laubinger thesis defense 2013-05-14x - state of · pdf fileootitis oo titis media, titis...

THE ASSOCIATION BETWEEN BREASTFEEDING

AND SELF-REPORTED POSTPARTUM

DEPRESSIVE SYMPTOMS IN OREGON MOTHERS

JEAN LAUBIN

GER

PRESENTED TO THE DEPARTM

ENT OF PUBLIC HEALTH AND

PREVENTIVE M

EDICINE

OREGON HEALTH AND SCIENCE UNIVERSITY

IN PARTIAL FULFILLM

ENT OF

THE REQUIREM

ENTS FOR THE DEGREE OF

MASTER OF PUBLIC HEALTH

INTRODUCTION

BREASTFEEDING

For mothers:

For mothers:

For mothers:

For mothers:

�Reduces

Reduces

Reduces

Reduces the risk of type 2 diabetes,

the risk of type 2 diabetes,

the risk of type 2 diabetes,

the risk of type 2 diabetes, rheumatoid

rheumatoid

rheumatoid

rheumatoid arthritis

arthritis

arthritis

arthritis, , , ,

cardiovascular

cardiovascular

cardiovascular

cardiovascular disease,

disease,

disease,

disease, breast

breast

breast

breast, and ovarian

, and ovarian

, and ovarian

, and ovarian cancer, weight loss

cancer, weight loss

cancer, weight loss

cancer, weight loss

For infants:

For infants:

For infants:

For infants:

�O OOOtitis

titis

titis

titis media,

media,

media,

media, upper

upper

upper

upper and lower respiratory tract infections,

and lower respiratory tract infections,



asthma

asthma

asthma

asthma, RSV,

, RSV,

, RSV,

, RSV, necrotizing

necrotizing

necrotizing

necrotizing entercolitis,

entercolitis,

entercolitis,

entercolitis, atopic

atopic

atopic

atopic derm

atitis,

derm

atitis,

derm

atitis,

derm

atitis,

gastroenteritis

gastroenteritis

gastroenteritis

gastroenteritis, , , , obesity

obesity

obesity

obesity, , , , celiac

celiac

celiac

celiac disease,

disease,

disease,

disease, diabetes

diabetes

diabetes

diabetes, , , , leukemia

leukemia

leukemia

leukemia and

and

and

and

sudden infant death syndrome




Categorizations: Exclusive vs Any vs None; Duration




The

The

The

The Oregon 2011 breastfeeding report

Oregon 2011 breastfeeding report

Oregon 2011 breastfeeding report

Oregon 2011 breastfeeding report card: 91.2

card: 91.2

card: 91.2

card: 91.2% of mothers

% of mothers

% of mothers

% of mothers

initiated breastfeeding; 62.5



initiated breastfeeding; 62.5% breastfeed at least 6 months.

% breastfeed at least 6 months.

% breastfeed at least 6 months.

% breastfeed at least 6 months. Background

POSTPARTUM DEPRESSION

Depression occurring in the first year after birth




diagnosed using questionnaire and clinical




consultation

consultation

consultation

consultation

Approximately 10

Approximately 10

Approximately 10

Approximately 10- ---15% of mothers will experience PPD

15% of mothers will experience PPD



Sym

ptoms include

Sym

ptoms include

Sym

ptoms include

Sym

ptoms include sadness, lack of energy, trouble

sadness, lack of energy, trouble



concentrating, anxiety, and feelings of guilt or




worthlessness, which can inhibit



worthlessness, which can inhibit maternal bonding

maternal bonding

maternal bonding

maternal bonding

In children PPD associated with inhibited language




development, emotional and behavioral problems




Background

SPECIFIC AIMS

SA1: To

SA1: To

SA1: To

SA1: To describe the prevalence of breastfeeding at least 8 weeks and self

describe the prevalence of breastfeeding at least 8 weeks and self

describe the prevalence of breastfeeding at least 8 weeks and self

describe the prevalence of breastfeeding at least 8 weeks and self- ---reported

reported

reported

reported

postpartum depressive sym

ptoms in the 2007 sample of Oregon mothers with







live births.

live births.

live births.

live births.

SA2: To

SA2: To

SA2: To

SA2: To measure the association between breastfeeding and self

measure the association between breastfeeding and self

measure the association between breastfeeding and self

measure the association between breastfeeding and self- ---reported

reported

reported

reported


ptoms while adjusting for potential confounders and







to examine the potential presence of effect modification using logistic regression.




�Examine additional risk factors that may influence the primary relationship

between breastfeeding and self-reported postpartum depressive sym

ptoms

including income, race/ ethnicity, education, and marital status, parity,

maternal age, insurance, sm

oking status, maternal body mass index, and

stressful life events.

SA3: To

SA3: To

SA3: To

SA3: To determ

ine if directionality and causality can be determ

ined using these

determ


ined using these

determ


ined using these

determ


ined using these

results combined with current literature on the association between




breastfeeding and self



breastfeeding and self- ---reported postpartum depressive sym

ptoms.

reported postpartum depressive sym

ptoms.


ptoms.


ptoms.

METHODS

PREGNANCY RISK ASSESSMENT MONITORING

SYSTEM

2007 PRAMS survey data




PRAMS is an epidemiologic survey



PRAMS is an epidemiologic survey used to evaluate

used to evaluate

used to evaluate

used to evaluate

maternal health behaviors with the goal of improving




the health outcomes of mother and infants.




Weighting: oversampling of underrepresented racial/




ethnic populations, low birth weight infants and non



ethnic populations, low birth weight infants and non- ---

respondents.

respondents.

respondents.

respondents.

S SSSampled 3067 mothers with 1894 completing the

ampled 3067 mothers with 1894 completing the



survey: a weighted response rate of 67%




Methods

PRIMARY DEPENDENT VARIABLE

Self

Self

Self

Self- ---Reported Postpartum Depressive Sym

ptoms

Reported Postpartum Depressive Sym

ptoms


ptoms


ptoms

PRAMS Questions:

PRAMS Questions:

PRAMS Questions:

PRAMS Questions:

1.

1.

1.

1. “Since your new baby was born, how often have you felt down, depressed, or

“Since your new baby was born, how often have you felt down, depressed, or



hopeless?”

hopeless?”

hopeless?”

hopeless?”

2.

2.

2.

2. “Since your new baby was born, how often have you had little interest or little

“Since your new baby was born, how often have you had little interest or little



pleasure in doing things?”




Possible Responses: Always, Often, Sometimes, Rarely, Never




Classification:

Classification:

Classification:

Classification:

Present: Response of Always or Often to either question




Absent: Response of Sometimes, Often, Never to both questions




Missing: No response or response of Sometimes, Often, Never to one question




and no response on other




Methods

PRIMARY DEPENDENT VARIABLE

Self

Self

Self

Self- ---reported postpartum depressive sym

ptoms (SRPPDS):


ptoms (SRPPDS):


ptoms (SRPPDS):


ptoms (SRPPDS):

Subset: 9.83% reported postpartum depressive sym

ptoms


ptoms


ptoms


ptoms

Methods

Question

Question

Question

Question

“Since your new baby was born, how often




have you felt down, depressed, or hopeless?”




Present

Absent

Missing

Total

“Since your

new baby

was born,

how often

have you had

little interest

or little

pleasure in

doing

things?”

Present

101

94

0174

Absent

73

1586

91682

Missing

02

29

38

Total

195

1668

31

1894

SRPPDS Total

SRPPDS Total

SRPPDS Total

SRPPDS Total

Combined

Overall

Present

Absent

Missing

Total

268

1586

40

1894

PRIMARY INDEPENDENT VARIABLE

Any breastfeeding at least 8 weeks




77% of the sample population breastfed at least 8 weeks




Methods

Qu

es

tio

nY

es

No

Mis

sin

gT

ota

l

Is y

ou

r b

ab

y a

live

no

w?

18

08

26

60

18

94

Is y

ou

r b

ab

y l

ivin

g w

ith

yo

ur

no

w?

18

00

76

11

86

8

Did

yo

u e

ver

bre

ast

fee

d o

r p

um

p b

rea

st

mil

k t

o f

ee

d y

ou

r n

ew

ba

by

aft

er

de

live

ry?

16

61

13

36

18

00

Are

yo

u s

till

bre

ast

fee

din

g o

r fe

ed

ing

p

um

pe

d m

ilk

to

yo

ur

ne

w b

ab

y?

10

96

55

51

01

66

1

Ho

w m

an

y w

ee

ks

or

mo

nth

s d

id y

ou

b

rea

stfe

ed

or

pu

mp

mil

k t

o f

ee

d y

ou

r b

ab

y?

< 1

we

ek

>

= 1

we

ek

Mis

sin

g

47

49

81

0

To

tal

va

lid

fo

r a

na

lys

is

Mis

sin

g/

Un

kn

ow

n

17

74

12

0

INCLUSION/ EXCLUSION

3067 Mother sampled in 2007

1894 Completed surveys

1774 completed breastfeeding

questions

1854 completed self-reported

postpartum depressive

symptoms questions

135 were excluded for not having

breastfeeding or self-reported


ptum

status

1759 Had inform

ation on both

breastfeeding and Self-reported


ptoms

and were included for analysis

120 were missing or unknown for

breastfeeding status

40 were missing or unknown for

self-reported postpartum

depressive sym

ptoms

Methods

CONFOUNDERS

De

mo

gra

ph

ic V

ari

ab

les

Po

ssib

le R

esp

on

ses

Co

din

g f

or

An

aly

sis

Ma

tern

al

Ag

e (

BC

)C

on

tin

uo

us

0=

>=

25

1=

<2

5

Ma

rita

l S

tatu

s (B

C)

Ma

rrie

d, n

ot

ma

rrie

d0

=M

arr

ied

1=

No

t M

arr

ied

Ma

tern

al

race

/e

thn

icit

y (

BC

) (E

M)

His

pa

nic

, NH

Wh

ite

, NH

Afr

ica

n A

me

rica

n,

NH

Asi

an

/P

aci

fic

Isla

nd

er,

NH

Am

eri

can

In

dia

n/

Ala

ska

n N

ati

ve

0=

No

n H

isp

an

ic W

hit

e, 1

=H

isp

an

ic,

2=

No

n H

isp

an

ic A

fric

an

Am

eri

can

, 3

=N

on

His

pa

nic

Asi

an

/P

aci

fic

Isla

nd

er,

4

= N

on

His

pa

nic

Am

eri

can

In

dia

n/

Ala

ska

n

Na

tiv

eM

ate

rna

l E

du

cati

on

(B

C)

Le

ss t

ha

n 1

2th

gra

de

, 12

thg

rad

e/

GE

D,

Gre

ate

r th

an

12

thg

rad

e0

= >

= 1

2th

gra

de

1=

12

thg

rad

e/

GE

D2

=<

12

thg

rad

eo

r0

= >

12

thg

rad

e1

= <

=1

2th

gra

de

Methods

SOME MORE CONFOUNDERS

PR

AM

S M

ea

sure

sP

oss

ible

Re

spo

nse

sC

od

ing

fo

r A

na

lysi

s

Pre

gn

an

cy I

nte

nti

on

(P

R)

I w

an

ted

to

be

pre

gn

an

t so

on

er,

I w

an

ted

to

b

e p

reg

na

nt

late

r, I

wa

nte

d t

o b

e p

reg

na

nt

the

n, I

did

n’t

wa

nt

to b

e p

reg

na

nt

the

n o

r a

ny

tim

e i

n t

he

fu

ture

0=

No

if r

esp

on

se i

s “I

wa

nte

d t

o b

e p

reg

na

nt

late

r” o

r “I

did

n’t

wa

nt

to b

e p

reg

na

nt

the

n o

r a

ny

tim

e i

n t

he

fu

ture

”1

=Y

es

if r

esp

on

se i

s “I

wa

nte

d t

o b

e p

reg

na

nt

soo

ne

r” o

r “I

wa

nte

d t

o

be

pre

gn

an

t th

en

”

Fa

mil

y I

nco

me

(P

R)

Le

ss t

ha

n $

10

,00

0, $

10

,00

0-$

14

,99

9,

$1

5,0

00

-19

,99

9, $

20

,00

0-2

4,9

99

, $2

5,0

00

-$

34

,99

9, $

35

,00

0-$

49

,99

9, $

50

,00

0 o

r m

ore

0=

0%

-2

00

% F

ed

era

l P

ov

ert

y L

ev

el

1=

20

0%

Fe

de

ral

Po

ve

rty

Le

ve

l o

r g

rea

ter

Insu

ran

ce b

efo

re p

reg

na

ncy

(P

R)

No

Ye

s0

= Y

es

1=

No

Pa

rity

(P

R)

Fir

st c

hil

d, s

eco

nd

ch

ild

, th

ird

ch

ild

, fo

urt

h

chil

d, f

ifth

ch

ild

or

mo

re0

=F

irst

ch

ild

1=

Se

con

d c

hil

d2

= T

hir

d o

r m

ore

Sm

ok

ing

No

w (

PR

)4

1 o

r m

ore

cig

are

tte

s2

1-4

0 c

iga

rett

es

11

-20

cig

are

tte

s6

-10

cig

are

tte

s1

-5 c

iga

rett

es

Le

ss t

ha

n 1

cig

are

tte

0 c

iga

rett

es

0=

0 c

iga

rett

es

1=

1 o

r m

ore

cig

are

tte

s

Bo

dy

Ma

ss I

nd

ex

(PR

)A

ra

tio

of

we

igh

t b

efo

re p

reg

na

ncy

an

d

he

igh

t w

ith

ou

t sh

oe

s0

=<

25

BM

I1

=2

5 o

r g

rea

ter

BM

I

Str

ess

ful

Lif

e E

ve

nts

(P

R)

(EM

)N

oY

es

0=

0 E

ve

nts

1=

1-2

Ev

en

ts

1=

3 o

r m

ore

Ev

en

ts

Methods

LOGISTIC REGRESSION ANALYSIS

Bivariate analysis used to determ

ine significant


ine significant


ine significant


ine significant

confounders for multivariate logistic regression




analysis and ensure correct variable categorization




Multiple logistic regression considered significant




variables from bivariate analysis for inclusion




Forward stepwise model building confirm

ed with


ed with


ed with


ed with

backward stepwise elim

ination


ination


ination


ination

Effect modification was considered via interaction term




with breastfeeding. “



with breastfeeding. “Testparm

Testparm

Testparm

Testparm

” was used to

” was used to

” was used to

” was used to

determ

ine significance of the term

. determ


. determ


. determ


.

Methods

RESULTS

BIVARIATE ANALYSIS

Va

ria

ble

N (

tota

l 1

75

9)

Un

we

igh

ted

%

Re

po

rtin

g

De

pre

ssio

n

We

igh

ted

%

Re

po

rtin

g

De

pre

ssio

n

We

igh

ted

Un

iva

ria

te O

R

95

% C

Ip

-va

lue

An

y b

rea

stfe

ed

ing

At

lea

st 8

we

ek

s

Ye

sN

oM

issi

ng

12

71

48

80

11

.17

19

.47

7.9

16

.2R

efe

ren

t2

.24

51

.37

, 3.6

70

.00

3

Pre

gn

an

cy

In

ten

tio

n

Ye

sN

oM

issi

ng

10

42

49

72

0

11

.42

16

.79

8.2

71

2.4

9R

efe

ren

t1

.58

1.2

5, 2

.01

0.0

01

Fa

mil

y i

nc

om

e (

%F

PL

)

<2

00

%>

=2

00

%M

issi

ng

95

76

81

12

7

16

.72

7.4

91

3.6

75

.34

2.8

1R

efe

ren

t1

.97

, 3.9

9<

0.0

01

Ma

rita

l S

tatu

s

Ma

rrie

dN

ot

Ma

rrie

dM

issi

ng

10

70

68

90

10

.93

17

.42

8.2

21

3.2

8R

efe

ren

t1

.71

1.0

1, 2

.89

0.0

46

Ma

tern

al

Ag

e

<2

5>

=2

5M

issi

ng

59

01

16

90

18

.98

10

.69

15

.91

7.0

82

.48

Re

fere

nt

1.6

2, 3

.80

<0

.00

1

Ma

tern

al

Ed

uc

ati

on

<=

Hig

h S

cho

ol

Dip

lom

a/

GE

D>

Hig

h S

cho

ol

Dip

lom

a/

GE

DM

issi

ng

94

7

80

2

10

16

.37

9.9

8

12

.56

7.1

Re

fere

nt

1.8

79

1.0

9, 3

.22

0.0

25

Results

BIVARIATE ANALYSIS

Va

ria

ble

N (

tota

l 1

75

9)

Un

we

igh

ted

%

Re

po

rtin

g

De

pre

ssio

n

We

igh

ted

%

Re

po

rtin

g

De

pre

ssio

n

We

igh

ted

Un

iva

ria

te O

R

95

% C

Ip

-va

lue

Ma

tern

al

Ra

ce

His

pa

nic

NH

Wh

ite

NH

AP

IN

H A

fric

an

A

me

rica

nN

H A

I/A

NM

issi

ng

41

55

88

29

32

03

25

82

11

.81

12

.24

10

.58

20

.69

16

.28

11

.87

8.6

11

1.5

22

0.2

4

16

.55

1.4

3R

efe

ren

t1

.38

2.7

2

2.1

0

0.8

45

, 2.4

2

0.4

9, 3

.92

1.5

9, 4

.67

1.1

6, 3

.82

0.1

71

0.5

23

0.0

01

0.0

18

Insu

ran

ce

Be

fore

Pre

gn

an

cy

Ye

sN

oM

issi

ng

95

17

96

6

10

.45

17

.09

8.2

21

2.2

2R

efe

ren

t1

.55

0.5

5, 4

.42

0.3

87

Pa

rity

1st

2n

do

r m

ore

Mis

sin

g

73

91

01

80

12

.58

14

.15

8.9

91

0.4

Re

fere

nt

1.1

70

.64

, 2.1

70

.58

8

Sm

ok

ing

No

w

Ye

sN

oM

issi

ng

23

81

49

42

7

23

.11

11

.78

16

.63

8.9

12

.04

Re

fere

nt

1.5

5, 2

.69

<0

.00

1

Bo

dy

Ma

ss I

nd

ex

No

rma

l <2

5A

ll o

the

r>=

25

Mis

sin

g

87

37

43

14

6

11

.45

15

.61

6.9

11

2.4

2R

efe

ren

t1

.91

1.4

6, 2

.48

<0

.00

1

Str

ess

ful

Lif

e E

ve

nts

01

-23

or

mo

re

52

96

96

52

1

6.6

21

0.7

82

3.6

1

6.0

59

.18

15

.12

Re

fere

nt

1.5

72

.77

0.7

2, 3

.43

0.6

8, 1

1.3

10

.24

10

.14

6

Results

FORWARD STEPWISE MODEL BUILDING

Ch

ara

cte

rist

ic*

Sta

ge

1 O

R

(T-s

tat)

P-v

alu

e

Sta

ge

2 O

R

(T-s

tat)

P-v

alu

e

Sta

ge

3 O

R

(T-s

tat)

P-v

alu

e

Sta

ge

4 O

R

(T-s

tat)

P-v

alu

e

Sta

ge

5 O

R

(T-s

tat)

P-v

alu

e

Fa

mil

y i

nc

om

e

<2

00

% F

PL

20

0%

FP

L o

r

mo

re

2.8

1 (

6.1

5)

<0

.00

1Referent

2.6

6 (

5.9

7)

<0

.00

1

Referent

2.4

6 (

3.8

0)

0.0

01

Referent

1.9

3 (

2.4

6)

0.0

24

Referent

1.9

6 (

2.2

3)

0.0

39

Referent

Sm

ok

ing

No

w

Ye

s

No

1.7

3 (

5.8

2)

<0

.00

1Referent

1.6

9 (

6.9

8)

<0

.00

1Referent

1.4

4 (

4.0

6)

0.0

01

Referent

1.4

3 (

4.4

3)

<0

.00

1

Referent

BM

I BM

I <

25

(R

ef)

BM

I >

=2

5

Referent

1.8

0 (

4.0

6)

0.0

01

Referent

1.9

9 (

5.1

2)

<0

.00

1

Referent

2.0

4 (

5.2

3)

<0

.00

1

Ma

tern

al

ag

e

<2

5 y

ea

rs

25

ye

ars

or

old

er

2.1

5 (

2.9

6)

0.0

08

Referent

2.1

9 (

3.1

6)

0.0

05

Referent

Ma

rita

l sta

tus

Ma

rrie

d

No

t M

arr

ied

No

t si

gn

ific

ant

Referent

-0.1

7 (

0.8

70

)M

ate

rna

l e

du

cati

on

<

=1

2th

/G

ED

>1

2th

/ G

ED

No

t si

gn

ific

ant

-1.7

5 (

0.0

98

)Referent

Results

FORWARD STEPWISE MODEL BUILDING

Ch

ara

cte

rist

ic*

Sta

ge

1 O

R

(T-s

tat)

P-v

alu

e

Sta

ge

2 O

R

(T-s

tat)

P-v

alu

e

Sta

ge

3 O

R

(T-s

tat)

P-v

alu

e

Sta

ge

4 O

R

(T-s

tat)

P-v

alu

e

Sta

ge

5 O

R

(T-s

tat)

P-v

alu

e

Pre

gn

an

cy

Inte

nti

on

Ye

s N

o

No

t si

gn

ific

ant

Referent

-0.2

3 (

0.8

23

)M

ate

rna

l ra

ce/

E

thn

icit

yH

isp

an

icN

H W

hit

eN

H A

PI

NH

Afr

ica

n

Am

eri

can

NH

AI/

AN

No

t si

gn

ific

ant

-0.3

1 (

0.7

64

)Referent

2.1

1 (

0.0

49

)1

.61

(0

.12

4)

1.1

4 (

0.2

70

)In

sura

nce

be

fore

p

reg

na

ncy

Ye

s N

o

No

t si

gn

ific

ant

Referent

0.7

8 (

0.4

45

)P

ari

ty1

stC

hil

d2

nd

or

mo

re

No

t si

gn

ific

ant

Re

fere

nt

1.6

4 (

0.1

17

)S

tre

ssfu

l lif

e

ev

en

tsO

Ev

en

ts

1-2

Ev

en

ts3

+ E

ve

nts

No

t si

gn

ific

ant

Referent

0.5

8(0

.57

1)

1.4

0(0

.18

0)

Results

DISCUSSION

FINAL MODEL

T TTThe

he

he

he odds of self

odds of self

odds of self

odds of self- ---reported postpartum depressive

reported postpartum depressive



sym

ptoms in women who breastfed less than eight

sym


sym


sym


weeks was 1.60 times the odds of self



weeks was 1.60 times the odds of self- ---reported

reported

reported

reported


ptoms in women who


ptoms in women who


ptoms in women who


ptoms in women who

breastfed at least eight weeks (95% CI: 1.004, 2.57)




�Confounders: Maternal age, Family income,

Current sm

oking status, Body Mass Index,

�Effect Modifier: Stressful life events

No Stressful life events: OR: 4.81



No Stressful life events: OR: 4.81 (95%: 2.28

(95%: 2.28

(95%: 2.28

(95%: 2.28- ---10.17

10.17

10.17

10.17) )))

Discussion

STRENGTHS AND LIMITATIONS

Strengths:

Strengths:

Strengths:

Strengths:

�Large stratified random sample

Large stratified random sample



�W WWWeighting to assess often overlooked populations

eighting to assess often overlooked populations



�Variety of behavioral and demographic variables

Variety of behavioral and demographic variables



Limitations:

Limitations:

Limitations:

Limitations:

�Cross

Cross

Cross

Cross- ---sectional PRAMS data

sectional PRAMS data

sectional PRAMS data

sectional PRAMS data- ---cannot assess directionality

cannot assess directionality



�Response rate of 67% may have biased results

Response rate of 67% may have biased results



�Recall bias of socially undesirable traits possible

Recall bias of socially undesirable traits possible



�Limited definition of main variables

Limited definition of main variables



CAUSALITY

HILL’S CRITERIA FOR CAUSE

1. Strength of the Association




2. Specificity of the Association




3.

3.

3.

3. Consistency

Consistency

Consistency

Consistency of the Association

of the Association

of the Association

of the Association

4. Biological

4. Biological

4. Biological

4. Biological Plausibility

Plausibility

Plausibility

Plausibility

5.

5.

5.

5. Temporal

Temporal

Temporal

Temporal Relationship

Relationship

Relationship

Relationship

6. Dose Response effect




7. Intervention effect




Causality

KNOWN ASSOCIATION

Bi

Bi

Bi

Bi- ---directional association:

directional association:



�Breastfeeding linked to multiple measures of

depressive sym

ptoms

�Depressive sym

ptoms linked to breastfeeding cessation

Common confounders

Common confounders

Common confounders

Common confounders: :::

�Maternal age, Education, Race/ Ethnicity, Parity,

Socioeconomic status, Marital status, Income, Body

Mass Index, Pregnancy Intention, Smoking

Causality

SRPPDS OUTCOMES

1st

Au

tho

r, Y

ea

rS

am

ple

Siz

eM

eth

od

Dir

ec

tio

nM

ea

sure

of

Ass

oc

iati

on

Tim

e P

oin

t o

f

Ass

oc

iati

on

Stu

dy

co

ntr

ibu

tio

n

Po

stp

art

um

De

pre

ssiv

e S

ym

pto

m O

utc

om

es

Wa

rne

r 1

99

62

37

5C

ross

-se

ctio

na

lN

ot

BF

->

EP

DS

>

12

OR

: 1

.52

(1

.12

-2

.06

)6

we

ek

sM

ag

nit

ud

e o

f a

sso

cia

tio

n

Ch

au

dro

n 2

00

14

65

Lo

ng

itu

din

al

Wo

rry

ab

ou

t B

F -

>

Dia

gn

ose

d

de

pre

ssio

n

(DS

M/

tre

atm

en

t)

Re

lati

ve

Ris

k:

3.0

(1

.04

-9.2

2)

4 w

ee

ks

Ma

gn

itu

de

of

ass

oci

ati

on

Yo

nk

ers

20

01

29

7C

ross

-se

ctio

na

lA

ny

BF

->

ED

PS

>1

2O

R:

0.6

0 (

0.4

4-

0.8

1)

3 w

ee

ks

Ma

gn

itu

de

of

ass

oci

ati

on

Me

zza

cap

pa

20

02

24

Lo

ng

itu

din

al

BF

vs

no

t B

F->

Ne

gM

oo

dA

NO

VA

: t(

21

)=

2.3

0;

p<

0.0

5D

uri

ng

b

rea

stfe

ed

ing

se

ssio

n1

-7 m

on

ths

Dif

fere

nce

in

E

xpo

sure

gro

up

s

Gro

ër

20

05

18

3C

ross

-se

ctio

na

lE

xclu

siv

e B

F v

s n

ot

BF

->

Mo

od

Sco

res,

S

tre

ss s

core

s

AN

OV

A t

=2

.89

p

=0

.00

44

-6 w

ee

ks

Dif

fere

nce

in

E

xpo

sure

gro

up

s

Ha

tto

n 2

00

51

85

Lo

ng

itu

din

al

An

y B

F v

s n

ot

BF

-> E

PD

S ≥

14

AN

OV

A p

<0

.05

6 w

ee

ks

Dif

fere

nce

in

E

xpo

sure

gro

up

s

Wa

tkin

s 2

01

12

18

5L

on

git

ud

ina

lD

isli

ke

BF

->

EP

DS

>1

2O

R=

1.4

2 (

1.0

4-

1.9

3)

8 w

ee

ks

Ma

gn

itu

de

of

ass

oci

ati

on

Ta

sha

ko

ri 2

01

21

50

Ca

se-c

on

tro

lN

ot

BF

vs

BF

->

EP

DS

>1

2X

2(t

=-2

.9,d

f=1

48

);

p=

0.0

04

8 w

ee

ks

Dif

fere

nce

in

E

xpo

sure

gro

up

s

Yst

rom

20

12

42

,22

5L

on

git

ud

ina

lB

F o

r M

ixe

d B

F

Fo

rm->

An

x &

De

p

Sy

mp

tom

s

β=

0.0

8(0

.05

-0.1

1)

(Mix

ed

) β

=0

.24

(0.2

1-0

.29

) (F

orm

)

6 m

on

ths

Dif

fere

nce

in

a

mo

un

t o

f B

F

Zu

ba

ran

20

12

89

Cro

ss-s

ect

ion

al

BF

se

lf-e

ffic

acy

->

EP

DS

>1

2,

R2=

0.1

25

; F

(1,8

7)-

12

.43

p=

0.0

01

2-1

2 w

ee

ks

Do

se-r

esp

on

se

ass

oci

ati

on

Causality

BREASTFEEDING OUTCOMES

1st

Au

tho

r,

Ye

ar

Sa

mp

le S

ize

Me

tho

dD

ire

cti

on

Me

asu

re o

f

Ass

oc

iati

on

Tim

e P

oin

t o

f

Ass

oc

iati

on

Stu

dy

co

ntr

ibu

tio

n

Bre

ast

fee

din

g O

utc

om

es

He

nd

ers

on

20

03

14

10

Lo

ng

itu

din

al

EP

DS

>1

2 -

> B

F

Ce

ssa

tio

nH

R=

1.2

5 (

1.0

3-

1.5

2)

Pri

or

to 8

we

ek

sM

ag

nit

ud

e o

f a

sso

cia

tio

n

De

nn

is 2

00

75

94

Lo

ng

itu

din

al

EP

DS

>1

2 -

>B

F

dis

con

tin

ua

tio

n

OR

: 0

.57

(0

.34

-0

.95

)E

PD

S 1

-we

ek

BF

4

-8 w

ee

ks

Ma

gn

itu

de

of

ass

oci

ati

on

Ga

ffn

ey

20

12

14

47

Lo

ng

itu

din

al

EP

DS

> 9

->

Lo

w

BF

in

ten

sity

OR

: 1

.57

(1

.16

-2

.13

)8

We

ek

sM

ag

nit

ud

e o

f a

sso

cia

tio

n

Causality

STRENGTH OF THE ASSOCIATION

The association is mildly to moderately strong




Significant p

Significant p

Significant p

Significant p- ---values and confidence intervals across

values and confidence intervals across



studies suggest confidence in significance of the




association

association

association

association

Moderate support for causal relationship




Causality

SPECIFICITY

Breastfeeding measures




�Exclusive vs Mixed vs Any vs Not

�Feeling regarding breastfeeding/ Self-efficacy

�Duration, intensity

Depressive sym

ptoms measures

Depressive sym

ptoms measures

Depressive sym

ptoms measures

Depressive sym

ptoms measures

�EPDS >9 or >12

�Mood scores

Mild to moderate support




Causality

CONSISTENCY

Relationship between measures of breastfeeding status




and measures of depressive sym

ptoms found


ptoms found


ptoms found


ptoms found

consistently across studies




Consistency provides moderate support for a causal




relationship

relationship

relationship

relationship

Causality

BIOLOGICAL PLAUSIBILITY

Oxytocin

Oxytocin

Oxytocin

Oxytocin

�Lactation and mood regulation

Prolactin

Prolactin

Prolactin

Prolactin

�Increases with breastfeeding

�Low levels associated with increased negative mood

Stress

Stress

Stress

Stress

�Inflammation triggers depression

�Induced via sleep deprivation

�Breastfeeding reduces stress levels

Provides moderate to strong support




Causality

TEMPORALITY

Breastfeeding Depressive sym

ptoms


ptoms


ptoms


ptoms

�Watkins 2011: Women with severe breastfeeding pain in the first day, first week, and

Watkins 2011: Women with severe breastfeeding pain in the first day, first week, and



second week more likely to be depressed at two months




�Ystrom

Ystrom

Ystrom

Ystrom2012: Adjusting for depressive sym

ptoms at 30 weeks gestation, postpartum

2012: Adjusting for depressive sym






depressive sym

ptoms at 6 months were predicted by increased bottle

depressive sym


depressive sym


depressive sym

ptoms at 6 months were predicted by increased bottle- ---feeding and

feeding and

feeding and

feeding and

mixed breastfeeding in the months prior. Additionally women with depressive




sym

ptoms during pregnancy had higher levels of depressive sym

ptoms postpartum

sym


ptoms postpartum

sym


ptoms postpartum

sym


ptoms postpartum

compared to women with out depressive sym

ptoms during pregnancy.







�Mezzacappa

Mezzacappa

Mezzacappa

Mezzacappa2002: From 10

2002: From 10

2002: From 10

2002: From 10 mins

mins

mins

minsbefore to 10

before to 10

before to 10

before to 10 mins

mins

mins

minsafter breastfeeding, mothers

after breastfeeding, mothers



reported a statistically significant decrease in negative mood compared with bottle



reported a statistically significant decrease in negative mood compared with bottle- ---

feeding. Breastfeeding mothers




Depressive sym

ptoms Breastfeeding

Depressive sym

ptoms Breastfeeding

Depressive sym

ptoms Breastfeeding

Depressive sym

ptoms Breastfeeding

�Henderson 2003: Women who experienced postpartum depression at anytime in the

Henderson 2003: Women who experienced postpartum depression at anytime in the



year after birth were more likely to stop breastfeeding and in most cases, onset of




depressive sym

ptoms preceded breastfeeding cessation

depressive sym


depressive sym


depressive sym


Causality

DOSE-RESPONSE & INTERVENTION

Ystrom

Ystrom

Ystrom

Ystrom2012:

2012:

2012:

2012:

�Bottle feeding higher risk factor for postpartum

depressive sym

ptoms than mixed breastfeeding

Provides Mild support




No intervention effect noted in these studies




Causality

SUMMARY

There is some evidence to suggest a causal relationship




Major question regarding directionality




Bias, confounding, chance not consistently found across




studies to explain association




�Small sample size, non-response, loss to follow-up (Hatton,

2005; Yonkers 2001; Dennis 2007).

�Misclassification: to lack of reevaluation of mothers or exclusion

of mothers based mixed breastfeeding status (Tashakori2012)

�Depressive sym

ptoms scores (Chaudron2001, Dennis 2007).

�Some confounders not considered: transition from

breastfeeding to solid foods and return to employm

ent

�Most studies subsets of larger studies not geared towards this

association

FUTURE STUDIES

Directionality

Directionality

Directionality

Directionality

Onset

Onset

Onset

Onset of depressive

of depressive

of depressive

of depressive sym

ptoms

sym

ptoms

sym

ptoms

sym

ptoms

Levels of

Levels of

Levels of

Levels of breastfeeding

breastfeeding

breastfeeding

breastfeeding

I IIIntervention

ntervention

ntervention

ntervention effect

effect

effect

effect

PUBLIC HEALTH IMPLICATIONS

Support

Support

Support

Support of breastfeeding practices to improve mental health

of breastfeeding practices to improve mental health



outcomes, im

prove mother

outcomes, im

prove mother

outcomes, im

prove mother

outcomes, im

prove mother- ---infant

infant

infant

infant bonding

bonding

bonding

bonding

If breastfeeding decreases SRPPDS encouragement of




breastfeeding initiation and duration could be used to decrease




severity of depressive sym

ptoms in mothers hesitant to use







antidepressants.

antidepressants.

antidepressants.

antidepressants.

If SRPPDS early on decreases breastfeeding intensity and duration




then early screening and recognition of depressive sym

ptoms


ptoms


ptoms


ptoms

and treatm

ent could im

prove breastfeeding outcomes

and treatm

ent could im


and treatm

ent could im


and treatm

ent could im


If the association is bidirectional then early recognition of




depressive sym

ptoms as well as support of breastfeeding

depressive sym


depressive sym


depressive sym


practice may confer improved mental health as well as other




health outcomes in mother and infants.




ACKNOWLEDGEMENT

Committee Members

Committee Members

Committee Members

Committee Members

�Dr. Ken Rosenberg

�Dr. Liz Adams

�Dr. TomiMori

Public Health and Preventive Medicine




Priya

Priya

Priya

Priya

Srikanth

Srikanth

Srikanth

Srikanth

Al Sandoval

Al Sandoval

Al Sandoval

Al Sandoval

laubinger thesis defense 2013-05-14x - state of · pdf fileootitis oo titis media, titis...

Documents