Graduate Thesis Proposal_GRAVES_5.2.2013.pdf

8/11/2019 Graduate Thesis Proposal_GRAVES_5.2.2013.pdf

1/47

! " # $ % & ' ! " # ) # * + ,

!"#$%&' %&)*&$ )+,-./* 0,"0*,*$%+&

&.$,%$%+& "1.2*$%+& -"##*'"# %& *&+&/%&" #"$$%&'3 * ,*&1+-%4"15

2+&$,+//"1 $,%*/6"7"22* 8 9,*:"#

!"# % ;


2/47

$

-+.,% #/ 0#1'%1'*()*+,-*.,+ ,+/)+0 1

,+&+,+23+4 56

7,89+3* 4.!!-,: 51

7,89+3* ;+43,)7*)82 58;4 %%

()!)*-*)824 154*.;: &(80 ;)-?,-! 1%7,89+3* *)!+()2+ 1%,+&+,+23+4 11

-77+2;)3+4 1@

%&&'()*+ *,-'.-/*01'(0 234'- 56%&&'()*+ **,7.-''(*(8 0993 5:%&&'()*+ ***,7.-''(*(8 7.-*&0 5;%&&'()*+ *


3/47

!"#$"%5!""#$%& (!"#$%#&' ")*!"#$%& ()

!"#$%5!"#$ &'( )*'+5!"#$ &"'(!)

!"#$%&'( *%+ ,-*.-#(

23'%"+'4"% 5%63%7


4/47

=

Overweight and obesity epidemic in the United States: The obesity epidemic in the United

States (US) is a well-established public health crisis associated with immense national healthcare

expenditures [1-4]. Obesity rates among American adults and children have been increasing over

the past three decades with a slight leveling off in recent years [1,9]. Over 35% of US adults and

almost 17% of US children and adolescents are classified as obese [9]. An effort to prevent and

reduce childhood obesity prevalence is of particular interest, because overweight or obese

children are more likely to be overweight or obese later in life, leading to chronic and possibly

fatal diseases such as type 2 diabetes and cardiovascular disease [5-8]. The first four months of

life has been identified as a potentially critical intervention period for prevention of both rapid

weight gain in the first year of life and subsequent overweight and obesity in childhood [9]. For

this time period, breastfeeding is recognized as the healthiest mode of infant-feeding, and may be

protective of rapid weight gain early in life [10, 11].

Self-regulation and normal infant intake for breastfed and formula-fed infants: Dewey and

colleagues [12] conducted a landmark study comparing weight-for-length and percent fat mass

between exclusively breastfed infants and exclusively formula-fed infants. There were 46

breastfed infants and 41 formula-fed infants included in the study, and the researchers controlled

for potential confounding variables such as parental socioeconomic status, ethnicity, maternal

education and anthropometric measurements, and infant sex and birth weight. Results indicated

that breastfed infants had a significantly lower weight-for-length compared to formula-fed

infants between 7 and 24 months, and the greatest difference occurred between 11 and 16

months. Also, the formula-fed infants reached a greater peak in percent body fat compared to


5/47

C

breastfed infants during the first year of life. Dewey and colleagues concluded that the probable

main contributing factor for these results was the difference in total energy intake between

groups. Formula-fed infants on average consumed significantly more kcal/day during the first

year of life compared to the breastfed infants, with intake ranging from 79-156 kcal more per day

for formula-fed infants [12].

Li and colleagues [13] further explored infant self-regulation to determine if breastfed

infants consumed fewer kcal per day due to the ability to better self-regulate intake, as they are

thought to be more in control of when a feed is terminated, and there is no external influence on

energy-density [14]. Breastfed infants must actively suckle to feed whereas formula-fed infants

can be more passive, and caregivers might manipulate the ability of formula-fed infants to self-

regulate. Therefore, it was hypothesized that infants who were fed expressed breastmilk from a

bottle would grow comparably to the formula-fed infants, despite differences in infant-feeding

mode. Li and colleagues [13] tested infant weight status in the first year of life as an outcome of

milk type (human vs. nonhuman milk) and feeding mode (breast vs. bottle). They found that

bottle-feeding was associated with greater infant weight status at one year of age, regardless of

the type of milk offered. In a separate study, Li and colleagues [15] found that infant weight

status at one year was positively associated with infant-initiated bottle emptying. This suggests

that breastfed infants may receive subtle signals over the course of a breastfeeding episode,

possibly in the form of increasing fat concentrations toward the end of a feed [13]. Theoretically,

secondary to emulsification of bottles of pumped breastmilk, regardless of milk type, bottle-fed

infants may not experience this variation in macronutrient profile, potentially affecting internal

hunger and satiety cues.


6/47

6

Need for intervention targeting formula-fed infants:Despite initiatives to increase worldwide

breastfeeding initiation and duration rates, many infants, particularly in the US, are being offered

infant formula and some are never breastfed. According to the Centers for Disease Control and

Prevention (CDC) 2011 Breastfeeding Report Card, about 25.4% of US infants and about 34.4%

of Tennessee infants are never breastfed [16]. The rates of exclusive breastfeeding continue to

decrease as infants age, with only 35.0% of US infants being exclusively breastfed at 3 months

of age and a mere 14.8% of US infants being exclusively breastfed at 6 months of age [16]. For

Tennessee infants, these rates at 3 months and 6 months are 27.9% and 12.8%, respectively [16].

Thus, breastfeeding initiation and duration rates in Tennessee are below the national averages.

Furthermore, Grummer-Strawn and colleagues showed that infants who were not being

exclusively breastfed were offered other milks, including cows milk and milk substitutes [17].

On average, the use of cows milk was not seen until the infants were about 9-10 months of age

[17]. Therefore, the younger infants who were not exclusively breastfed were primarily offered

infant formula. There are many risks associated with incorrect preparation of infant formula,

whether the caregiver is over-concentrating, diluting, or modifying the formula with the addition

of infant cereal and/or solid foods [18-24].

Modifying infant formula, overfeeding, and associated consequences: Behaviors such as

adding cereal to bottles, early introduction of solid foods, and overfeeding, contradict infant-

feeding recommendations for infants less than 4 months of age [18, 25]. Both addition of cereal

and early introduction of solid foods are thought to induce sleep, causing the infant to sleep for

longer periods of time, possibly motivated by maternal convenience [26-28]. Also, there is some

evidence supporting a moderate decrease in reflux or spit-up when cereal is added to the infant


7/47

:

formula [29]. However, adding cereal to an infants bottle could increase the risk for dental

caries, allergies, and bacterial infections [18, 21-24, 30-32], and early introduction of solid foods

could lead to overweight/obesity in childhood [33, 34].

Overfeeding is another concerning infant-feeding behavior. There are a variety of

practices leading to overfeeding and excess infant caloric intake other than adding cereal to the

bottle or early introduction of solid foods. For example, intentional and unintentional over-

concentration or offering more than 20 kcal per ounce of formula could be one mechanism

resulting in a heavier infant [35]. Premature infants are offered formula that is 24 kcal per ounce

to stimulate catch up growth [35]. Therefore, theoretically, even a small manipulation in

formula concentration may lead to unintentional weight gain over time for healthy infants.

Conversely, dilution of infant formula ( 185% Federal Poverty Level) population is more

likely to be overweight or obese later in life [36, 37]. Not only are low-income mothers more

likely to formula feed, but they are also at a greater risk for adding cereal to the bottle and early


8/47

#

introduction of solid foods [27, 28, 38, 39]. Crocetti and colleagues [27] found that out of 102

female primary caregivers, 76% were WIC participants, 39% of their infants were never

breastfed, and 44% introduced solid foods, including infant cereal, prematurely. Infants who

were given infant cereal before four months of age were also 31 times more likely to receive

other solid foods before four months of age. Of the mothers who offered solid foods prior to four

months of age, 80% claimed that their infants were not satisfied by formula or breastmilk alone,

and 53% stated that solid foods induced better sleep for their infants. Heinig and colleagues [28]

also discovered from focus groups with WIC participants, that many mothers reported the use of

fluids and solid foods to influence infant behavior and to induce infant sleep.

In addition, this population of low-income mothers has been shown to have increased

likelihood of inaccurate infant formula preparation practices that may increase the risk for later

overweight/obesity [37]. WIC provides infant formula for formula-feeding mothers but not

enough to sustain infants for the entire month [40]. Since WIC is a supplemental nutrition

program, the mother is responsible for purchasing formula with her own means after her supply

each month is depleted. Formula is much more expensive when compared to both infant cereal

and solid foods, so supplementing formula feeds with these alternatives may seem to be the most

logical and affordable option for low-income mothers. More convenient and cost-effective

behaviors can be tempting, especially if mothers are unaware of the potential long-term

consequences of incorrect infant formula preparation practices.

Online nutrition education in low-income populations: The US Department of Health and

Human Services, Education, and Commerce agrees that internet access in the home is a priority

for US citizens for health promotion and education purposes [41]. Reaching individuals by


9/47

;

means of the internet is a more cost-effective tool to promote nutrition education messages

compared to delivery in a traditional classroom setting [42, 43]. This approach allows for greater

dissemination of messages to a broader audience, including those living in rural areas, who might

be otherwise isolated from nutrition education due to lack of transportation and general lack of

resources [42-44]. Research indicates that about two-thirds of low-income individuals living in

rural areas have internet access [41, 42]. WIC, among other government health organizations, are

currently implementing online learning as an alternative to the traditional classroom learning

approach [43]. Even if online learning modules are not proven to be practical as a stand-alone

teaching strategy, online modules could be used in conjunction with a traditional classroom

approach [42]. This computer-based teaching strategy must be proven effective with regards to

initiating positive health behavior changes. Limited research is currently available testing the

effectiveness of online nutrition interventions among low-income mothers using randomized-

controlled study designs [41].

Implications for Future Research:There is a need to address the incorrect formula preparation

practices among low-income, formula-feeding mothers, including over-concentration, dilution,

and modifications/additions to infant formula. Initiating formula preparation behavior changes

among mothers may lead to positive short-term and long-term health outcomes for infants as

well as a decrease in associated national healthcare expenditures. An early intervention teaching

mothers how to correctly prepare formula bottles in a cost-effective, convenient manner must be

tested for validity before being distributed for use in various government health organizations,

such as WIC.


10/47

D"

Conclusion: There is an unmet opportunity to address inappropriate formula preparation

practices among US mothers such as over-concentration, dilution, and modifications or additions

to infant formula. More research is needed in the field of online nutrition education, especially

among the low-income population. Future research should address this gap by testing behavior

changes among low-income, formula-feeding mothers as a result of an online nutrition education

intervention.

5%/%"%1&%*

DE .FGH %1I .FJKHL 7M 8ABCDE# DF GHD'IJBF "FI "IK'BCGBFECL! #$%& '&()*+%&)$ ,-./0 $""#I M1N75D,56E

$E 9NOPQ .3I .FKKLGG 1)I .RKSHQ 3-I 1T)LUPGG 1%I 0FVFW .XI 2GPNFG >1M 7JBO"'BFGB KP KOBJQBDRHE

"FI KABCDE# DF EHB .FDEBI 4E"EBCS 5MMMT%66>I %Z]PO 13I ']]PSS &1I &KPPTP 1%I )RQNPK )@M -CCKGD"EDKF ABEQBBF VKCEF"E"' G"EGHT

XV RJKQEH "FI KABCDE# DF GHD'IHKKIN VJKCVBGEDOB GKHKJE CEXI#L8M )FE"ZB "FI RJKQEH KP AJB"CETPBI"FI PKJWX'"TPBI DFP"FEC DF JB'"EDKF EK EHB EDWDFR KP DFEJKIXGEDKF KP GKWV'BWBFE"J# PKKICN EHB

;-,()2? CEXI#L ;"ODC -JB" ,BCB"JGH KF ("GE"EDKFS )FP"FE 2XEJDEDKF "FI ?JKQEHL1*./ 3/-(%/.+

D;;5I [%N;;;,D""6E

D5E 3H -I 1FNFOHF XI 2PHQ 7@I 8KR]]PK,7SKFUQ 31M ,DCZ KP AKEE'BTPBBIDFR PKJ J"VDI QBDRHE R"DF

IXJDFR EHB PDJCE #B"J KP 'DPBL1+*8 3-(%/.+ 1()$-6* ,-( $"D$I 5@@N=5D,=56E

D=E \LONPY '%I XLZQYLQ 73I \RNZPY 79I \LJWHQYLQ X1I @RSSP (2I 2HYZPK X9M ;BOB'KVWBFE KP EHB

JBCVKFCDOBFBCC EK GHD'I PBBIDFR GXBC CG"'BL1>>-.%.- $"D5I @UN$D",$D;E


11/47

DD

DCE 3H -I 2PHQ 7@I 8KR]]PK,7SKFUQ 31M -CCKGD"EDKF KP AJB"CEPBBIDFR DFEBFCDE# "FI AKEE'BTBWVE#DFR

ABH"ODKJC "E B"J'# DFP"FG# QDEH DFP"FEC\ JDCZ PKJ B]GBCC QBDRHE "E '"EB DFP"FG#L3-(%/.+%*6 $""#I

5%% 4XVV' %N7::,#=E

D6E 9KFG -I @F`]FQ 3I %YYFO %I BHVVPQ]P`PK 3I @RZKLU XI %RYSHQ %I @F`]FQ '9M )''DGDE IJXR

B]VKCXJB DF V"EDBFEC BO"'X"EBI PKJ "''BRBI GHD'I "AXCB "FI FBR'BGEL3-(%/.+ '=-+: #/+- $"DDI

%YN=;",=;CE

D:E 8KR]]PK,7SKFUQ 31I 7TFQGLQ >7I 2PHQ 7@M )FP"FE PBBIDFR "FI PBBIDFR EJ"FCDEDKFC IXJDFR EHB

PDJCE #B"J KP 'DPBL3-(%/.+%*6 $""#I 5%% 4XVV' %N756,=$E

D#E >GPHQ]FQ -M 3-(%/.+%* 94.+%.%)& ?/&(0))@ACSZ POQ$""=E

D;E 1LKSLQ %1M 4#WVKCDXW @N :KXFR VBKV'BS "JEDPDGD"' FXEJDEDKF "FI EJ"FCDEDKF"' G"JBL *HB

FXEJDEDKF"' GH"''BFRBC KP EHB #KXFR "IX'E QDEH G#CEDG PDAJKCDCN EJ"FCDEDKFL3+)* 94.+ B)* $"";I

@[N=5",=="E

$"E 1LGHQF >1I >HPG` 1M .FIBJCE"FIDFR IBVJBCCDOB C#WVEKWC "WKFR HDRHTJDCZS VJBRF"FES -PJDG"FT

-WBJDG"F QKWBFLC)=-&6 ?-/$.8 2664-6 $"DDI %5N$;5,5"5E

$DE (LKKHY X1I @FKKHNF >I >GHQNPQY]HSZ 8I \Lbb]FQ 1I 'HYPQVFKSZ 87I 'KGHTZ \%I -PUPKY 1M *DWDFR KP

DFDED"' GBJB"' B]VKCXJB DF DFP"FG# "FI JDCZ KP DC'BE "XEKDWWXFDE#L!1,1 $""5I %M6ND:D5,D:$"E

$$E .FKGPSSH .I .FSSFQPL %M >KWB VJBV"J"EDKF KP VKQIBJBI DFP"FE PKJWX'"N DC DE C"PB^1*./ 3/-(%/.+

$""#I MYNDD5D,DD5$E$5E -HTZFKOY 81I 8RKSGPK X@I @PRTZFS 3-M 4XJODO"' "FI RJKQEH KP +FEBJKA"GEBJ C"Z"_"ZDD DF DFP"FE

JDGB GBJB"' JBGKFCEDEXEBI QDEH Q"EBJS WD'ZS 'D`XDI DFP"FE PKJWX'"S KJ "VV'B aXDGBL! 1>>$ ,%*+)0%)$

$""CI MMN#==,#C"E

$=E &LLGP X%I @FKKHNF >I 3PRQN )4I \Lbb]FQ 1I 'HYPQVFKSZ 87I -PUPKY 1I (LKKHY X1M *DWDFR KP DFDED"'

B]VKCXJB EK GBJB"' RJ"DFC "FI EHB JDCZ KP QHB"E "''BJR#L3-(%/.+%*6 $""6I 55YN$D:C,$D#$E

$CE .RVPKL XI .ZFQTGLQ @I 7FQTZP[ 7I -H^PKL 1I -LOKHNRP[ %@I @FKKHNF .M )WVJKODFR EHB `X"'DE# KP

DFP"FE C'BBV EHJKXRH EHB DFG'XCDKF "E CXVVBJ KP GBJB"'C BFJDGHBI QDEH EJ#VEKVH"FS "IBFKCDFBT

U\TVHKCVH"EBS "FI XJDIDFBTU\TVHKCVH"EBL94.+ 9-4+)6*% $"";I 5%N$:$,$#"E

$6E .KLTPSSH 1I )ROFY -I >KRN]FQ 7M 7"JBFE"' AB'DBPC "FI VJ"GEDGBC JBR"JIDFR B"J'# DFEJKIXGEDKF KP

CK'DI PKKIC EK EHBDJ GHD'IJBFL#$%& 3-(%/.+ D38%$/E $""=I )I >F^FQFNZ,&KLTZFYWF >I .LZPQ -I &FQTZRGF XM ="JJDBJC EKGKWV'D"FGB QDEH DFP"FETPBBIDFR JBGKWWBFI"EDKFC "WKFR 'KQTDFGKWB QKWBFL! ?4= F/*.

$""6I %%N$:,5#E

$#E \LK^FSZ %I )[HPTZTHFK[ &I 7[FcPUYWF \M *HB BPPBGE KP EHDGZBFBITPBBI DFEBJOBFEDKFC KF

R"CEJKBCKVH"RB"' JBP'X] DF DFP"FECN C#CEBW"EDG JBODBQ "FI WBE"T"F"'#CDC KP J"FIKWD_BIS

GKFEJK''BI EJD"'CL3-(%/.+%*6 $""#I 5%%NPD$6#,D$::E

$;E &-bC "AKXE ="A# &KJWX'"S =KEE'B &BBIDFR "FI )FP"FE 2XEJDEDKFLBLKGO \PFGSZ 9KNFQH[FSHLQa

$"";E

5"E 0)3N &JB`XBFE'# "CZBI XBCEDKFC "AKXE GBJB"'L1FK`GFQO )PJFKS]PQS Lb \PFGSZE

5DE _HPNGPK %,8 77I \RVPK )I \R]]PG 1I @LQHbFTHL 'M +"J'# DFP"FE PBBIDFR "FI JDCZ KP IBOB'KVDFR

E#VB 5 ID"ABEBCT"CCKGD"EBI "XEK"FEDAKIDBCL!1,1 $""5I %M6ND:$D,D:$#E

5$E 1T1HGGPQ *.I -FSSFQFSKF` 3I )RbbHPGO X%I 1LKKHYLQ X3I 1FT3FRNZGHQ 71I 8PQSHGH 7I 1RZGZFRYGPK @7M

*HB B"J'# KJDRDFC KP '"EBJ KABCDE#N V"EHQ"#C "FI WBGH"FDCWCL1(5 'G> ,-(


12/47

D$

5CE 7HQNZ 8>I >LNFQ 1)I %(-=%)$ $""#I 5[N6#$,6;CE

56E 9GH^PHKF < -'I 9G]YSPO XI 8ZPGbH 31M *HB 0)3 7JKRJ"WN ="GZRJKXFIS *JBFICS "FI )CCXBC &KKI

-CCDCE"FGB "FI 2XEJDEDKF ,BCB"JGH ,BVKJE $""$E

5:E 3H -I )FKGHQN (I 1FRKHTP 'I @FKWPK 3I 8KR]]PK,7SKFUQ 31M =JB"CEPBBIDFR J"EBC DF EHB .FDEBI

4E"EBC A# GH"J"GEBJDCEDGC KP EHB GHD'IS WKEHBJS KJ P"WD'#N EHB %66% 2"EDKF"' )WWXFD_"EDKF

4XJOB#L3-(%/.+%*6 $""CI 55UNP5D,5:E

5#E 1PQQPGGF X%I _HPNGPK &I @KHPbPG -I (L^FW 0M &BBIDFR )FP"FEC "FI *KII'BJC 4EXI#N EHB E#VBC KP

PKKIC PBI EK >DCV"FDG DFP"FEC "FI EKII'BJCL! 1= ;%-. 166)* $""6I 56@N7;6,D"6E

5;E %SWHQYLQ (3I @HGGHQN %7I )PY]LQO 71I 8LGO -7I 0LRKQFY,\FKOS %M -CCBCCWBFE KP EHB FXEJDEDKF

"FI VH#CDG"' "GEDODE# BIXG"EDKF FBBIC KP 'KQTDFGKWBS JXJ"' WKEHBJCN G"F EBGHFK'KR# V'"# "

JK'B^! #)==4&%.H ?-/$.8 $"":I 1%N$=C,$6:E

="E .FYP & .1I \HQL XM 8F'DFB 2XEJDEDKF +IXG"EDKFN +FH"FGDFR 8VVKJEXFDEDBC PKJ (DWDEBIT,BCKXJGB

(B"JFBJCL!)4+&/$ )I 'G.-&6%)& $"DDI


13/47

D5

!"#$%&' 8499+":

Preparing infant formula incorrectly lead to many short-term and long-term negative healthoutcomes for infants [18, 21-24, 30-32]. Potential consequences include a greater risk for

bacterial infections, dental caries, allergies, and additional calories for the infant that may beassociated with unfavorable weight gain [18, 21-24, 30-32]. Rapid weight gain during the first 4months of life is predictive of overweight and obesity later in life [5, 6, 9, 45]. Overweight andobese children and adults have an increased risk of several associated co-morbidities, such astype 2 diabetes and cardiovascular disease, and the obesity epidemic is a well-established publichealth crisis in the United States (US) [5-8]. Preventing rapid weight gain in infants has beenidentified as a potential intervention point for obesity prevention, particularly for formula-fedinfants [9, 13].

There is an unmet opportunity to address improper formula preparation practices among US

mothers such as over-concentration, dilution, and modifications or additions to infant formula.The United States Department of Agriculture (USDA) has developed a series of core

nutrition education messages for mothers of children ages two and older [46]. These

messages were designed to promote healthy eating practices among mothers and their

children; however, the messages do not speak to the unique nutrition needs of infants. Inorder to address this gap, the research team has established points of intervention based on thequalitative data analysis of one-on-one, audio-recorded, in-depth interviews with low-income,formula-feeding mothers. The research team developed two nutrition education messages, usingthe nominal group process, that were designed to help mothers prepare infant formula correctly[47]. These messages emphasize the importance of preventing the overfeeding of infants and thepotentially harmful effects of additions, such as cereal or the premature introduction of solid

foods. An online intervention will be developed based on these nutrition education messages andtested using a randomized-controlled study design. The research team will determine if theintervention results in formula preparation behavior change among mothers. The purpose of thisstudy is increase availability of a novel nutrition education intervention to effectively improveformula preparation practices among low-income, formula-feeding mothers, thus potentiallyimproving short-term and long-term health outcomes for US infants.

Low-income, formula-feeding mothers with infants < 3 months of age will be recruited tocomplete online modules designed to test the nutrition education messages. Infant intake will berecorded by mothers for 24 hours (baseline), after which the mothers will be randomly

assigned to one of three groups: intervention online module 1 (educational video), interventiononline module 2 (educational video plus interactive segments), or online control module(educational sham video). Within two weeks of module completion, infant intake will berecorded by mothers a second time for 24 hours (follow-up). Intake data will be analyzed forsignificant differences in behavior (ex: amount of infant formula offered and/or amount of infantformula consumed).


14/47


15/47

DC

rates in Tennessee are below national averages. Infants less than four months of age who are not

breastfed are most likely offered infant formula instead [17]. There are risks associated with

incorrect preparation of infant formula, whether the caregiver is over-concentrating, diluting, or

modifying the formula with the addition of infant cereal and/or solid foods [18, 21-24, 30-32].

Modifying infant formula, overfeeding, and associated consequences:Behaviors such as

adding cereal to bottles, early introduction of solid foods, and overfeeding, contradict infant-

feeding recommendations for infants less than 4 months of age [18, 25]. Both the addition of

cereal and the early introduction of solid foods are thought to induce and extend infant sleep, and

are behaviors potentially motivated by maternal convenience [26-28]. Mothers also report adding

infant cereal to bottles of infant formula, in an effort to manage reflux or spit-up [28, 29].

Though there is limited evidence indicating a moderate decrease in reflux or spit-up when cereal

is added to the formula [29], this is only to be indicated in specific, diagnosed situations [49] and

is otherwise recommended against for infants less than 4 months of age [18, 25]. In fact, adding

cereal to bottles of infant formula is thought to increase the risk for dental caries, allergies, and

bacterial infections [18, 21-24, 30-32]. In addition, cereal in the bottle could theoretically

contribute to excess calories for the infant and subsequent overfeeding [34].

In addition to cereal in the bottle and early solid food introduction, there are other practices

leading to excess infant intake such as over-concentration, dilution, or too many ounces offered.

For example, intentional and unintentional over-concentration, or offering more than 20 kcal per

ounce, of formula could be one mechanism resulting in a heavier infant [35]. Premature infants,

if also low birth weight, are offered formula that is 24 kcal per ounce to stimulate catch up


16/47

D6

growth [35]. Therefore, theoretically, even a small manipulation in formula concentration may

lead to unintentional weight gain over time for healthy infants. Conversely, dilution of formula

( 185% Federal Poverty Level) is more likely to be overweight

and obese later in life [36, 37]. Not only are low-income mothers more likely to formula feed,

but they are also at a greater risk for adding cereal to the bottle and for early introduction of solid

foods [27, 28, 38, 39]. In addition, this population of low-income mothers has been shown to

have increased likelihood of inaccurate infant formula preparation practices that may increase the

risk for later overweight/obesity [37]. WIC provides a standard amount of infant formula per

month [40]. However, as infants grow larger, the standard amount is eventually less than the

infant needs for the entire month [40]. Since WIC is a supplemental nutrition program, parents

are responsible for purchasing formula after this monthly supply is depleted. This is of concern,

as formula is much more expensive than both infant cereal and solid foods, and supplementing

formula feeds with these alternatives may seem to be the most logical and affordable option for


17/47

D:

low-income populations. More convenient and cost-effective behaviors can be tempting,

especially if mothers are unaware of the potential long-term consequences of improper infant-

feeding practices or if the behaviors are considered incongruent with real world needs [28].

Preliminary data from previous phases of this project: Preliminary data suggest that mothers

eligible for WIC are frequently preparing powdered infant formula in a way that may be

detrimental for their infants short-term and long-term health (pilot work). In addition, during in-

depth interviews, mothers report never being advised by healthcare providers about how to

correctly prepare powdered infant formula. Pilot work in a low-income, formula-feeding

population, conducted in this laboratory, has revealed a population that recalls anxiety and

discomfort associated with learning to prepare a bottle of infant formula. Mothers in this

population report seeking advice from family members and peers, in addition to seeking advice

from health care professionals. This is consistent with findings from previous research conducted

by Heinig and colleagues [28].

Results from earlier phases of this study suggest that infants were being fed significantly more

than the recommended amount, and possibly even more than reports of actual intake in the

literature [12, 50]. During Phase I of this project, the same infants were followed over time and

data were collected at ~2 months and ~4 months of age. Preliminary data from Phase I suggest

that bottle preparation at ~2 months is linked to weight status at ~4 months. During qualitative,

in-depth interviews, occurring in Phase II, mothers reported that offering too many ounces per

feed is related to a variety of outcomes such as spitting up and bellyaches. These data illustrate


18/47

D#

the importance of intervening with WIC-eligible mothers and the potential implications of

preparing infant formula incorrectly.

Justification of an online intervention: The research team will deliver nutrition education

messages, targeting formula preparation practices, to WIC-eligible mothers through an online

learning module. The US Department of Health and Human Services, Education, and Commerce

agrees that internet access in the home is a priority for US citizens for health promotion and

education purposes [41]. Reaching individuals by means of the internet would be a more cost-

effective tool to promote nutrition education messages, compared to delivery in a traditional

classroom setting [42, 43]. This approach allows for greater dissemination of messages to a

broader audience, including those living in rural areas, who might be otherwise isolated from

nutrition education due to lack of transportation and general lack of resources [42-44]. Second,

recent research indicates that about two-thirds of low-income study participants living in rural

areas have internet access [41, 42]. Not only would an online intervention be more cost-effective

and reach a larger audience, but offering this type of tool would be in alliance with the strategies

government health organizations, such as WIC, are currently implementing [43]. Even if the

online learning module is not shown to be practical as a stand-alone teaching strategy, the

module could be used in conjunction with a traditional classroom approach [42]. Offering access

to the online tool would require minimal additional effort for the staff of government and

community health organizations and would provide another level of education that may enhance

knowledge retention and result in positive health behavior changes.


19/47

D;

Despite valid justification of utilizing online delivery of nutrition education messages, this

teaching strategy must be proven effective with regards to initiating positive health behavior

changes. Limited research is currently available testing the effectiveness of online nutrition

interventions among low-income mothers using randomized-controlled study designs. The

proposed study will address this gap by not only testing the nutrition education messages, but

also by developing an online teaching strategy, thus enhancing the literature available regarding

online delivery of nutrition education messages and associated behavior changes.

Need for intervention: Although breastfeeding is the recommended mode of infant feeding,

there is a critical need to establish interventions targeting proper formula preparation and infant

feeding behaviors among those who choose to formula feed or are unable to breastfeed their

infants [10, 11, 16]. Efforts to increase national breastfeeding rates should continue so US infants

can be as healthy as possible. In the meantime, the large population of formula-feeding mothers

needs to be offered guidance in order to prepare infant formula in an appropriate manner to

potentially decrease the negative health outcomes associated with incorrect formula preparation.

Initiating formula preparation behavior changes among mothers would lead to positive short-

term and long-term health outcomes for infants as well as a decrease in national healthcare

expenditures. Early intervention is critical to prevent mothers from engaging in inappropriate

formula preparation practices that may be detrimental to their infants short-term and long-term

health. Given the public health importance, there is a need for a standard education module to

illustrate to formula-feeding mothers how to correctly prepare infant formula. An early

intervention, teaching mothers how to properly prepare formula bottles in a cost-effective,

convenient manner must be tested for effectiveness before being distributed for use in various


20/47

$"

government health organizations such as the Expanded Food and Nutrition Education Program

(EFNEP) and WIC.

Dr. Katie Kavanagh, the principal investigator (PI), and an experienced researcher in the field of

infant-feeding, will be supervising the project and collaborating with two Co-principal

investigators (Co-PIs), Dr. Janie Burney and Dr. Betty Greer with University of Tennessee

Extension. Dr. Burney and Dr. Greer have extensive experience with message transmission and

online module development, so the research team will be highly qualified to implement this

online intervention.

Innovation: To the research teams knowledge, there are no current preventative nutrition

education interventions targeting infant formula preparation, specifically addressing the issues of

over-concentration, dilution, and modification to infant formula.

Government health organizations, such as WIC, are beginning to offer services online and are

implementing computer-based learning strategies on-site [51]. The online learning module

generated from this study will coincide with the increasing use of technology by government

health organizations. The online nutrition education modules developed and tested as an

objective of this project could be offered in addition to traditional classes or as a stand-alone

teaching strategy. More research is needed in the field of online nutrition education, especially

among the low-income population [41]. This study will address this gap by testing behavior

changes among formula-feeding, low-income mothers as a result of an online intervention. The


21/47

$D

research team will provide an innovative approach that addresses the current and future needs of

government health organizations by utilizing computer-based learning.

>.$%&'36%*

Objective 1:Conduct a randomized-controlled trial, to determine if targeted nutrition

education messages, designed to increase compliance with infant-feeding

recommendations and delivered in an online format (intervention groups 1 and 2), will

result in knowledge and behavior change, compared to messages unrelated to infant

nutrition (control group), and to do so in a population of low-income, formula-feeding

mothers with infants less than 3 months of age.

Objective 2:Determine if messages, delivered using embedded interactive

subcomponents (intervention group 2), result in greater knowledge and behavior change,

compared to a more traditional slideshow format (intervention group 1).


22/47

5%*%+"&? @%'?#=*

Brief description of Phase I and Phase II: All phases of this study recruited WIC-eligible,

formula-feeding mothers (>16 years of age) with healthy infants < 3 months old.

Phase I:During Phase I, the research team collected infant formula samples for one 24 hour

period at two different time points (~2 months of age and ~4 months of age). At each time point,

infant weight, length, and head circumference were measured by trained research staff. In

addition, mothers completed infant intake forms for 48 hours (samples collected for the last 24

hours) to assess the amount of formula offered to the infant, amount leftover after feeding,

whether or not cereal was added to the bottle, how much cereal was added (if any), amount of

spit-up (if any), and who fed the infant the bottle. The intake forms also accounted for calories

obtained from other sources, such as juice or solid foods, and how many hours the infant was

sleeping in between feeds. The 54 mothers who participated at both time points also completed a

detailed questionnaire at each time point about infant-feeding behaviors and beliefs.

Phase II:Phase II focused on developing nutrition education messages based on the findings

from Phase I and additional one-on-one, in-depth recorded interviews with mothers about

feeding their infants also conducted during this phase. The script for the interview was designed

by an expert committee, and the research team received extensive training from Dr. Suzie

Goodell, a qualitative research specialist. Prior to the interview, mothers were asked to

demonstrate to the research assistants how they normally prepare formula bottles, and

observations were recorded. A total of 13 interviews were completed, and all members of the


23/47

$5

research team, including Dr. Goodell, agreed that saturation was reached after the first 10

interviews.

Nutrition education messages were developed by the research team based on prevalent improper

formula preparation practices encountered during Phase I and during the Phase II interviews and

observations. These messages were presented to a professional work group populated by local

community members who coordinate programs serving low-income, formula-feeding mothers

with young infants. During the professional work group, messages were ranked by perceived

importance using the nominal group process and then were refined based on input and group

discussion. This ranking and refining process was repeated among paraprofessionals working

directly with the target population. Focus groups with low-income, formula-feeding mothers

were held to further discuss and refine the messages. This process resulted in two final nutrition

education messages. The first message conveys that breastmilk and formula provide all of the

nutrients infants need until 4-6 months of age, and advises against modifying infant formula with

cereal and early introduction of solid foods. The second message focuses on preventing

overfeeding infants and the small size of the infant stomach. These messages will be combined to

form the intervention for Phase III, which will further test the developed messages along with

testing teaching strategies.


24/47

$=

Message 1:Breastmilk and formula are fillingand provide everything your baby needs to be

healthy until 4 to 6 months.

!

Formula and breastmilk arepacked with vitamins andminerals. Adding cereal, juice,or extra water takes awayimportant nutrients from yourbaby.

! Cereal and juice have a lot ofsugar and dont have as muchfat and protein as formula orbreastmilk. Fat and protein helpyour baby to stay fuller, longer.

Message 2:All babies have small stomachs.Feed your baby small amounts more

frequently.

!

All babies spit up, butoverfeeding may lead tobellyaches and can makespitting up worse.

! Save formula, time, and moneyby making small amounts.Small amounts heat up fasterand you waste less.

Phase III: The purpose of Phase III is to use a randomized-controlled study design to test the two

developed nutrition education messages by assessing infant formula preparation knowledge and

preliminary behavior changes as a result of an online intervention. Resources from University of

Tennessee Extension will be utilized to develop a strategic and effective online learning module.

The online learning modules are currently being finalized by the research team with the

assistance of resources available to the Co-PIs in University of Tennessee Extension. As

previously described, three online modules will be developed: intervention online module 1

(educational video), intervention online module 2 (educational video plus interactive segments),

and an online control module (educational sham video focusing on preparing the home for a

crawling infant). All modules will be followed by a short questionnaire in order to ensure

completion of the online segment, and modules will be designed to take no more than 10 minutes

to complete, but mothers will have no time restrictions. The online learning modules for the two

intervention groups will communicate the two nutrition education messages developed in Phase


25/47

$C

II. Topics covered in the intervention modules will include the satiating effect of breastmilk and

infant formula and feeding age-appropriate amounts to young infants with small stomachs.

The two intervention modules will incorporate the same animation, and the messages will be

delivered by an animated infant. The only difference between the two intervention modules will

be the presence of interactive components in intervention module 2. For example, the mother

might have to choose the correct, age-specific, amount of formula to offer an infant after

receiving the education pertaining to the size of the infant stomach. She may then answer

questions about accurate formula preparation, given different ingredients to choose from, after

receiving education about the satiating effect of formula and breastmilk. If a mother chooses an

incorrect option, she will be corrected, provided with explanations, and prompted to repeat

response to the question until the correct answer is chosen. If she answers correctly, she will be

provided with an explanation for why her answer is correct. The interactive module will

designed so that the research team can determine which incorrect responses were chosen prior to

the mother choosing the correct response (if any) and how many times the mother answered

incorrectly (if applicable). The online control module will not address infant feeding or any

related topics that may indirectly influence maternal feeding behaviors. To reduce bias, the

control module will cover a topic pertinent for mothers of infants ~3 months of age, such as

safely preparing the home for a soon to be crawling infant. The control module will be strictly

educational with no interactive components.

Objective 1:Conduct a randomized-controlled trial, to determine if targeted nutrition education

messages, designed to increase compliance with infant feeding recommendations and delivered


26/47

$6

in an online format (intervention groups 1 and 2), will result in knowledge and behavior change,

compared to messages unrelated to infant nutrition (control group), and to do so in a population

of low-income, formula-feeding mothers with infants less than 3 months of age.

Phase III recruitment:Research participants for this project will be WIC-eligible, formula-

feeding mothers (>16 years of age) with < 3 month old, healthy infants in the Appalachian region

of the US (will be screened for eligible zip codes). Parity is not an inclusion/exclusion criterion

for this study. Mothers will be recruited from one of several sources. The first source will be the

Knox County WIC clinic, via flyer distribution (Appendix I). Another recruiting source for

participants will be EFNEP, which will be coordinated by one of the Co-PIs working at

University of Tennessee Extension. Similar to previous phases, recruitment flyers (Appendix I)

will be distributed at local community health organizations serving the target population, such as

the Pregnancy Help Center, Helen Ross McNabb, and the Hope Resource Center. Also, online

recruitment will continue by posting recruitment advertisements on Facebook, Twitter, and

Craigslist. The research team has been successful with online recruiting during previous phases

of this project. A total of 90 mothers must be recruited (30 per group) in order to have a sample

size large enough to detect differences in infant-feeding behaviors, such as total ounces offered

and consumed. Recruitment will continue until this sample size is achieved.

Phase III detailed methods:Potential participants who contact the Infant, Child, and Adolescent

Nutrition (ICAN) Lab at The University of Tennessee, Department of Nutrition, will be screened

for eligibility. Eligible participants will be those who are primarily formula-feeding (offering no

more than 2 breastfeeds a day), using powdered or from-concentrate infant formula, and who are


27/47

$:

low-income (based on 2012 2013 WIC eligibility guidelines). These standards will be revised

in accordance with the new version of WIC eligibility guidelines, which will be released in July

2013. Eligible participants will be mothers with infants who were normal birth weight (> 2500 g

or 5.5 lbs) and whose infants are free from any chronic health conditions affecting infant intake,

including but not limited to heart conditions and respiratory conditions. In addition, infants must

be < 2.5 months of age (76 days) at recruitment, as this will allow time for completion of the

online learning module before the infant is 3 months of age. Eligible participants must have a

reliable phone for communication of study activities and must have access to the internet. They

must also reside in the Appalachian region of the US, so they will be screened for eligible zip

codes. A screening tool similar to the ones from previous phases will be used (Appendix II).

2012-2013 WIC Eligibility Criteria

Persons inFamily orHousehold Size

Annual MonthlyTwice-

MonthlyBi-Weekly Weekly

1 $20,665 $1,723 $862 $795 $398

2 27,991 2,333 1,167 1,077 539

3 35,317 2,944 1,472 1,359 680

4 42,643 3,554 1,777 1,641 821

5 49,969 4,165 2,083 1,922 961

6 57,295 4,775 2,388 2,204 1,102

7 64,621 5,386 2,693 2,486 1,243

8 71,947 5,996 2,998 2,768 1,384

Each Add'lMember Add

+$7,326 +611 +306 +282 +141

Upon determination of eligibility, a brief explanation of study activities will be given, and an

expression of interest by the participant will be obtained. Thereafter, two copies of the Consent

Form (Appendix IV) and 24 hour infant intake forms (Appendix V) will be mailed to the


28/47


29/47

$;

knowledge retention and reaction to the online module. At this time, research assistants will once

again collect 24 hour infant intake data over the telephone. Participants will be asked to mail the

completed intake record to the research lab, using a self-addressed, stamped envelope (SASE)

supplied by the research team. Only participants from the two intervention groups will be asked

to complete the exit interview. Following the exit interview (if applicable) and after completing

the follow-up 24 hour infant intake record, participants in all three groups will be compensated

with the remaining $20 gift card to a national discount merchant, which will be mailed to them.

Only data on infant intake, formula preparation, and knowledge will be collected, placing the

participants at minimal risk when completing this study. In addition, all online modules will be

password protected, increasing confidentiality. The research assistants will assign each

participant with a unique user name and password in order to access the online modules. All data

will be collected at baseline (~2.5 months of infant age), and within 2 weeks of completing an

online module. The length of time from baseline to follow-up will be no more than 3 weeks.

Objective 2:Determine if messages, delivered using embedded interactive subcomponents

(intervention group 2), result in greater knowledge and behavior change, compared to a more

traditional slideshow format (intervention group 1).

Data analysis: The sample size of 30 participants per group was calculated based on literature

describing intake of normal, healthy infants at ~4 months of age (means and associated standard

deviations) [12, 50]. A power of 0.8 and a significance level of 0.05 were used in the sample size

calculation. SPSS (version 20.0) will be used for statistical analysis of baseline and follow-up


30/47

5"

infant intake data. Statistical analyses will determine if there was a significant difference in

formula modifications and overall infant intake between groups.

After data collection, data will be cleaned to account for missing variables. Then, frequencies

and correlations will be calculated to determine if randomization was successful and if any

independent variables affected the outcome variables of ounces offered and consumed. If, for

example, infant age (days) is correlated with amount of infant formula offered (ounces), then

infant age would be a confounding variable, and would be accounted for in the forthcoming

statistical analyses. Furthermore, if by chance, one group was comprised of significantly more

male infants compared to the other two groups, despite randomization, and infant gender was

related to the outcome variables, then this would be accounted for in the upcoming analyses as

well. Following descriptive statistical analyses, the research team will compare the outcome

variables, ounces of infant formula offered and consumed, based on group assignment. This will

be completed by running a multivariate analysis of variance (MANOVA) test. This analysis will

determine if there were statistically significant differences in the ounces of infant formula

offered and consumed between groups. A subsequent post-hoc analysis will then determine

where the statistical differences occurred (ie: between which groups). Linear regression analyses

would be used to control for potential confounding variables.

These analyses will be used to determine if receipt of the message alone impacted behavior and

if differences were significant between the intervention group receiving just the educational

video compared to the intervention group receiving the educational video plus the interactive

components: thus, testing the effectiveness of the nutrition education messages and the


31/47

5D

effectiveness of the different teaching strategies. As a secondary exploratory analysis, the

research team will look at differences in cereal use between groups based on group assignment.

This will not be a primary outcome tested, however, because the sample size for this project was

calculated based on infant intake data, not based on data for cereal usage. The research team

anticipates that the sample size might not be large enough to reach statistical significance for

differences in cereal use based on group assignment. However, this will be an interesting

outcome to explore to observe if the messages changed infant-feeding behaviors in regards to

addition of cereal to infant formula, and may be used as pilot data for future projects and

proposals.

Application of the Research Results: The results and online learning module will be made

available to community programs such as EFNEP and WIC and other professionals who work

with low-income, formula-feeding mothers with young infants. In addition, findings will be

presented at a national conference and published in a peer-reviewed journal.

2393'+'3#1*

Despite plans for successful recruitment strategies, the research team anticipates difficulties

recruiting WIC-eligible, formula-feeding mothers with < 3 month old, healthy infants. However,

recruitment has become less of a problem as the study has progressed. In addition, Phase III will

not include a home visit, most likely enhancing recruitment. Phase I and Phase II both required a

home visit, but Phase III will be primarily internet-based with brief telephone interaction.

Mothers will be able to complete study activities at a time and location that is convenient for

them. Therefore, the research team anticipates fewer drop-outs compared to previous phases. In


32/47

5$

addition, the online data collection methodology allows the research team to reach a larger

geographic area than previous phases, because driving distance will not be a limiting factor for

study recruitment.

8'4=: A,#7 ;3+B"+9

!"#$%&' -39%,31%

Study

Activity Mar-2013

Apr-2013

May-2013

Jun-2013

Jul-2013

Aug-2013

Sep-2013

Oct-2013

Nov-2013

June-2014

2014-

2015

Begin recruiting

participants

Data

Collection

Statistical analysis of data and

interpretation of results

Defend graduateThesis

Disseminate findings/Publish

7TKPPQPO bLK PGHNHVHGHS`

$= ZLRK VFYPGHQP HQbFQSHQSFWP

-FQOL]H[PO

'dJPKH]PQSFG]LORGP $

&LYS $= ZLRKHQbFQS HQSFWP

%QFG`[P

'dJPKH]PQSFG]LORGP D


.LQSKLG]LORGP


%QFG`[P

'dTGROPO


33/47


34/47

5=

$$E .FKGPSSH .I .FSSFQPL %M >KWB VJBV"J"EDKF KP VKQIBJBI DFP"FE PKJWX'"N DC DE C"PB^1*./ 3/-(%/.+

$""#I MYNDD5D,DD5$E

$5E -HTZFKOY 81I 8RKSGPK X@I @PRTZFS 3-M 4XJODO"' "FI RJKQEH KP +FEBJKA"GEBJ C"Z"_"ZDD DF DFP"FE

JDGB GBJB"' JBGKFCEDEXEBI QDEH Q"EBJS WD'ZS 'D`XDI DFP"FE PKJWX'"S KJ "VV'B aXDGBL! 1>>$ ,%*+)0%)$

$""CI MMN#==,#C"E

$=E &LLGP X%I @FKKHNF >I 3PRQN )4I \Lbb]FQ 1I 'HYPQVFKSZ 87I -PUPKY 1I (LKKHY X1M *DWDFR KP DFDED"'

B]VKCXJB EK GBJB"' RJ"DFC "FI EHB JDCZ KP QHB"E "''BJR#L3-(%/.+%*6 $""6I 55YN$D:C,$D#$E

$CE +"CE *BFFBCCBB 3HD'IJBF\C >KCVDE"' )FP"FE &BBIDFR ?XDIB'DFBCL$""CE

$6E .RVPKL XI .ZFQTGLQ @I 7FQTZP[ 7I -H^PKL 1I -LOKHNRP[ %@I @FKKHNF .M )WVJKODFR EHB `X"'DE# KP

DFP"FE C'BBV EHJKXRH EHB DFG'XCDKF "E CXVVBJ KP GBJB"'C BFJDGHBI QDEH EJ#VEKVH"FS "IBFKCDFBT

U\TVHKCVH"EBS "FI XJDIDFBTU\TVHKCVH"EBL94.+ 9-4+)6*% $"";I 5%N$:$,$#"E

$:E .KLTPSSH 1I )ROFY -I >KRN]FQ 7M 7"JBFE"' AB'DBPC "FI VJ"GEDGBC JBR"JIDFR B"J'# DFEJKIXGEDKF KP

CK'DI PKKIC EK EHBDJ GHD'IJBFL#$%& 3-(%/.+ D38%$/E $""=I )I >F^FQFNZ,&KLTZFYWF >I .LZPQ -I &FQTZRGF XM ="JJDBJC EK

GKWV'D"FGB QDEH DFP"FETPBBIDFR JBGKWWBFI"EDKFC "WKFR 'KQTDFGKWB QKWBFL! ?4= F/*.

$""6I %%N$:,5#E

$;E \LK^FSZ %I )[HPTZTHFK[ &I 7[FcPUYWF \M *HB BPPBGE KP EHDGZBFBITPBBI DFEBJOBFEDKFC KF

R"CEJKBCKVH"RB"' JBP'X] DF DFP"FECN C#CEBW"EDG JBODBQ "FI WBE"T"F"'#CDC KP J"FIKWD_BISGKFEJK''BI EJD"'CL3-(%/.+%*6 $""#I 5%%NPD$6#,D$::E

5"E &-bC "AKXE ="A# &KJWX'"S =KEE'B &BBIDFR "FI )FP"FE 2XEJDEDKFLBLKGO \PFGSZ 9KNFQH[FSHLQa

$"";E

5DE 0)3N &JB`XBFE'# "CZBI XBCEDKFC "AKXE GBJB"'L1FK`GFQO )PJFKS]PQS Lb \PFGSZE

5$E _HPNGPK %,8 77I \RVPK )I \R]]PG 1I @LQHbFTHL 'M +"J'# DFP"FE PBBIDFR "FI JDCZ KP IBOB'KVDFR

E#VB 5 ID"ABEBCT"CCKGD"EBI "XEK"FEDAKIDBCL!1,1 $""5I %M6ND:$D,D:$#E

55E 1T1HGGPQ *.I -FSSFQFSKF` 3I )RbbHPGO X%I 1LKKHYLQ X3I 1FT3FRNZGHQ 71I 8PQSHGH 7I 1RZGZFRYGPK @7M

*HB B"J'# KJDRDFC KP '"EBJ KABCDE#N V"EHQ"#C "FI WBGH"FDCWCL1(5 'G> ,-(


35/47

5C

=$E .FYP & .1I \HQL XM 8F'DFB 2XEJDEDKF +IXG"EDKFN +FH"FGDFR 8VVKJEXFDEDBC PKJ (DWDEBIT,BCKXJGB

(B"JFBJCL!)4+&/$ )I 'G.-&6%)& $"DDI P]PGLK )I 7HQNZ %I -FTP &9I 7LSL BV"EK'KR#S "FI 2XEJDEDKF c2-47?>-2d "FI EHB +XJKVB"F 4KGDBE# PKJ

7BID"EJDG ?"CEJKBFEBJK'KR#S >BV"EK'KR#S "FI 2XEJDEDKF c+47?>-2dL!)4+&/$ )I 3-(%/.+%*

K/6.+)-&.-+)$):H /&( 94.+%.%)& $"";I


36/47

C))%1=3&%*

C))%1=3D


37/47

C))%1=3D


38/47

7TKPPQHQN 0LLG ?JFNP $A


39/47

C))%1=3D


40/47


41/47


42/47

C))%1=3D FE


43/47

*QbFQS *QSFWP 2LK] ?]HOOGP JFNPY, ]RGSHJGP TLJHPY JPK JFTWPSA


44/47

*QbFQS *QSFWP 2LK] ?GFYS JFNPA


45/47

C))%1=3D F


46/47

*QbFQS *QSFWP 2LK] ?]HOOGP JFNPY, ]RGSHJGP TLJHPY JPK JFTWPSA


47/47

*QbFQS *QSFWP 2LK] ?GFYS JFNPA

Documents

Graduate Thesis Proposal_GRAVES_5.2.2013.pdf