Journal of Affective Disorders 202 (2016) 115–119

Contents lists available at ScienceDirect

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

Research paper

Breastfeeding and mental health in adulthood: A birth cohort study in
Brazil
Christian Loret de Mola a,b,n, Bernardo Lessa Horta a, Helen Gonçalves a,
Luciana de Avila Quevedo c, Ricardo Pinheiro c, Denise Petrucci Gigante a,
Janaína Vieira dos Santos Motta a,c, Fernando C. Barros a,c
a
Postgraduate Program in Epidemiology, Federal University of Pelotas, Brazil
b
Nursing Department, Federal University of Pelotas, Brazil
c
Health and Behavior Postgraduate Program, Universidade Católica de Pelotas – UCPEL, Pelotas, RS, Brazil

art ic l e i nf o a b s t r a c t

Article history: Background: Breastfeeding is negatively associate with behavioral and internalization problems, psy-
Received 13 February 2016 chological stress, and depressive/anxiety symptoms. However, studies evaluating specific mental health
Received in revised form disorders are scarce. We aimed to assess the association between breastfeeding and mental health
23 May 2016
outcomes in young adults.
Accepted 25 May 2016
Methods: In 1982, hospital deliveries in Pelotas (Southern Brazil) were identified; liveborns were ex-
Available online 26 May 2016
amined and their mothers interviewed (n¼5914). Information on breastfeeding was collected in early
Keywords: childhood. In 2012–13, at 30 years of age, we used the Mini International Neuropsychiatric Interview
Depression (MINI) for the diagnosis of major depression (MD), generalized anxiety disorder (GAD) and social anxiety
Mental health
disorder (SAD). In addition, we used the Beck Depression Inventory (BDI-II) and the Self-reported
Anxiety
Questionnaire (SRQ-20), to evaluate depressive symptoms severity and common mental disorders
Breastfeeding
Cohort (CMD), respectively. We used multivariable regression models to evaluate the association between
Brazil breastfeeding and mental health outcomes.
Results: We evaluated 3657 individuals. Prevalence of CMD, MD, GAD and SAD was 24.3%, 7.9%, 12.7% and
3.6%, respectively. In multivariable models the odds of having a more severe case of depression (BDI-II)
was smaller among those breastfed for 6 or more months (OR ¼0.69 95%CI [0.53–0.89]). We observed a
similar pattern for MD and CMD, however, confidence intervals included the reference.
Limitations: We had no information on home environment characteristics during childhood. Lack of
power and a small effect size could explain why we did not detect an association between breastfeeding
and MD.
Conclusion: Breastfeeding reduced the odds of having more severe depressive symptoms.
& 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).

1. Introduction In addition, breastfeeding is negatively associated with beha-
vioral and internalization problems (Hayatbakhsh et al., 2012;
Besides the short-term benefits of breastfeeding, evidence Heikkila et al., 2011; Liu et al., 2006, 2014; Yi et al., 2005) as well as
shows that been breastfed for a longer time has long-term con- psychological stress (Montgomery et al., 2006) and other mental
sequences. It has been reported that breastfeeding may reduce the health outcomes like depressive and anxiety symptoms (Hayat-
likelihood of obesity and type-2 diabetes. (Horta et al., 2015a; bakhsh et al., 2012; Oddy et al., 2010; Reynolds et al., 2014), major
Kelishadi and Farajian, 2014), increase performance in cognitive depression (Allen et al., 1998, Peus et al. 2012a) and attention
tests (Horta et al., 2015a) and it is associated with a higher income deficit disorder (ADHD) (Mimouni-Bloch et al., 2013; Stadler et al.,
in adulthood (Victora et al., 2015). 2015).
Merjonen et al. (2010) reported that breastfeeding might de-
crease the risk of depression possibly associated with the C/C
n
Corresponding author at: Postgraduate Program in Epidemiology, Federal Uni-
genotype of the estrogen receptor 1 gene. However, other studies
versity of Pelotas, Rua Marechal Deodoro, 1160-3° Piso, Pelotas, RS 96020-220,
Brazil . have failed to observe an association between breastfeeding and
E-mail address: chlmz@yahoo.com (C. Loret de Mola). later mental health (Anselmi et al., 2008; Kramer et al., 2008;

http://dx.doi.org/10.1016/j.jad.2016.05.055
0165-0327/& 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

1 MD 272 7. 2015) We tested the in- teraction between sex and breastfeeding over the evaluated out- comes. 2011). 3–5. sociated with socioeconomic status (Muntaner et al. Variable Na Mean (s.. 1986).5 M array (Illumina. in multivariable models. we tried to In the 2012–13 visit.d.7 ratios (PR) (Barros and Hirakata. assets index. we estimated the pro.0 and 6 months or more.9 months. (Supplementary Ta- presence of a major depressive episode during the last two weeks. moderate from 20 to 28 and schooling Mother’s age at birth (years) 3541 26.2 Multivariable models were adjusted for the following con- s. in. the pro- generalized anxiety disorder (GAD). USA). 3524. Those individuals whose BDI-II score was between 0 and 13 Female 1841 52 points were considered as minimal/no depression. the Human Genome Diversity Project (HGDP). We divided the time of breastfeeding in 3–5. breastfeeding. in Brazil. a southern Brazilian signed an informed consent. Sociodemographic characteristics at birth. Native Americans from the HGDP. anxiety disorder. (Barros et al. maternal a Number of individuals with data on at least one mental health outcome and marital status (married/living with partner or not married). breastfeeding. Birth weight (grams) 3541 3224 (526) Information on duration of breastfeeding was gathered in the Breastfeeding 1984 (mean age of 19 months) and 1986 (mean age of 42 months) o 1 month 753 21. genotyped using the Illumina Omni 2. all participants In 1982. less than 1 month. 1986). BDI ¼Beck depression inventory. respectively. those whose gical interviews using the Mini-international psychiatric interview mothers had between 5 and 8 years of schooling and whose family (MINI) V5. had information on CMD. at a mean age of 30.1 sion was defined by a score of 14-19. For each individual.. follow-up rate of 68. and 93 cluding depression (major depression and depression severity). when the 1982 Pelotas birth cohort started.)/Median Prevalence % and there was no evidence of a lifetime episode of mania/hypo- [IQR] mania according to the MINI.. Severe 179 5. 2014). 2013. maternal age at birth. about one in every three subjects had a order (GAD) in the last 6 months.0 four categories. 1996). 2000) and assessed the income ranged from 1 to 6 minimum wages. maternal smoking during pregnancy. social anxiety disorder (SAD) portion of European.. family and mental health have been done in high-income settings. validated for Brazil (Mari and Williams. (Kramer et al. number of Most of the studies on the association between infant feeding previous gestations.8 test was used to evaluate the proportional odds assumption. ble 1)Table 1 shows that 52% of the individuals included in the a lifetime episode of mania/hypomania.. where income at delivery. and nucleotide polymorphisms shared by samples from the HapMap breastfeeding was independent of socioeconomic status. and the et al. represented a research clinic to be interviewed and examined. 97% of which males with eight or more were considered as positive for CMD had information on duration of breastfeeding (n ¼3542). San 2011) However. father living in et al.¼ standard deviation. maternity hospitals in Pelotas. (Horta et al. Because mental health outcomes are negatively as. which follow the whole cohort and participants were invited to visit the added to the 325 subjects known to have died. and the earliest available data on age at weaning was used 1–2.7 to minimize recall bias. 83 African Americans.. The fol- (Mari and Williams. / Journal of Affective Disorders 202 (2016) 115–119 Kwok et al.2 years. generalized anxiety dis- analysis were females. Lind et al. 2003) of mental health outcomes BDI-II using the MINI and CMD. CMD ¼ Common mental dis- founding variables: sex. we used the Beck Depression In- Table 1 ventory (BDI-II) to evaluate the intensity of depressive symptoms.3 visits. CA. MD¼ Major depression. Males with a score of six or more and fe. we performed psycholo. For the BDI-II.9 Poisson regression with robust adjustment of the variance was GAD 438 12.0 (6. Methods We obtained ethical approval for the study from the ethics committee in the 'Universidade Federal de Pelotas’. 262 Europeans (American and Italian). African American.. cestry was estimated. Pelotas cohort.2) severe from 29 to 63 points (Beck et al. 77 admixed Mexican Americans. Admixture analyses were based on 37. The following HapMap samples were used as ex- This study was aimed at assessing the association between ternal panels: 266 Africans. and mental health outcomes in young adults. and 3445 individuals. Mild 375 10. In addition. low-up rate at 30 years was higher among females. if a major depressive episode was reported in the last two weeks.8 portional odds ratios (OR) using ordinal regression and the Brant Moderate 270 7.0 validated for Brazil (Amorim. In the 30 years visit. and social anxiety disorder (SAD) in the last month. mild depres.1%. health outcome was available for 3661 individuals. Data on at least one mental the presence of CMD.0539 single was little awareness of the benefits of breastfeeding. The 5914 liveborns whose families lived in the urban area of the city were examined and their mothers interviewed. genomic ancestry (percentage of African orders.. breastfeeding and mental health out- We considered an individual as having major depression (MD) comes at 30 years in individuals from the 1982 Pelotas birth cohort. GAD¼ Generalized anxiety disorder.9 months 909 25. to assess GAD. (SRQ-20). Z 6 months 1064 30. parental schooling (highest grade achieved at school). birth weight.. 3701 subjects were interviewed. 1–2.7 used in crude and multivariable analysis to estimate prevalence SAD 127 3. Mother’s with four or less years of 1135 32. IQR¼ Interquartile range. we used the self-reported questionnaire 3459. In addition. SAD¼ Social ancestry).116 C. were visited daily and all births were identified. CMD 853 24. 2008) From June 2012 to February 2013. Loret de Mola et al. 2011). there Diego. MD. (Brion Project. 5332 had information on breastfeeding. 3456.9 months. 2. Of the 5914 livebirths included in the 2015b). mother reference of nerve pro- breastfeeding duration is positively associated with income (Brion blems (proxy of mothers general mental health). . of these.. city. the same house and paternal history of psychiatric problems. cohort.d.9 months 816 23. 2004) most Genomic ancestry was estimated from DNA samples that were of these studies are subject to residual confounding. Results been followed in several occasions. ered from the hospital records). (Lima-Costa et al. and SAD at 30 years. These individuals have 3. and Native American an- and common mental disorders (CMD). type of delivery (information was gath.

89]).55) 0. type of delivery.79 (0.1) showed a smaller risk of disease. (Reynolds et al. 10. the confidence intervals barely included the reference in most of the analysis. measured using chi-squared tests for heterogeneity.98) Adjusteda PR (95%CI) p¼ 0.18 1 0. income at birth.63a p o 0..1) 493 (27.8) 93 (5. 6 or more months had a lower risk of major depression (PR¼ 0.02b 1 0. but found no association with anxiety. by the BDI-II.001 po 0.9) 179 (12.95) Adjusteda PR (95%CI) p¼ 0.77 (0.53  1.69 (0.13) 0.69 95%CI[0. SAD¼Social anxiety a longer duration of breastfeeding was associated with a lower risk disorder. and for this reason.84 [95%CI:0.001a po 0.23c 1 0.001a cluded the reference (Table 3). those who were breastfed for a longer time also Sex p¼ 0. The pattern of cording to the MINI was 7.99) 0.9) 76 (20.1) 34 (2.4) 119 (7.71 (0.1) 9–11 99 (25. fidence intervals included the reference. Liu et al.95]).6 to mother with four or less years of schooling at delivery.51  0.3) 320 (17. GAD ¼ Generalized anxiety disorder.17a p o 0.75  1.77 1.12) GAD Crude PR (95%CI) p¼ 0.7) 45 (11. Mental Disorder Breastfeeding p-Value o1 m 1–2.0) 19 (4. Among 1 month. ranging from 3. .70 1. however.2) any clear pattern of association and all confidence intervals in- Income tertiles p¼ 0. we believe that our findings suggest that had GAD. Loret de Mola et al. showed pression were higher. however.89) a Adjusted for sex.44c 1 0. previous gestations. family Studies evaluating the effect of breastfeeding over specific income and maternal schooling.9) 216 (18. and mental health was fect of breastfeeding over outcomes like general behavior (Heikkila associated with maternal income and schooling.7) 26 (2. moderate and severe depression was breastfeeding is associated with mental health in early adulthood. assets index.14b 1 0.35 1. smoking during pregnancy.3) lyses by sex.72 (0.0) 35 (7.90 (0.59  1.4) 108 (10.09) 0. Prevalence of major depression ac- specially depression or depressive symptomatology.72  1.71 0. magnitude of the associations slightly decreased and the con- acteristics at birth.2%. b Test for linear trend. However. However.20) 0.05) MD Crude PR (95%CI) p¼ 0.9%. the evidence had a lower prevalence of CMD (PR ¼0. 5.07–2. even in fully adjusted models (OR ¼1.0) 178 (16.83 (0. Concerning GAD and SAD. All other p-values were calculated less likely to have more severe symptoms of depression. we did not evidence Female 554 (30. the Breastfeeding and mental health variables according to sociodemographic char.7) 63 (13.91 (0.69 Hayatbakhsh et al. of minimum We did not evidence differences between the measures of as- wage sociations in males and females. 2006.64 similar results..97) 0.94 (0. breastfed for 6 months or more was independent of gender.0) 96 (7.001a p o0.18) 0. females had a mental disorders are scarce.87 (0.91 (0.57  1.61 (0.001a p o0.4) 337 (30. for most of the mental health outcomes.4) 178 (16.48  1.29) 0.06) SAD Crude PR (95%CI) p¼ 0. individuals who were breastfed for 6 months or more those few studies evaluating specific mental illness.70 0. 2011.89 (0. in a southern Brazilian city.66 (0.0) 29 (7. et al.2) 300 (24.72 (0.79 (0..1.78  1.53  1.54c 1 0. Therefore. (2012) found that breastfeeding for 4 or more Table 3 Crude and multivariate models evaluating the association between breastfeeding and mental health outcomes at 30 years in the Pelotas birth cohort of 1982.55  0.001a po 0.63  1.37) 0. Most studies show the protective ef- higher prevalence of mental disorders.11) 0. Maternal p¼ 0. birth weight.01c 1 0.2) 118 (7.2) 4. / Journal of Affective Disorders 202 (2016) 115–119 117 Table 2 95%CI[0. Allen et al.08) 0. In addition.9 m 3–5. GAD ¼ Generalized anxiety disorder.2) 70 (4.79 (0. all interactions terms had a 1 359 (32.91 (0. the study might have been underpowered. 2005) or mental well- Table 3 shows that compared to those breastfed for less than being. marital status. CMD ¼Common mental disorders.85 (0. maternal education.001 p o0.0) 47 (3.94) 0.44  1. On the other hand.51  0. those subjects who breastfed for [1. c Wald test p-value for heterogeneity. the confidence intervals in. association was similar for all depressive like outcomes. which was Table 2 shows that the proportion of subjects who were not the case for anxiety outcomes.98]). Z 12 179 (36. we cannot exclude that these associations were due to chance.64  1.1) 2 355 (28. father living together and father history of psychiatric illness. after controlling for confounders.9%.69  1. mother nerve problems.53  0.5) 68 (5.78 (0. maternal age.69 (0. of disease.81 (0.77 (0. SAD¼ Social anxiety disorder.52  1. Models evaluating major depression.4) 55 (5. Yi et al.40c 1 0.16) 0.7% (MINI diagnoses).92 (0. In crude analyses.17) 0.88 (0.51  0.02 (0.49  1.75 (0. The odds of having a Breastfeeding CMD MD GAD SAD more severe depressive symptoms (BDI-II) was smaller among at 6 months those breastfed for 6 or more months even after controlling for N (%) N (%) N (%) N (%) N (%) confounding variables (OR ¼0.001a po 0.71 1.84 (0. Because the pre- valence of most of the outcomes was small. C.08) 0.68  1. respectively.7) 104 (9.3) 387 (25. is not conclusive.9) 12 (3.27c 1 0.71  0.7% For this reason.94 p o 0.2) 46 (3. 7.9) 123 (10.20) Adjusteda PR (95%CI) p¼ 0.001a po 0.1) we found that individuals breastfed for more than 6 months were a p-values using chi-squared test for trend.0) 360 (21.15) BDI-II Crude OR (95%CI) p¼ 0.001 po 0.63  1. CMD ¼ Common mental disorders.51]).9) 137 (11. we did not stratified the ana- 3 350 (29.001 Male 510 (30.2) 57 (5.38  1.8%.2) 324 (28.79  1.2) In this cohort followed since birth.08c 1 0..0) 202 (11. and one of 12.11) 0.07) 0. Prevalence of mild.71  1.03) Adjusteda PR (95%CI) p¼ 0.13) 0.4) 10 (2. (1998) found that among those After controlling for confounders. every four were positive for common mental disorders and 12. MD ¼Major depression. BDI-II ¼ Beck depression inventory. ancestry.001a schooling in years 0–4 366 (32.38) 0. 2014) in children and adolescents. 2014. Discussion 5–8 418 (27.62  0.16) 1. subjects who were never breastfed the odds of having major de- cluded the reference.8) p-value 40.9 m Z6 m b CMD Crude PR (95%CI) p¼ 0. MD ¼Major depression.86 (0.93 (0.96) Adjusteda OR (95%CI) p¼ 0.53  0.

obesity. C. Rev. Med. A. C. of mothers misclassified breastfeeding duration measured in Kelishadi.. Messe. Horta.P. et al.C. 2011.J. In adults. Psiquiatr. Seeley.. breastfeeding on IQ. D. Furthermore. de Mola. References Some studies suggest that home environment and/or maternal care during childhood could influence the appearance of later Allen.. in this follow-up.J.L. explaining the association between breastfeeding and IQ. C. De Saude Publica rates for several baseline characteristics suggest that the present 42 (Suppl 2). and that breastfeeding is associated with 513–529. Rey.C..S. A. Farajian. A. Minten. G. TX. Matijasevich. 2008. Quigley. Psychopathol.L.C.U. Therefore. Acta Paediatr.. 1990). J. H.. N. Childhood IQ and adult mental disorders: a test of .. Najman. B.... 132. 2015a). adjusted estimated tend to be quite different in re- 1982 Pelotas birth cohort. M. Methodol. F.. Recall of childhood neglect and showed in a previous report that 244 individuals died during the physical abuse as differential predictors of current psychological functioning. 1996. C.. Bor. Dos Santos Motta.. de Mola. non-differential misclassification had a small effect in our Horta. However. spective study.. Kelly. F. D. them died by external causes. Anselmi. 44 441-441e. to mother-infant bonding or a com. L.A. Concerning selection bias.. Rev. associated with common mental disorders at 23 years. whereas in the presence of residual other hand. Martin. mortality due to suicide should be ex..A. eicosapentaenoic acid (EPA)..L. Oakley Browne risk for psychopathology in childhood and adolescence. Anselmi et al... Amorim.L.. B. Beria. In summary. Goncalves... 1995). The protective effects of breastfeeding on chronic 3-month categories. F. Beck. 3. in this de entrevista breve para diagnóstico de transtornos mentais.. / Journal of Affective Disorders 202 (2016) 115–119 months reduced anxiety/depression symptoms. Southern Brazil.. Loret de Mola et al. B. Long-term consequences of breast- feeding on cholesterol. Me- mortality in individuals with depression due to suicide. B. Barros.org/10.M. systolic blood pressure. (2008). fore we are unable to assess whether the protective effect of Prevalence and early determinants of common mental disorders in the 1982 birth cohort. 2012. components of breast milk.N. 22. in 2 weeks. Biomed. Methodology of the Pelotas with respect to breastfeeding duration. C. less than 3% (n ¼79). the observed associations are unlikely to be due included the reference in fully adjusted models. R..A.055. Breast feeding and child behaviour in the millennium cohort study. 2000. involved neighboring categories (Huttly et al. 3. C.L. What are the causal effects of 325 deaths that occurred. we consider that this systematic review and meta-analysis. We Gauthier. M. Negl. Epidemiol.. D. L.E. breastfeeding on mental health is attributable to the biological 26–33.G. L. Steer. (2012b) reported that the odds of being because in our cohort there was no clear social patterning of depressed were higher among those breastfed for less than breastfeeding according to socioeconomic status. (2013) found that breastfeeding had no effect over depressive Concerning residual confounding... P. Res. C. Do mothers and 1986 was 0. D... Breastfeed.2016. We should also mention that we did not ask for current psy.. Gigante. J. Furthermore. Caspi. Pelotas. 2008).: Off. Roberts... R.. 2015a. Hayatbakhsh. 10. In Psychological Corporation. Breastfeeding and intelligence: sys- treatment. tional studies: an empirical comparison of models that directly estimate the prevalence ratio. J.R. and 42 months for the remainder. Information on duration of breastfeeding was collected at 19 Horta. we believe that the observed symptoms in a birth cohort of 11 years-old children. Cohort profile update: the 1982 the weighted kappa comparing the information provided in 1984 Pelotas (Brazil) birth cohort study. and type-2 diabetes: psychiatrist during the last year. Vaughan. B. Adv. 40.. 2014. Hirakata. Int.R. 104.. Huttly. but this association was reduced and confidence intervals tions. Southern Brazil.J.. 2003. et al. Monitoring estimates.P. 1995. G.C... K. mortality in Pelotas birth cohort from 1982 to 2006.G. O’Callaghan. 7.. observed that breastfeeding was not lation to the crude ones. Harrington. tremely low and did not underestimate the association. Prenatal and perinatal influences on mental health disorders (Gauthier et al. Victora. Victora. 2014)..P.. 1996.L.. Renfrew. A. Peus et al. Anselmi. chological treatment in this follow-up. Smith.. Loret de Mola. the absence of differential follow-up Barros. present in human milk Supplementary data associated with this article can be found in and one of the possible components involved in the mechanisms the online version at http://dx.. we did not have information on home environment char.. Merjonen et al. W. 2015a. 30–37..C.. J. C. D. months for 96% of the sample.G. B. therefore some individuals 635–642. 572–575.05. Rev. Int. addition.. T. Epidemiol.jad..P.J. the negative results cannot be explained by a higher Brion. 20. Med. 14–19. 106–115.80.. 96. acteristics during childhood nor maternal-infant bonding.. Acad. visited a Horta. Horta. reason why we were not able to find an association shown in longitudinal studies that children with lower IQ have an with the MINI diagnoses. Among the nezes. Alternatives for logistic regression in cross-sec- bination of both..L.. obesity and blood pressure? Evidence from comparing high-income with middle-income cohorts. with negative symptoms could actually be positive cases in Horta. 1996. adjustment for several possible socioeconomic decreased the inherent risk of depression present in individuals and demographic confounders resulted in slight changes in the with the C/C genotype of the estrogen receptor 1 gene. L. of adolescents at misclassification tends to decrease the magnitude of the associa- age 14. C. Res. along with docosahexaenoic acid (DHA). Barros. results are unlikely to have been affected by selection bias.. Therefore. 480–486. Gigante. and 44 between 5 and 24 years. Barros. Barros. Child.. De Saude Publica 42 (Suppl 2).118 C. first 4 years of life. De study.G. C. of tematic review and meta-analysis. Similarly. Victora. M. B. 108–114. 2008. Am. On the measures of association. It has been might be small.. Rev. Epidemiol. K. R. Aronoff..T.. Koenen. BMC Med. 3. breastfeeding and mental health could be the relationship be. San Antonio.. 2015b. M. F. it is important to bear in mind that any non-differential Poulton. Kwok to recall bias...C. Victora. De Saude Publica 42 (Suppl 2). and only 19 of Child Abus. Arch. A. Victora. 670–680. 104.G. our findings suggest that breastfeeding reduces A possible pathway explaining the association between the odds of having more severe depressive symptoms. GK.. Dis. but in nearly all such cases misclassification non-communicable diseases in adulthood: a review of evidence. S. A. Similarly.. Araujo.. (2010) found that breastfeeding the present study. Gigante.P. Lawlor. F. In addition. 21. and the similar follow-up birth cohort study from 1982 to 2004–5. in a previous study using the confounding..G.. Heikkila. Bras. 2009). J. However. Candiota. and a recent meta-analysis have shown that breastfeeding is associated with a higher IQ (Horta et al. V. Breastfeed. those with no mental disorder according to the MINI. 1990.. Mini International Neuropsychiatric Interview (MINI): validação parent-child interaction qualities (Papp. mainly violence or accidents (Horta Association of breastfeeding and adolescents’ psychopathology: a large pro- et al. Barros.. Victora. J. L. no Appendix A. J.M.T. J. associations are not due to confounding by socioeconomic status. Sacker. Manual for the Beck Depression Inventory-II. G.. Nonetheless. S. J. L. I... C. H. 2009.. Supplementary material previous study have formally tested the mediating effect of IQ.1016/j.B..L. 7–15.. has been proposed as a diet supplement in the treatment of depression (Martins. J. the effect of breastfeeding over other mental health outcomes tween breastfeeding and cognitive development. A va- overestimate breast feeding duration? An example of recall bias from a study in lidation study done in a subsample of the cohort showed that 24% southern Brazil.G. Victora. 2011. Williams.. suggesting a high degree of agreement. Horta. Dev. 2009).O.R..L. increased risk of developing adult depression (Koenen et al.doi. Acta Paediatr. J. Kubzansky.M. Southern Brazil.. 549–559.L. 2008. Stollak. P.G. there.M. A. G.C. M.. 1998. Moffitt. Y.D. Victora.. Oliveira. However. Gigante... Lewinsohn. B. C.

. Zhang. 2014. 81. Ment. P. Papp... P. L. 1995.. 17. Gouveia.L. 2004. Redelin. 26. mother-child interaction qualities in early childhood. M. Arch.. 243–245.. R. L. Fombonne. Lehtimaki. E.. 50–57. Pediatrics 121. P. / Journal of Affective Disorders 202 (2016) 115–119 119 the cognitive reserve hypothesis. Williams. Dis. Tarazona-Santos.. Paul. Peus.. D... Perceptions of poor maternal care are associated with adolescent Contemp. D. Victora. S... B. Eaton.. Liu. Loret de Mola.. S96. Horta. Rey. Jacoby. C. Genomic ancestry and ethnoracial self-classification 40. 13 (3). Recalled in- Martins. Breast feeding and resilience against Bogdanovich. E. 2014. E. Epidemiol.T. Longitudinal associations between breastfeeding and observed Kehdy. J.J.... Oddy. Redelin. . 9. B. Clin. Perrine. E. J.. Breastfeeding and later psycho- hood: a retrospective analysis. Muntaner... 2010. J. Chin. K. Silburn. Br. G.I. Genet. Lancet Glob.. 2013.. B. Bogdanovich..J.S.... self-esteem and depression: Hong Kong’s ‘Children of 1997′ birth feeding on child and adolescent mental health: a pregnancy cohort study fol- cohort.. Long-term behavioural consequences of infant feeding: the limits of adult women. A validity study of a psychiatric screening questionnaire child mental well-being and protective against obesity at 9 years of age? Child: (SRQ-20) in primary care in the city of Sao Paulo. Affect.. A. hyperactivity disorder. Hennessy. birth cohort study from Brazil... Vanilovich. C. H.. 9812. Rodrigues.. Dahhou. N. Psychol. g. Association between breastfeeding and in- association between breastfeeding and child’s later adult depression: the role telligence.A. Psychosom. Re- N..K.G.. 148.. Salo. 2012a. 8. Child.E. Kwok. 882–890. X. Shannon. L. 437–448.. of breastfeeding intervention trial study. e435–e440. B. P.. Nigg. based on 5871 community-dwelling Brazilians (the Epigen Initiative). F. Gass. M. J. Nutrients 6.M. W. Shuper.. Child evidence from a meta-analysis of randomized controlled trials. F. F. Arch. 2015. Landau. X. observational studies. E... Am.. Sacker. Lima-Costa. I. L. L. Pediatr. Eur.J. Disord.. 34.. with gender and infant breastfeeding in preschool children. Hum. S36–S41... 525–542. Scharnholz... P. E. Breast feeding and early adolescent Zubrick. Dis. Yang. T. 2006.R. Breastfeeding may protect from developing attention deficit and habil. T. J. 91. Promotion psychosocial stress. J. 1986. J. 990–994. 1995. J. Ma... Yi. N... Tremblay.. M.. Epidemiol. Kendall. Association of breastfeeding with behavioral problems hood: A retrospective analysis. Disruptions in childhood parental care as risk factors for major depression in 2011. Perinat. P.. 189–190. 156. and temperament development in children aged 4–5 years. and income at 30 years of age: a prospective of estrogen receptor gene polymorphism. Care Health Dev. Lederbogen. 740–746.J. 2010.. Pereira. 2009. Aust. E. 277–278. S...D. Loret de Mola et al. Linder. Pediatrics 134 (Suppl 1). Kachevanskaya... 25..B. M. X. Rev. J. Breast-feeding in infancy and major depression in adult- Lind. Schooling. Keltikangas-Jarvinen. Miech. Scharnholz. Peus. M. 2012b.. Stanley. R.S. Quevedo.. Socioeconomic position clusive breastfeeding on child behavior and maternal adjustment: evidence and major mental disorders. J. N.. Ehlin. M. Jokela. M. Rodrigues-Soares. Fombonne. G... 5.N. Psychother. lowed for 14 years.. 500–506.M. 568–574. J.. R. Abnorm.. P. Pinheiro.M. R.D.C. V. P. F.T. Leung. Paul. from a large. 98. Liu.. Sci. Barros. randomized trial. L. F.F.L..G.P. O’Campo.H.C. S. No Pagination Specified 28. Matush. Psychiatry 29. T.B. 2005. J. Psychother. E. Kramer. A. Polek. Paediatr. R. Am. de Klerk..H.G. Is breastfeeding in infancy predictive of children’s internalizing behavior problems. R. Raveh. Deuschle. Mironova. C.. J. Victora.. 81. 76–89.. Montgomery. Li. P. e199–e205.. P. Pediatr.A. Mari.G... A.. Horta.R. 40. Yi.. P. 2014.M. F. M. S. Gigante. 2014.F.. 189–190. Twin Res... Sci. Platt. A.W.Z. D. Oakley Browne. A. Epigen-Brazil. M.. F. J... F. Rep.. E.W. L. 2013.. Joyce... Child.L. 887–894. C.. Zhang.. Deuschle. 95–100. The long-term effects of breast- behaviour. Merjonen. Nutr. A. 23–26.. Paediatr. Psychiatry 166. E. Bushnell. J.. J. M... 2006.W. Leung. B.. M. Stadler. Mimouni.. [Chinese]. Igumnov..M. Correlation of behavioral problems Mimouni-Bloch.. Child: Care Health Dev. The Goncalves. A.R. Musser. Breastfeeding and active bonding protects against Reynolds. J. social development of children at 6 years of age..E. C. EPA but not DHA appears to be responsible for the efficacy of itiation and duration of maternal breastfeeding among children with and omega-3 long chain polyunsaturated fatty acid supplementation in depression: without adhd in a well characterized case-control sample. L.. Psychosom. M. Robinson. 2008. Zhou. Schieve. Effects of prolonged and ex.J. Wells. M. A. Deuschle. Barreto. J.. W. Gass. L..R. Mambrini.. depression. 2015. V..G.S.A.E. J. C. Coll.. 334–337. Lederbogen.. Matush. M.. Holton. Health 3. N. J.. Chalmers. J. Chin. Li. Platt. J. Neurol. Kramer. C.. Deuschle. 2015.. Psychiatry: J. Hornblow.. 53–62. Breast-feeding infancy and major depression in adult- Liu. educational attainment....