://publications.gc.ca/ collections/Collection/MP32-28-01-1-2E.pdfS (accessed 22.10.12). McIntyre, L., Glanville, N.T., Raine, K.D., Dayle, J.B., Anderson, B., Battaglia, N., 2003. Do low income lone mothers compromise their nutrition to feed their children? Canadian Medical Association Journal 168, 686–691.

129

Meaney, M.J., Szyf, M., Seckl, J.R., 2007. Epigenetic mechanisms of perinatal programming of hypothalamic-pituitary-adrenal function and health. Trends in Molecular Medicine 13, 269–277. Meaney, M.J., 2010. Epigenetics and the biological definition of gene  environment interactions. Child Development 81, 41–79. Melchior, M., Caspi, A., Howard, L.M., Ambler, A.P., Bolton, H., Mountain, N., Moffitt, T.E., 2009. Mental health context of food insecurity: a representative cohort of families with young children. Pediatrics 124, e564–e572. Milligan, K., Stabile, M., 2008. Do Child Tax Benefits Affect the Wellbeing of Children? Evidence from Canadian Child Benefit Expansions (Working Paper No. 14624). Available from:/http://www.nber.org/papers/w14624S (accessed October 22, 2012). Nemeroff, C.B., Vale, W.W., 2005. The neurobiology of depression: inroads to treatment and new drug discovery. Journal of Clinical Psychiatry 66 (7), 5–13. Nolen-Hoeksema, S., Larson, J., Grayson, C., 1999. Explaining the gender difference in depressive symptoms. Journal of Personality and Social Psychology 77, 1061–1072. Rainville, B., Brink, S., 2001. Food Insecurity in Canada, 1998–1999 (R-01-2E). Available from:/http://publications.gc.ca/collections/Collection/MP32-29-01-2E.pdfS (accessed October 22, 2012). Rangan, A.M., Blight, G.D., Binns, C.W., 1998. Iron status and non-specific symptoms of female students. Journal of the American College of Nutrition 17, 351–355. Rutter, M., 2003. Poverty and child mental health natural experiments and social causation. Journal of the American Medical Association 290, 2063–2064. Slopen, N., Fitzmaurice, G., Williams, D.R., Gilman, S.E., 2010. Poverty, food insecurity, and the behavior for childhood internalizing and externalizing disorders. Journal of the American Academy of Child and Adolescent Psychiatry 49, 444–452. Statistics Canada, 2010a. Household Food Insecurity, 2007–2008 (Report No.: 82– 625-X). Available from: /http://www.statcan.gc.ca/pub/82-625-x/2010001/ article/11162-eng.htmS (accessed October 22, 2012). Statistics Canada, 2010b. National Longitudinal Survey of Children and Youth (NLSCY), Cycle 8. Available from: /http://www23.statcan.gc.ca/imdb-bmdi/ document/4450_D4_T9_V8-eng.pdfS (accessed October 22, 2012). Stetler, C., Miller, G.E., 2011. Depression and hypothalamic-pituitary-adrenal activation: aquantitative summary of four decades of research. Psychosomatic Medicine 73, 114–126. Szyf, M., McGowan, P., Meaney, M.J., 2008. The social environment and the epigenome. Environmental and Molecular Mutagenesis 49, 46–60. Taylor, S.E., Lerner, J.S., Sage, R.M., Lehman, B.J., Seeman, T.E., 2004. Early environment, emotions, responses to stress, and health. Journal of Personality 72, 1365–1393. Taylor, S.E., 2010. Mechanisms linking early life stress to adult health outcomes. Proceedings of the National Academy of Sciences of the United States of America 107, 8507–8512. To, T., Guttmann, A., Dick, P.T., Rosenfield, J.D., Parkin, P.C., Tassoudji, M., Harris, J.K., 2004. Risk markers for poor developmental attainment in young children: results from a longitudinal national survey. Archives of Pediatrics and Adolescent Medicine 158, 643–649. Weaver, I.C., Cervoni, N., Champagne, F.A., D’Alessio, A.C., Sharma, S., Seckl, J.R., Meaney, M.J., 2004. Epigenetic programming by maternal behavior. Nature Neuroscience 7, 847–854. Weinreb, L., Wehler, C., Perloff, J., Scott, R., Hosmer, D., Sagor, L., Gundersen, C., 2002. Hunger: its impact on children’s health and mental health. Pediatrics 110, e41. Whitaker, R.C., Phillips, S.M., Orzol, S.M., 2006. Food insecurity and the risks of depression and anxiety in mothers and behavior problems in their preschoolaged children. Pediatrics 118, e859–e868. Wilhelm, K., Parker, G., Geerligs, L., Wedgewood, L., 2008. Women and depression: a 30 year learning curve. Australian and New Zealand Journal of Psychiatry (New South Wales) 42, 3–12.