Lrsa R. YAwEr, MPH I Drarur M.

BEcrrn, ScD MPH

TnnvN F. Moy, MS RD r Jonl GtrrELSoHN, PHD

DvaNN MarsoN KoprMAN, DnPH MPH CHES

Project Joy: Faith Based

Cardiovascular Health Promotion
for African American Women

SYNOPSIS

O b l e c t i v eT. h e a u t h o r st e s t e dt h e i m p a c to n c a r d i o v a s c u lrai sr kp r o f i l e so f
A f r i c a nA m e r i c a nw o m e na g e s4 0 y e a r sa n d o l d e ra f t e ro n e y e a ro f p a r t r c r -
n a t i o ni n o n e o f t h r e e c h u r c h - b a s endu t r i t i o na n d p h y s i c aal c t i v i t ys t r a t e -
g i e s :a s t a n d a r db e h a v i o r agl r o u p i n t e r v e n t i o nt h, e s t a n d a r di n t e r v e n t r o n
s u p p l e m e n t ewdi t h s p i r r t u aslt r a t e g i eos r, s e l f - h e l spt r a t e g i e s .

. o m e nw e r e s c r e e n ea
M e t h o d sW d t b a s e l i na e n d a f t e ro n e y e a ro f p a r t t c t -
pationT a ti t h i n g r o u pa n d b e t w e e n
. h e a u t h o r sa n a l y z e idn t e n t i o n - t o - t r ew
g r o u p su s i n ga g e n e r a l i z eeds t i m a t i negq u a t i o nas d j u s t m e nf ot r i n t r a - c h u r c h
M s . Y a n e k , D r . B e c k e r ,a n d M s . M o y a r e c l u s t e r i n gB.e c a u ssep i r i t u asl t r a t e g i ewse r e a d d e dt o t h e s t a n d a r d interven-
with the Center for Health Promotion, tion by participanth s e m s e l v e tsh, e r e s u l t s f r o m b o t h a c t i v eg r o u p sw e r e

J o h n s H o p k i n s U n i v e r s i t yS c h o o l o f
M e d i c i n e ,B a l t i m o r e , M a r y l a n d .M s . Y a n e k
a n d M s . M o y a r e R e s e a r c hA s s o c i a t e s ;D r
B e c k e r i s a P r o f e s s o ro f M e d i c i n e .D r .
G i t t e l s o h n i s a n A s s o c i a t eP r o f e s s o r ,
I n t e r n a t i o n a lH e a l t h ,J o h n s H o p k i n s
University BloombergSchoolof Public
r n d ,t h u s ,c o m b i n e df o r c o m p a r i s o nwsi t h t h e s e l f - h e l gp r o u p .
s i m i l aa
R e s u l t sA. t o t a l o f 5 2 9 w o m e nf r o m l 6 c h u r c h eesn r o l l e dI.n t e r v e n t i opna r -
s b o d yw e i g h t( - l . I l b s ) ,w a i s t
t p r o v e m e n ti n
t i c i p a n t se x h i b i t e ds i g n i f i c a ni m
c i r c u m f e r e n c(e- 0 , 6 6i n c h e s )s, y s t o l i cb l o o d p r e s s u r e( - 1 . 6m m H g ) d i e t a r y
e n e r g y( - l l 7 k c a l )d, i e t a r y t o t a l f a( -t Bg ) ,a n ds o d i u mi n t a k e( - 1 4 5m g ) T h e
s e l f - h e lgp r o u pd i d n o t , I n t h e a c t i v ei n t e r v e n t i ognr o u p ,w o m e ni n t h e t o p
A^.il^ f^. '^,oi.hf lncc a t o n e y e a r h a d g v g n l 3 r g e r ,c l i n i c a l l m
y eaningful

Health. Dr. Koffman is a Project Officer c h a n g ei sn r i s ko u t c o m e (s- 1 9 . 8l b s ) .

with the National Center for Chronic C o n c l u s i o n sI n. t e r v e n t i opna r t i c i p a nat sc h i e v ecdl i n r c a lilm
y p o r t a nitm p r o v e -
D i s e a s eP r e v e n t i o na n d H e a l t h mentsin cardiovascular disease riskprofilesone yearafterprograminitiation,
Promotion, Centers for Disease Control w h i c hd i d n o t o c c u ri n t h e s e l f - h e l g p r o u p ,C h u r c h - b a s ei ndt e r v e n t i o ncsa n
a n d P r e v e n t i o n ,A t l a n t a , G e o r g i a . y n e f itth e c a r d i o v a s c uhl aera l t ho f A f d c a nA m e r i c a n
s i g n i f i c a n tbl e women.

PUBLIC HEALTH REPORTS ' 2OOI SUPPLEMENT 1 ' VOLUN4E I16

 rban African American women aged 40
yearsand older bear a marked excessrisk
of obesity,sedentarylifestyle, and atten-
dant cardiovascular disease(CVD) mor-
bidity and mortality. Public health srrare-
gies to lower the risk of lifestyle-relatedcardiovascular
disease in this group have met with little long-term
s u c c e s sI .
In African American communities, the church com-
munity remains the primary source of social support and
community leadership,particularly among older African
American women.2,3 Churches are well-acceptedsitesfor
cancer screening,a,5 blood pressur" .o.rtrol,o, weight loss
programs,6-8 cholesterol education,T,e smoking cessa-
tion,z'to'ttdiabeteseducation,8strokeprevention,12 physi-
cal activity,6'7'r3
and nutrition education.laHowever,eval-
uation of these programsfor long-term results has been
sparse,and health promotion activitiesthat are integrated
into church culture have not been tested previously in
the African American community.
ProjectJoy was designedto addressthe need for well-
evaluated, culturally integrated programs focusing on
lifestyle changein African American women. Participants
in the pilot project named the program from a Bible
verse, "...for the Joy of the Lord is your strength"
(I'{ehemiah8:10b). Project Joy was designedto test sev-
eral strategiesin the church environment to reduce car-
diovascular risk in urban communities where most
African American women are regular churchgoers.l5,r6
The overall objective was to determine the impact of
active nutrition and physical activity interventi,onson
one-yearmeasuresrelating to lifestyle risk factors and
CVD risk profiles compared with a self-help (control)
group. The study was also designed to determine the
extent to which a strong spiritual component and ele-
ments of church culture strengthenthe impact of stan-
dard behavioralgroup interventionsin the church.
MrrHoDS
This study was approved by the Johns Hopkins Joint
Commission on Clinical Investigations,the Institutional
ReviewBoard for the Johns Hopkins University Schoolof
Medicine.
Intervention development. We designed, imple-
mented, and compared three intervention strategiesfor
their impact on cardiovasculardiseaserisk behaviorsand
outcomes.These included a behavioralmodel based on
standard group methods with weekly sessions(SI), the
PUBLIC HEALTH REPORTS.2OOI SUPPLEMENT 1 . VOLUME
same behavioralgroup model supplementedwith a spiri-
tual and church cultural component (SP), and a control
group of non-spiritual, self-help interventions (SH). All
interventions were developed through a partnership
between community members and investigators, and
were basedon a community action and social marketing
model developedoriginally by the Health and Religion
Project of the PawtucketHeart Health Program.rT
We designedinterventions at the individual level to
enhanceself-efficacybut implemented them through the
churches in group sessionsto assurestrong support and
incorporationinto the social milieu of women of this age
group.18Focus groups were held with churchgoing
women from the African American community in Balti-
more to determine what kinds of nutrition or physical
activity interventionswould be most appropriate.Based
on these focus groups and additional in-depth interviews
with 53 churchgoingwomen, we designedthe interven-
tions and tested the questionnairesto assessnutrition,
physical activity,smoking cessation,and operationaland
feasibility aspectsof program implementation.We incor-
poratedinto the interventionsthe themes,suggestedses-
sion formats, and materialselicited and examinedduring
the focus groupsand in-depth interviews.
One church servedas a pilot venue where we tested
and refined the spiritual and church-culture component
intervention over a 20-week period. In addition, we
formed a Community Expert Panel to review and further
refine the interventions and measurements.This group
was comprised of four African American churchgoing
women and two African American pastorsfrom the com-
munity. None of the focus group, in-depth interview, or
pilot church participants, or Community Expert Panel
members, participated in the trial. This intensive com-
munity involvement in the design of the interventions
assuredcultural relevanceof the interventionsand study
protocols and assistedultimately in community "owner-
ship" of resultingprogramsand disseminationof results.
Church recruitment and randomization. Churches
that met eligibility criteria were identified from more
than 700 inner city churchesin Baltimore. Initial criteria
included location in the urban core of Baltimore, a pri-
marily African American congregation,and a known level
of high interest and participation in local activities,such
as revivals,conferences,and religiousevents.Once iden-
tified, churcheswere selectedby the study'spastoralcon-
sultants and recruited from denominationalstrata. Bap-
tists comprise the largest religious denomination in
Baltimore and in African American communities in gen-
116
eral (as high as 62Voin some areasof the southernUS).15
Therefore, we required that 50% of participating
churches be Baptist and the remaining 50% be either
independent or externally governed churches, such as
Roman Catholic, Methodist, or Holiness. Eligible
churchesalso had at least B0% AfricanAmericancongre-
gations,averageSunday attendanceof at least I 50 indi-
viduals, and no currently active program in weight con-
trol, exercise,or smoking cessationfor women aged 40
years or older. Becauseexact data on these variablesare
rarely formally maintained by a church, Project Joy staff
and pastoral consultants determined eligibility through
discussionswith pastorsand direct observationof Sunday
church servicesprior to recruitment.
Once a church was determined to be eliglble, we sent
information on the proposed study to the pastor of the
church. The principal investigator and project staff then
met with the pastor and any key lay leadersto describethe
project in detail, including the three interventions.To main-
tain the desired ratio of church denominations,envelopes
were prepared for each denominational stratum to ensure
appropriate representationin each of the three interven-
tions. For the first 23 churches, if the pastor agreedto the
project, he or she randomly selectedan envelopeto deter-
mine which intervention the church would offer. After
selection of the intervention, the pastor and staff signed a
cwenant (a term more meaningful and acceptableto pas-
tors than agreement)indicating their commitment to sup-
port Project Joy.Churches were enrolled sequentiallyuntil
the desiredsamplesizeof 490 women was obtained.
We found that the first 23 pastorsalmost universally
indicated a discomfort with the randomizationprocess.
They consistentlywanted to be told at the beginning of
the meeting to which intervention their church was
assigned,even though they understood assignmentwas
random. After realizing that the randomizationprotocol
seriouslyhinderedenrollment,we requestedand received
formal permissionfrom the Institutional ReviewBoard at
Johns Hopkins to pre-randomizethe church assignment
and presentthe study to the pastorwith full disclosureof
the pre-randomizationand descriptionof the other inter-
ventions. Thus, for the last I 5 churches, the staff
describedthe project, all intervention contingencies,and
the specific intervention to which that pastor'schurch
had been randomly assigned.If the pastor agreed, the
covenantwas signed.
Participant recruitment. Lay leaderswere designated
to assist with the project by the pastor of each church
after the first meeting. They were all African American
females, usually known "influence leaders" who were
PUBLIC
highly respectedin the church, and often included the
pastor'swife. The number per church varied by church
sizeand preferenceof the pastorsand church staffs.Pro-
ject staff met with these leaders to design participant
recruitment strategiesspecific to the proceduresof each
church. After a one-month pubhcity phase during which
church bulletin inserts advertisedthe program and fea-
tured tear-off forms with instructionsfor potential partic-
ipants to give to the designatedlay leaders,each church
scheduledat least one Recruitment Sunday.Posterswere
displayedin the church and announcementswere made
from the pulpit by the pastor and lay leadersthroughout
the month for all three interventions.Choir rehearsals,
Bible study groups, and other women's groups also
received announcements. On Recruitment Sunday,
announcementswere made from the pulpit by both the
pastor and the project staff, usually including the study's
principal investigator,inviting all women aged 40 years
and older to attend a recruitment meeting immediately
following each service,where the intervention was pre-
sented in greater detail. Each interested woman com-
pleted an eligibility form at that time, and signed up for
screeningappointments.Women who did not schedule
an appointment at the recruitment meeting, or who later
heard about the project by word of mouth, were called
later by project staff.All women who scheduleda screen-
ing appointment receiveda reminder card, map, instruc-
tions, and parking information.
Eligible participants were ages40 years or older, were
not pregnant or planning to become pregnant in the com-
ing year, had not had a myocardial infarction or stroke in
the past six months, had not felt any chest pain or angina
requiring the use of nitroglycerine in the past sir months,
did not have cancer currently under treatment, were not
undergoing renal dialysis,and were able to obtain permis-
sion to participate in the program from their physicians.
Women who did not have a physician were referred to a
federally funded communify health center when necessary.
Despite aggressiveand persistent recruitment and the
offers of retreats and free risk factor screening,most
women from churches randomizedto the self:help contin-
gency were not interested in the self-help intervention.
Women were avidly interested in active sessionsand
expressed a strongdesireto receivedirectedassistance'For
this reason,the numbersrecruitedfor the self:helpcontin-
gency were lower. During recruitment in each church,
even the numbers of women who merely stayed for the
recruitment talk were much lower than fbr the two active
contingencies.We attempted to increasenumbers in the
self-help comparison group by recruiting other church-
going African American women through advertisements;
HEALTH REPORTS ' 2OOI SUPPLEMENT 1 ' VOLUME ] I6
this attempt addeda small number (16) of women who are for support and information, for the remain-
included in the self-help group referencecomparisons. der of the year.

Standard bdhatioral intervention. Standard intervention
churches held weekly sessionson nutrition and physical
activity in their own facilities. FemaleAfrican American
health educatorsfrom the study staff taught the curricu-
lum, standardizedfor the first 20 weeks of sessions(Fig-
ure), with the assistanceof church lay leaders.Lay lead-
ers were self-identified or assignedby the pastor and
completed four hours of formal training by staff health
educatorson nutrition and fitness.
Each intervention sessionbeganwith a weigh-in and
group discussion,followed by a 30- to 45-minute nutri-
tion education module that included a taste rest or
cooking demonstration. The sessions,based on social
learning theory were designed to enhance individual
self-efficacy.lBEach session included 30 minutes of
moderate intensity aerobic activity, the nature of which
varied by church; physical activities included brisk
walking, water aerobics, or Tae Bo (Tae Kwan Do-
dance-boxing). After the first 20 weeks, Iay leaders
offered weekly sessions,with health educatorsavailable
Spiritual intervention. Churches offering the spiritual
intervention received the same sessionsas the standard
interventionchurches,with the addition of spiritual com-
ponents and church contextualcomponentsdesignedby
the Community Expert Panel and investigators. All
weekly sessionsincorporated group prayers and health
messagesenriched with scripture (see Figure). Physical
activitiesincluded aerobicsto gospelmusic or praiseand
worship dance.Telephonecalls from lay leadersand word
of mouth from other participants motivated attendance.
Church bulletins included weekly sessionreminders and
printed messagesfrom Project Joy, called the Joy of
Heahk, on healthy eating and physical activity,accompa-
nied by salient scriptures. The pastors offered regular
information on healthy eating and physical activity from
tip sheets supplied by Project Joy and distributed a
monthly health newsletter, called Fronctlce Pastor'sDesk,
to the congregation.Churches alsoparticipatedin at least
one eventper year sponsoredby ProjectJoy,such as walk-
a-thons,faith and worship dancerecitals,or fruit sales,all

Figure. Topics discussed in first 20 weekly sessions, standard and spiritual interventions, Project Joy,
Baltimore, 1997-1999

Session Nutrition topic Exercisetopic

I The Food Guide |ramid Short-Term Benefitsof PhysicalActivity

2 Fat Counting FitnessWalking
3 Portion Size Proper Footwear and Clothing for Exercise
4 Food Labeling ExerciseBuddies
5 EnergyBalance Principlesof Exercise,Flexibility
6 Fruits Heart Rate
7 Why We Eat Exerciselntensity
I Vegetables-Benefirs Health Benefits of PhysicalActivity
9 Vegetables-Preparation Realistic Expectations from Exercise
l0 Fats/Three-MonthReview Hot Weather Exercising
II Meats Safety lssueswith Exercise
12 Meat Alternatives PhysicalActivity That Can Become Exercise
13 Grains and Fiber Other Aerobic Exercise
14 Dairy Foods Distraction from Exercise
15 Salt/Sodium Minor lnjuries
16 Shop'Til You Drop (the Fat) ExerciseCues and Prompts
17 Dining Out ExerciseInterruptions
18 Breakfast,Lunch,and Snacks Long-Term Maintenanceof Exercise
19 Holiday Eating Long-Term Maintenanceof Exercise(continued)
20 Six-Month Evaluation Vacation ldeas

PUBLIC HEALTH REPORTS.2OOI SUPPLEMENT I . VOLUME 116

We believe it is not possible to maintain a non-:piritual
intervention within the African Am e rican church
environment.
activities that exposed other church members to the
health activitiesof ProjectJoy.
It should be noted that while the standardintervention
was originally designedwithout spiritual elements, partici-
pating women in all Sl-designated churches introduced
spirituality into their sessionsfrom the beginning, without
staff assistance. Just as the pastorsand lay leadersunder-
stood randomization but believed that in a world created
and maintainedby God nothing is random, the participating
women did not believe there could be any church-based
program that was not spiritual. They initiated sessionswith
prayer and selectedtheir own relevant scriptures.Standard
intervention participants also sought ways to include the
entire church and the pastor. Ultimately, the standardand
spiritual interventionsoperatedalmost identically.
Self-help comtrolimtervemtiom. The self-help inteivention
was included as a control or reference contlngency.It
included materialsfrom the American Heart Association
on healthy eating and physical activity and information
targeted to the participants'personal screening results.
For example,pamphletsfrom the American Heart Associ-
ation on smoking cessationor cholesterolwere used for
smokersor women with elevatedblood cholesterol.Each
participant receiveda gift-wrappedbox with her name on
it containing feedback about her personal screening
results, a place to list her personal goals for the year,
materials to allow self-monitoring,and the Project Joy,
NIH, and YMCA educational materials and pamphlets.
The behavioralgoalsfor the standardand spiritual inter-
vention groups were also included in the materials and
were offered during the self-helpretreat. In addition, the
lay leadersfbr SH-designatedchurchesreceivedthe same
lay leader manual as the SI lay leaders,which contained
not only content information and handouts for sessions
but instructionson how to run sessionsif churcheschose
to implement this part of the intervention. No further
help was offered directly but a hot line number was avail-
able for consultation from the professionalProject Joy
healtheducators.
PUBLIC
Baseline screening. A11eligible participantsin all inter-
ventions completed a baseline health assessment
between March 1997 and March 1999 at the project
offices, a non-clinical site. Before taking any baseline
screeningmeasurements,staff expiainedthe programand
screening procedures to all participants, and obtained
informed consentfor their participation.
D emograp'hics and medical'history. Standardized question-
naire items included employment status,years of educa-
tion completed, household income, and current marltal
status.The project nurse reviewedeach participant'sself-
reported medical history and probed for details on hyper-
tension, hlpercholesterolemia,diabetes,angina,myocar-
dial infarction, congestiveheart failure, cancer,arthritis,
or other serious health problems. Current medicatlons
were verified by examinationof the participant'smedica-
tion containersbrought to the first screeningvisit. Each
participant's physician received a standardizedform
requestingmedical clearancefor participation.
Anthropom,etrics.Body weight was measuredusing a cali-
brated digital scale (SECA Alpha Model 770) with the
participant wearing light indoor clothing and no shoes.
Height was measuredwith a set squareagainsta straight
wall. Body mass index (BMI) was calculated as weight
(kg) divided by height (m)2.Bioelectricalimpedancewas
used to assesspercent total body fat using appropriate
formulae,lebased on resistanceand reactancemeasure-
ments (RJL Systems,BIA-101Q analyzer,Clinton, MI).
Waist circumference was measured using guidelines of
the National Obesity Expert PanelReport.l
Blood presswreand.'l'teartrate. Blood pressurewas mea-
sured with a mercury sphygmomanometeraccording to
American Heart Association guidelines.20 All partici-
pants were seated quietly for at least five minutes prior
to measurement, and had not ingested caffeine or
smoked for at least 30 minutes prior to measurement.
The averageof three readingstaken at least 30 minutes
HEALTH REPORTS. 2OO1 SUPPLEMEN,I 1 . VOLUME 116
apart was used to characterizeblood pressure.Current
hypertension was defined as having an averageblood
pressure>\40/90 mmHg or taking current antihlper-
tensive pharmacotherapy.
Blood lipid levelsand glucose.Blood was obtained after
the participants had fasted for l2 hours overnight.Total
cholesterol,high-density lipoprotein (HDL) cholesrerol,
triglycerides (TG), and glucose were measured directly
from plasma by the CDC-standardizedJohns Hopkins
Chemistry Laboratory.The coefficient of variation in this
laboratoryfor total cholesterolmeasurementis less than
2%. Low-density lipoprotein (LDL) cholesterol was
calculated using the Friedewaldformula,2l as no partici-
pants had TG >400 mg/dl. Diabetes was defined as
having glucose >126 m{dl or receiving current hlper-
glycemic (oral or insulin) therapy.
Dietary nwtrient intake. The Block Food Questionnaire,a
food frequencyinstrument (1995 scannableversion),was
administered to participants by a registereddietitian or
researchstaff trained and superwisedby the dietitian. All
questionnaireswere sent to Block Dietary Data Systems
in Berkeley,California,for scanningand analysis,yielding
individual nutrient intake measuresof energlr,total fat,
sodium, and percent of energyfrom total fat.
Smoking amd.carbomwomoxide.Smoking status was self-
reported and verified by measured exhaled carbon
monoxide (CO) using a Vitalograph EC50 CO monitor.
The highest number that appearedon the monitor after
exhalationinto the monitor was recorded.If a participant
was a stated non-smoker and the CO was measured
>8 pp-, the test was repeated. If both measureswere
>8 pp- and the participant denied smoking,the partici-
pant was classifiedas a smoker.
Physicalactittity.Physicalactivity was assessedusing the
YalePhysicalActivity Survey,22 from which energyexpen-
diture was calculated.
Retreats. Separateretreats for each church (16) were
held after all participantsfrom a given church had com-
pleted baseline screeningand before the formal weekly
sessionswere initiated in the active intervention (SP and
SI) churches. For the standardand spiritual intervention
groups, the retreat included one 3-hour Friday evening
event and a full day Saturday,while in the self-helpinter-
vention churches, the retreat included only a full Satur-
day.The retreats servedas a kick-off for the program with
PUBLTC HEALTH REPORTS . 2001 SUppLEMENT 1 .
initial information and demonstrationson nutrition and
physical activity,and as an incentive for participation,as
this was an especiallysocial event for all women partici-
pating in the project in a given church. Retreats were
held at a local hotel and included: (a) motivational ses-
sions; (b) introductory nutrition education sessions;(c)
physical activity sessions,such as water aerobics,weight
training, line dancing, and walking; and (d.)an informa-
tional sessionin which participantsreceivedtheir individ-
ual results from the baseline screeningand a registered
nurse discussedCVD risk factors with the group. With
the participant'spermission,a copy of the samescreening
results was also mailed to each participant'sphysician.
Nicotine patcheswere provided to smokersfor the dura-
tion of the retreat so that all smokerswould be able to
remain smoke-freeduring the retreat.
Self-help church groups received the same materials
and were oriented to the samegoalsas membersof spiri-
tual and standard church groups. All participants
receivedthe sameinformation, alongwith personalboxes
containing general risk factor information and informa-
tion specific to each participant, as well as a leader's
guide to allow all women to conduct sessionsin their own
churches. The spiritual retreats had sessionsthat were
supplementedwith scripturewhile the standardand self-
help retreatsdid not include any scriptures.Interventions
began in each church within one month after their
retreat.
One year follow-up screening. After one year of inter-
vention sessions,we recruited participants to repeat all
measurementstaken at baseline.Participantsunwiliing to
return to the project offices were offered home or work-
site visits, or church-sitefollow-up screeningsinclusiveof
only the biologicaloutcome measures.Incentivesto com-
plete follow-up were devisedby the pastoralconsultants
and Community Expert Panel.These included tickets to
a gospel play, a bus trip to a nearby outlet shopping, or
gift certiflcatesto local establishments,and were offered
from the beginning of follow-up screening.Pastorsfrom
participatingchurcheswrote letters to their own congre-
gants encouraging completion of follow-up screening,
and all pastorsannouncedfollow-up screeningfrom the
pulpit. Announcementswere alsoplaced in each church's
bulletin. The only additional recruiting efforts employed
were additionalpastors'announcementsand staff and lay
leader telephone calls. Participantswho completed fol-
low-up screeningreceived a mailed letter describing all
baselineand follow-up results; a copy of this letter was
also mailed to each participant'sphysician.
VOLUME tl6
Intervention behavioral objectives. Goal behaviors
included exerciseof 30 minutes or more, 5 to 7 daysper
week; consumptionof at least 5 servingsof fruits and veg-
etableseveryday; fiber consumption of at least 25 glday;
fat consumptionof 40 gldayor less;consumptionof 1200
to 1800 dietary calories per day; and dietary sodium
intake of 2400 mglday or less. Smoking cessationwas
also a goal. Weight management was emphasizedbut
achievementof a specific "ideal" weight was not empha-
sized.Women were encouragedto optimize dietary and
physical activity and to achieve "reasonable"weight, or
that level of body weight that they reached when they
were meeting all of the abovebehavioralgoals.
Statistics. We conducted all analyseswith SAS soft-
ware, version 7.0, using the intention-to-treatapproach;
participants who did not complete follow-up measures
retained the same measurementsat follow-up that they
had during baseline assessment.We calculated simple
frequenciesand means,and used the chi squaredistribu-
tion for analysesof variance and contingency tables to
compare baseline and follow-up measures within and
between groups.We calculated changesin measuresby
subtracting the baseline measure from the follow-up
measure.We used paired t-tests and analysisof variance
to evaluatechange in continuous variables,and McNe-
mar'stest to evaluatechangesin categoricalvariables.We
performedmultiple linear and logistic regressionanalyses
using the Generalized Estimating Equations (GEE)'?3
approachto account for potential within-church cluster-
ing effects.
RssuLrs
Of the first 55 churchesidentified by our pastoralconsul-
tants and community experts,43 met all eligibilitycriteria.
After introductory letters and packets were mailed to their
pastors,38 (88.4%) scheduledpreliminary meetingsand
18 (47.3%) enrolled. Of the 20 churches that did not
enroll, 5 refused to participate during the period before
pre-randomization;2were eliminatedbecauseno partici-
pants enrolled (both were self-help churches); 13 were
eliminated becausethe church calendarwas too full to
allow expedientrecruitmentinitiation (within sir months);
and 2 were eliminatedby ProjectJoy for other reasons.Of
the 16 churchesenrolled,8 were Baptist (50%),3 werc
independent(that is, Holiness),and 5 were externallygov-
erned (Roman Catholic, United Methodist, and AME).
Four churches offered the spiritual intervention, 5
churches offered the standard intervention. and the
PUBLIC
remaining 7 churches offered self-help and included the
l6 women recruitedthrough advertising.
Of the 966 potential participantsself-identifiedat the
recruitment meetings,from the newsletter,or by word of
mouth, 920 (95Vo)women were eligible. Of the eligible
women, 702 (76%) scheduled appointments, and 559
(61%) werc screened.Of those who completed baseline
screening,529 (58%)women enrolled.There were no dif-
ferencesin schedulingor enrollment ratesbetween spiri-
tual and standardinterventionparticipants;however,self-
help participantswere significantlylesslikely to schedule
appointments (P = 0.001), to keep appointments once
scheduled (P = 0.001), or to enroll in the program (P =
0.001).Of thosewomen who enrolled,50.5Vo (267) were
in the spiritual group, 35.5% (188) were in the standard
group, and only 14.0%(74) were in the self-helpgroup.
Baseline characteristics are shown in Table L Most
participants had completed high school, three out of four
were employed, sllghdy more than half had hlpertension,
and more than two out of five had arthritis. Demographic
variableswere generallysimilar among intervention groups.
Standardintervention participants attended an aver-
age of 6.4 (standard deviation +6) professionalhealth
educator-ledsessions,while spiritual interventionpartici-
pants attendedan averageof 7.0 (standarddeviation+ 6,
P = 0.235). Attendance ranged from 65% at the first ses-
sion to 26.1%at the last sessionacrossall churches,and
on any given sessionattendancewas one-third to one-half
of overallparticipants.This varied slightly by church, but
not by spiritual vs. standard intervention. None of the
self-helpchurchesheld any sessions.
Fifty-sixpercent (294) of participantscompletedone-
year follow-up biological measuresand of these, 67.7%
( 199) completedall follow-up measures,including behav-
ioral outcomes in diet and physical activity. Participants
who completedfollow-up were significantlyolder (54.7 +.
9 versus 50.9 + 9, P = 0.0001), had lower incomes
( $ 3 6 , 6 2 8t $ 2 1 , 1 0 0 v e r s u s$ 3 8 , 9 0 3 t $ 2 4 , 0 0 0 ,P =
0.3052), were less likely to be currently employed(79Vo
versus 67Vo,P = 0.0036), and attended more sessions
(7.8 + 6 versus3.4 t 4, P = 0.0001)than thosewho did
not complete follow-up. Spiritual (159/267, 59.6Vo)and
standard( 106/188, 56.4%)interventionparticipantswere
significantly(P = 0.0074) more likely to return for follow-
up than self-help(29174, 39.2Vo)participants.
Changes in risk factors and behaviors. AII follow-up
and change measurementswere based on intention-to-
treat analyses.As noted, there were no real differences
between the spiritual and standard interventions, so
HEALTH REPORTS . 2OO1 SUPPLEMENT 1 ' VOLUME I I6

resultspresentedhere show both SI and SP groupscom-
bined and comparedwith the self-help group (Table 2).
Within the active intervention groups, there was a statisti-
cally 5ignifi.unt change in a favorabie direction for I I of
the 13 cardiovascular risk factor outcomemeasures,and a
near significant change for energy expenditure, although
the magnitudeof the changeswas modest.In the control
group, there was a statistically significant change in only
one of 13 outcomes:percent of energyof fat. Further,the
magnitude of any other changeswas far smaller than in
the intervention group, with a gain in weight over the year.
When only the women who returned for follow-up were
used in the analyses;the magnitudeof the changesin all
variablesin the active intervention groups was higher than
when intention-to-treatanalyseswere used. The self-help
group changeswere even smaller than observed,with an
averageweight gain over the year of 2.1 pounds (standard
deviationt 1.3),more than twice that in the intention-to-
treat analyses.
The group in the top decile for weight loss (average20
pounds at one year) also achievedclinically meaningful,
highly favorable changesin most other risk behaviorsand
outcomes;these changes,however,occurred only in the
PUBLIC HEALTH REPORTS . 2OO1 SUPPLEMENT 1 . VOLUME
::: :
rrai:r::l
:,i::r:;i
combined activeinterventiongroups(Table3). In the top
decileof weight lossin the self-helpgroup,changesin risk
behaviorsand outcomes were more modest, with an aver-
ageweight loss at one year of only 7 pounds.All analyses
of other changes within this group were underpowered,
but Table 3 demonstratesthe notable clinical differences
between the self-help and active intervention groups.
The prevalenceof achieving the behavioralgoals of
the intervention significantly improved from baseline to
follow-up in the combined intervention group (Table a),
particularly in the top decile for weight loss where the
increasein the proportion meeting behavioralgoals mir-
rored the positivebiologicalchangesobservedon Table 3.
The proportion of the self-helpgroup meeting behavioral
goals at follow-up decreasedor stayed exactly the same
(for consumption of fruits and vegetablesand dietary
fiber consumption),or increasedby an averageof only 7%
to 2% for meeting any of the goals.The greatestchange
among self-help participantswas an absoluteincreaseof
5.4% of participants who consumed <2.4 grams of
sodium per day.
Multiple GEE church-adjustedlogistic regression
predicting the "most successful"participants,those in the
I I6 75
 ,.';.,..'',,,,"''.',',i',','.'.,.;

-.
b4c9qe. ."' Mean 'deviotioi Mean
'meon
i'
.P "' t4gon": dei.
:i Qytcome !"e; . lt*o', i
.:
tVeight(pounds). ..-.. . . l191.6 e ! : 6 tt45 4s --!,1.
! : l t0.42.
t 0 . 4 2 . 0.0089 J 8 - 4 ::..*,!.1
o . 0 o 8 t J8-4:9, :e*: .; ' 0y83,,,t.0.5,?.. .0.'t!42. Q,SF,,
': YB.M!,qg/T:);.;.,.;:,..
deh..{p:unds).."....
t2:6. t.7 :.:-Q..17 0.p.14s ' $1,7:."i:.t:8. ,,q.i1;1.10.0e 0.'1157: .:
: .t9;tiZ , 9.W12
,' r?J, i.7: :': -0.66:to"l! 99ry1. lg'1,t +g,. .e72? o.gQa.I
,.WgI(incheg.
gddyyhhtt::((ppeer cr :ce::......,
en nt )t.). . .: .4.: .".11. ::001... 66... 4 . 3377 . f ' 0O. .1l ! : 0 ..@ ,.;{.@7 tga0 .,
' qg @0066.: :.:.. ffrr..//..: , J! ,::77' .", . ..--99..1111: :. . .t 0t 0: :{: 0{00.: ??
. 441199..: : :99. 1., 1
'8888
66
j " "'
, SBft(F.rmHg).
Ti:tf'T9. j........ '119i?9 .!1
135.0't20 -1.6 t9'f1 '19!::
t0.5,1 0.0037 136.0.,...*30: .;,,.;0.95. tl'l- 0.3812 013l:1
- 0 . 3 6 t 0 . 2 e 9Tl1 ,*19, " :" ,r,9:ll'
, . 9 . 2 2 . . , * 9 , 6:?0.. 7 ? s e?'1.qT,,
t:t,1. 9,j!9a
;ojf ( g r m u s . )
: . : r . . . . 82.1 il0 q 2 l l p , s 2 , 1 , ,* . 1 . 1 ..0.1.?91,,
(mi/dl). . .: . .. 125.8 t 3s -1..8 t0.92 0.0370 .'..132.9:.i: t:37:'..::;:;"0.1-! +:!.6:. ,9.:?a|50:1S-6ii:
,'LQL"-C r.
UQt--C'(mSldl) . .:... . . 58.8 ! 16 0.13 10.33 0.7019 58,2;;,.;t:1.{,.:;.:,.0"92 ,19.7Q .:.9:!.9L1.,.9"3q3.! ::
.,erier,gtintat<e . " . . , . i , , ' , " . ,: .' : i . . j,
" . , . ' .; ,,.,.
, - i r " i i l C i 1 1 . r . . , , . . . . . . 2 0 7 7t 6 9 s -l17 t16 O . O W J2, ,2 5 : 1 . ' . . . : t : ;8.6; 78::.,: ; { 1 2 : : :O . 8 1 9 0 . : 9 . 0 0 3 8 , . , , ,
. T o t 4 l { a( gt / d a y: .). . . . . 91.8 t3e € . 1 t 0 . e 9 0 . m 0 1. l 0 l : 3 . . . . : 4 , 4 5 ., :.t?: . 4; + , 1 . 0 : , , "
' . . . . . :" g , i J Z j ; ; $ Q 2 s 0 , . . :
j,
...Ener'$rfromfat : .... ..: ,,...,.i.. .., , . . . . . . . . 1. . . ' , , ' . , , ' . '
: i (pgrreqt) '' '':''''--.

26szt1008 -l4s t2s

i:" -' , i': :..'.'.:::..

::lodrl{'n{*1,..:.:. 0So.',,,tt13:,,,.,1
1?r.2 ,.,:.f,:,J:91..,,,9911,1
::t.12.12.''.:.€..;'f.92.' : 0.811:3:
, lneiga &pelrditure
' . - . . - ; . . ' ' . ; . - - -::-::-:-::-. : . : . . . ' ,.9:9!97.,-,
;...:; ::: t.t.t:.,: .:
q * r i - , ,J7? ::::::: ',i,,-j:j:.:,.,,r:,i.,,-.:r,:1,.,::r,::::r:a.r::r:::::::i::i::r::i::i::i:::i:i:i::i;i:::r::r::r:1..!,:::..':-
:'

,:. ..::(kJ/day) s e r !47"3

; i # i /.; i l - . . . . . . . . . . . . set lo.^ 4 8 . .t,?t.,...
38 r . ? ,r 0,0614. ",",;{fle,
,!9..} {1i: , 9.5314::,0{l?.2
:'
::
::::;l-:-
'
:
;:iir*l*t::lil:r::ie:nnr;1.::ry
:r:*if ,
ii^;...
NOIES: *-,Signmcanceof within.group comparisons, P for intention-to.-rret stq.d.gqt's;pai4edlt*e$t$compaflng baselinpto one-.yearfrollow-up . ;.;:
.
mgTyf.unt.Significance of be.ryveengroup comparisons, Pfor ingendon-toirrs+AflO-V.+pf!-a-aiu$.e-d.fc':.!!.u-lSf.:F"*,:.gTl4lnS.ht9r.. ., ..
." .

' )Fl= '..

;.,oei=ci.teticbtood.pressure :. , , ; . , , : , .. i . . , ; . .; . , ..;. .. ,; , .:,' . ' i . . . . . ' . " . . . ; . ; . ,. . . ' .
. : :.:.:'.i.' .:
;'LbL,C.=low{ensitylifoproteincholesterol .. . i:: :i:' .'..'i.::.'.i.:.:.:'i i
hign-densiE lipoprotein cholesterol
,, .HD-I+-C,=

top decile for weight loss,showedthat the higher number improved anthropometric measures,blood pressurelev-
of sessionsattended(P <0.0001), older age (P = 0.03), els, diet and, to a lesser extent, physical activity at one
higher baselineBMI (P <0.0001), and inclusion in the year in the active intervention groups,although the mag-
active intervention groups (P = 0.0014), were all signifi- nitude of the effect was modest. Our overallfindings are
cant independent predictors,while having diabetes(P = generally similar to previous church-basedweight loss
0.24) and hypertension(P = 0.39), and religiousdenomi- and dietary interventionprograms.One grouplaincreased
nation (P = 0.73),were not. the proportion of rural African American church mem-
bers in North Carolina consumingat least five fruits and
DrscussloN vegetablesper day from 23% to 33%. The PATHWAYS
studyBshoweda weight lossof 10.0pounds and waist cir-
Effect of intervention. This study demonstratesthe cumferencedecrementof 2.5 inches after l4 weeks of a
potentially important behavioraland biologicalrisk modi- church-basedintervention,comparedto a weight gain of
fication effect of activechurch-basedgroup interventions 1.9 pounds and waist circumference decrement of 0.4
for African American women. We observedsignificantly inches in a control group. Lighten Up, a church-based

PUBLIC HEALTH REPORTS . 2OO1 SUPPLEMENT I . VOLUME II6

,
lifestyle program,2ashowed an average2.3-pound weight
loss after 10 weeks.Another study6showed a 6.0 pound
weight loss at eight months, six months after an S-week
weight control program of weekly diet and exerciseses-
sions, and yet another25found a 4.2-pound weight loss
one year after an B-week church site program directed
explicitly at weight control.
There are some important differences between our
study and those of others. Follow-up for our study was
longerthan most, occurringone year after programinitia-
tion and 32 weeks after our health educatorswithdrew
from the church. Further, most other studies have not
focused on globally healthy lifestyles, but on a single
behavioralgoal, such as fruit and vegetableconsumption
or weight loss.The effect of our program overall,as well
as the particularly large sustainedimpact on changesin
multiple behaviorsand biological risk factors in a small
subgroup, is encouragingand supports the findings of
other shorter,more focal studies.
It is unlikely that the observedeffect in this study
representsartifact or regressionto the mean, as the self-
help control group did not show benefits of this magni-
PUBLIC HEALTH REPORTS.200t SUppLEMENT I .
tude. Although the number of participants in the self-
help control group was much smallerthan we anticipated
or than would have been ideal, we did find statistically
significant differences between the active intervention
group and the self-help control group for most outcome
measures.Further, the absolutenumber of biologicaland
behavioralrisk factors that changedwas much greaterin
the activeinterventiongroup than in the self-helpcontrol
group. The fact that the number of sessionsattendedwas
strongly related to a beneficial outcome probably indi-
catesthat the most motivatedwomen achievedthe great-
est benefits. The active interventions remained signifi-
cantly and independentlymore likely to result in women
achievingthe greatestclinical benefit.
More importantly, l}Vo of participants in active
church-basedinterventionsachievedhighly clinically sig-
nificant improvementsin CVD risk profiles one year after
program initiation. The change in CVD risk in this top
decile is of greater magnitude than has been shown in
most prior studies in this hlgh-rlsk population, and of a
magnitude that is notable given the large group nature of
the interventions.The fact that changesof this magni-
VOLUME ll6
,i
tude were existent in even l\Vo one year after program
initiation is also important, as most studies find short-
term effects without enduring sustainableimpact.
Implementation issues. Prior studies have shown that
churches are exceilent sites for accessing people for
community-basedhealth promotion programs.2'4-r0'r3'14churches was that the materials were not spiritual
The current study confirms the interest of churchgoing
women in active healthy lifestyle programs offered in the
church, with a particularly strong interest in a spiritually sessionsand added gospel music to their exerciseses-
based program. As noted, there were no differences in
outcomes between the spiritual and standard interven-
tion churches, despite the fact that all materials in the
standard intervention were originally designed without
spiritual elements. No attempt was made to include the
entire church in the standard behavioral interventions,
yet in the caseof every church, the participatingwomen
incorporateda spiritual element from the first sessionof
the retreat. Participants invented ingenious ways to
involve the entire church and the pastor, initiated ses-
sions with prayer,and sought relevant scripturesof their
own selection. Some standard intervention churches
even designateda lay leader to seek and integratescrip-
tures. Even in the spiritual interventionchurches,women
incorporated more spirituality than was originally
included in the session materials and protocols. Ulti-
mately,the spiritual and standardinterventionsoperated
almost identically,so it is not surprisingthey did not show Women simply wanted the active intervention sessionsat
any differencesin outcomes.
Most women were unaware that Project Joy differed
in fbrm among the various churches in Baltimore, and
PUBLIC
churchesgenerallywere not in contact with one another.
The first church started months before any others and
was a standard intervention church. Spirituality was
incorporatedby the women from the very first group ses-
sion. The chief programmaticconcern expressedin post-
intervention assessmentsin the standard intervention
enough. In everychurch at everyjuncture, women incor-
porated prayer and scripture into their weekly didactic
sions. Having had this experiencewith severalhundred
women, we believe it is not possibleto maintain a non-
spiritual interventionwithin the African American church
environment. Finally, the degree to which a pastor was
involved was not dependenton the role assignedto him
or heq but was highly individualized.Even in many stan-
dard intervention churches,the pastor chose to incorpo-
rate the activitiesof Project Joy into the Sunday services
and into the consciousness of the church.
strategies.One important finding of
Acthte versussel.f-kel.p
this study is the almost total lack of interest in a self-help
program.At each level of our recruitment process,pas-
tors. churches. and women were least interested in the
self-help intervention despite the fact that they received
free screeningwith feedback,individualizedintervention
materials, and a retreat to promote healthy lifestyle.
their churches. Even when the pastor agreedto partici-
pate in the self-help contingency,aggressiveand persis-
tent efforts to recruit individual women in church ser-
HEALTH REPORTS . 2OOI SUPPLEMEN'I I . VOLUN,IE ! I6
vices and by word of mouth were only modestlysuccess-
ful. Women consistentlyexpressedan interest in having
weekly sessionsoperatedby knowledgeableprofessional
leaders.They felt their peers were not qualified to lead
interventiongroups.
This finding appearsto contradict previous work sug-
gesting that volunteers are sufficient to run church-based
health promotion programs.26Process data from atten-
dance logs showed consistently that churches had lower
weekly attendance after the professionalhealth educators
stoppedleading the weekly sessions,in spite of the transfer
to trained lay leaders.In our post-study focus groups and
in-depth interviews,women indicated that this decline was
due to a lack of confidence in the capabilities of peers
whom they did not believe had the same expertiseas the
professionalhealth educators.It was difficult for them also
to shift their peer relationships with these women and to
accept them as leaders, as they had all attended the ses-
sions together. It is possible, however, that had the inter-
ventions begun with trained lay volunteers,without partici-
pants ever being exposed to the professional staff, that
acceptance of volunteer lay leaders would have been
greater.At the pastoral level, pastorswere more interested
in expert-ledprogramsfor their churches, also, becauseof
greater confidence in professional support staff. Thus,
while the study was designed to examine self-help as an
intervention, our findings suggest,as have others, that edu-
cation by itself is not sufficient to engagewomen in work-
ing on healthy lifestyles, although there was some modest
benefit accrued by women in this group.
Randomization. The unwillingness of pastors to take
part in the randomization process was surprising, and
one that may be embedded in the ethos of the church
environment and specific individuals. While pre-ran-
domizationin the later recruited churches was not ideal
from our research perspective, it was more acceptable
to the pastorsand yielded a lower refusal rate. Previous
studies had shown churches to welcome the randomiza,
tion process;7however,they were not in the context of
offering a spiritual versus non-spiritual intervention,
which almost certainly affected the willingness to be
randomized. Pastorsdid not wish to be seen as having
"determined" which intervention their churches would
receive.This is basedon their personal belief that "God
guides their hands." Once informed of the spiritual
intervention, they all preferred to have it and did not
wish to participate without it. However, they were far
more accepting of a non-spiritual intervention if their
churches were randomly assigned to a program in
PUBLIC HEALTH REPORTS . 2OOI SUPPLEMENT I .
advance,stating their ability to attribute the pre-assign-
ment to divine providence.
Attendance. Most women attended the retreat and were
very excited to begin the program at that point. All
churches began weekly intervention sessionswithin one
month after the retreat, usually within one week, so we
believe a time gap was not the reason for the drop in
attendance. Process data suggestthat women stopped
coming becausethey did not see immediateresults,simi-
lar to the reasons they gavefor not retuming for follow-up
screening.Telephone calls from staff health educatorsand
social support from the pastor and church community was
clearly not enough to motivate attendance.This should
not be a surprise, as it is found in every other behavioral
interventionreportedin the literature,especiallyphysical
activity interventions. Our participants have suggested
offering a choice of multiple days per week, offering ses-
sions at night, and opening the program to the whole fam-
ily. This was beyond what we had funding to do. Anecdo-
tally, we found that the churches with the strongest
support from the pastor'swife had the best attendance,
while the churches with minimal support from the pastor's
wife had the lowest attendance.Future studieswill need
to develop additional strategiesto increaseattendance.
Ckwrck "oumership"of program. Prior work has demon-
strated the need to work in partnership with churches, not
merelyto imposethe researchagendaon the church.2'7'12'27'
2eIn each church, we worked with the pastor and church
leaders to train lay leaders and to give them the program
materials outright, allowing each church to make the pro-
gram its ov,n in whatever form desired after the 20 weeks
of health educator-managed sessions.Becauseof this or.m-
ership building process, sessionsfollowing the 20 weeks
varied by church. Some participants were more interested
in nutrition and left the weekly sessionsearly to avoid the
physical activity.One church was more interested in physi-
cal activity, and wished to drop the nutrition portion alto-
gether but did not. Pastoralcommitment appearedto be an
important factor in motivating continued participation;
even more essentialwas a commitment to the project from
the pastor's wife or another woman acting as a spiritual
leader of the church. Although attendance by the original
cohort for this studydecreased after the first 20 weeks, the
ownership building was quite successful.To date, eight of
the nine active intervention churches are holding weekly
sessionsled by their own lay leaders.The pilot intervention
church continues to meet weekly, more than four years
after the inception of Project Joy.
VoLUME 116 79
Pastoralcommitmentappearedto be an importantfactor in motivating
continuedparticipation;evenmore essentialwas a commitmentto the
projectfrom the pastor'swife or anotherwoman actingas a spiritual
leaderof the church.
Follow-up. Rates for follow-up were lower than
expected, although comparable to other church-based
studies. For example,in one study 757oof the interven-
tion group returned for follow-up while only 36% of the
control group returned.ePathwayshad 78% experimental
group return, and 90Vo control group at 14 weeks.8
Another had a 77% return rate two yearsout,rawhile still
others had 4l7o return at eight months,6 and 50% return
at one year.25Data from our post-intervention focus
groups and in-depth interviews, including some women
who dld not return for follow-up, suggestedthat the pri-
mary reason for not returning was participant failure to
achievethe goalsor biologicaloutcomeswithin the con-
text of the church, indicating a personalbreak with the
covenantthey had made in their spiritual lives. Virtually
none of the pastor-designedincentives or pastoral
entreatiesto complete follow-up worked. This may be a
problem unique to churchgoingwomen, but it does sug-
gest that the women who did not return had a higher
probability of having gained weight or some unfavorable
measurable variable. However, even the conservative
i n t e n t i o n - t o - t r e a t a n a l y s e ss t i l l s h o w e d s i g n i f i c a n t
changesin the desireddirection.
Implications. Despite their relatively small impact in
the active intervention groups, our church-basedinter-
ventions attenuated the common increase in body
weight shown to occur within as little as one year in
some women, as was observedin the self-help control
group.30These women, many of whom were post-
menopausal,both lost weight and improved their diets,
thus improving their lifestyle. The fact that this weight
loss was accompaniedby healthful behavior changesin
PUBLIC
nutrition suggeststhat there is an overallimprovementin
pro-health lifestyles in a group at very high risk for
chronic diseases.
Conversely,many community studies have shornn an
intervention'simpact on knowledgeand self-reportedbehav-
iors but havenot demonstratedany biologicalbenefit. In the
current study, the convergenceof modest to significant
dietary behavior change with favorable biological change
suggeststhat community interventions at the level of the
church, where there is continuing support and reinforce-
ment, havea reasonablechanceof influencing the health of
participants.If the statisticalfindings of this study (that is,
changesof this magnitudein the overallgroup accompanied
by changesof the magnitude found in l0% of the group)
were appliedto the entire populationof churchgoingAlrican
American women at risk for cardiovasculardiseases,it is
possible they could help shift the public health burden of
diseasein ways that more labor-intensiveclinical interven-
tions cannot. Our church-based interventions can reach
more people and sustain effects longer through the contin-
ual reinforcementof community systems.If interventionsof
this nature were disseminatedthrough largenational organi-
zations,such as the Congressof National Black Churches,
for example,which representsvarious denominationswith
65,000churchesand more than 20 million people,the pub-
lic health impact could be quite considerable.3l
This study was supported in part by CooperativeAgreement No.
U481CCU309674-01 from the Centers for DiseaseControl and Prevention
(SlP l7W-95) and the Of{iceof Women's Healththroughthe National
Heart, Lung,and Blood Institute.
We would like to thank all of our particiPating
churches,without whose
cooperationand commitmentthis proiectwould not havebeenpossible.
HEALTH REPORTS . 2OO1 SUPPLEMENT 1 . VOLUME 116
References

t. Expert Panelon the ldentificationEyaluationand Treatmentof Over- LeFebvreRC, LasaterTM, CarletonRA, PetersonG. Theory and deliv-
weight and Obesityin Adults. Clinicalguidelinesfor the identification, ery of health programmingin the community:the PawtucketHeart
evaluation,and treatmentof overweightand obesityin adults:the evi- HealthProgram.Prev Med | 987;| 6:80-95.
dence report. National Heart, Lung, and Blood Institute. Bethesda t 8 . BanduraA. SocialLearningTheory. EnglewoodCliffs (NJ): Prentice-
(MD): NationalInstitutesof Health; 1998. Hall;1977.
L. L e v i nJ S .T h e r o l e o f t h e b l a c kc h u r c hi n c o m m u n i t ym e d i c i n eJ. N a t l KushnerRF,KunigkA, AlspaughM, AndronisPT, LeitchCA, Schoeller
Med Assoc 1984;76:477 -83. DA, Validationof bioelectrical-impedance analysis as a measurement of
3. Eng E, HatchJ, CallanA. Institutionalizing socialsupport through the changein body compositionin obesity.Am J Clin Nutr 1990;52:2 | 9-23.
churchand into the community.HealthEducQ | 985;| 2:8| -92. FrohlichED, Grim C, LaBartheDR, MaxwellMH, PerloffD, Weidman
4. Eng E. The SaveOur SistersProject: a social network strategyfor WH. Recommendations for humanblood pressuredeterminationby
reachingrural blackwomen. Cancer 19931.72:1071-7. sphygmomanometers. Reporcof a specialtask force appointedby the
Davis DT, BustamanteA, Brown CP, Wolde-TsadikG, SavageEW, S t e e r i n gC o m m i t t e e , A m e r i c a n H e a r t A s s o c i a t i o n .H y p e r t e n s i o n
C h e n gX , e t a l . T h e u r b a nc h u r c ha n d c a n c e rc o n t r o l :a s o u r c eo f | 988;I l:209A-22A.
s o c i a l i n f l u e n c ei n m i n o r i t y c o m m u n i t i e s .P u b l i c H e a l t h R e p FriedewaldWT, LevyRl, FredricksonDS. Estimationof the concentra-
| 994;| 09:500-6. tions of low-densitylipoprotein(LDL) cholesterolin plasma,without
KumanyikaSK, CharlestonJB. LoseWeight and Win: a church-based use of the preparativeultracentrifuge. Clin Chem 1972;18:499-502.
weight loss programfor blood pressurecontrol amongblackwomen. 22. DiPietro L, CaspersenCJ,OstfeldAM, Nadel ER.A surveyfor assess-
P a t i e nEt d u cC o u n s1 9 9 2|;9 ( l ) : | 9 - 3 2 . i n g p h y s i c a la c t i v i t ) ,a m o n g o l d e r a d u l t s . M e d S c i S p o r t s E x e r c
LasaterTM, Wells BL, CarletonRA, ElderJP.The role of churchesin 1993:25:628-42.
d i s e a sper e v e n t i o rne s e a r csht u d i e sP. u b l i cH e a l t hR e p 1 9 8 6 ; l 0 l : 1 2 5 - 3 1 . 23. Zeger SL,LiangK-Y, Albert PS.Modelsfor longitudinaldata:a general-
McNabb WL, Quinn M, Kerver J, Cook S, Karrison T. The PATH- ized estimatingequationapproach.Biometrics1988;44:
1049-60.
WAYS church-based weight loss programfor urbanAfricanAmerican 24. OexmannMJ,ThomasJC, Taylor KB, O'Neil PM, GarveyWT, Lack-
women at risk for diabetes.DiabetesCare | 997;20:| 5 | 8-23. land DT, et al. Short-term impact of a church-basedapproachto
Wiist WH, FlackJM. A church-based cholesteroleducationprogram. risk in African Americans.Ethn Dis
lifestylechangeon cardiovascular
P u b l i cH e a l t hR e p 1 9 9 0 : 1 0 5 : 3 8 1 - 8 . 2000;10:17-23.
t 0 . VoorheesCC, StillmanFA, SwankRT, HeagertyPJ,LevineDM, Becker 25. Wells BL, DePueJD, LasaterTM, Carleton RA. A report on church
DM. Heart, Body,and Soul:lmpactof church-based smokingcessation site weightcontrol. HealthEducRes | 988;3:305- | 6.
interventionson readiness to quit. PrevMed I 996;25:| -9. LO. D e P u e J D , W e l l s B L , L a s a t e rT M , C a r l e t o n R A . V o l u n t e e r sa s
StillmanFA, Bone LR, RandC, LevineDM, Becker DM. Heart, Body, providersof heart health programsin churches:a report on imple-
and Soul:a church-based smoking-cessation programfor urbanAfrican mentation.Am J HealthPromot | 990;4:36l-6.
Americans.Prev Med | 993:22:335-49. 27. LasaterTM, CarletonRA, Wells BL. Religiousorganizations and large-
t2. OkwumabuaJO, Martin B, Clayton-DavisJ, PearsonCM. Stroke Belt s c a l e h e a l t h r e l a t e d l i f e s t y l e c h a n g e p r o g r a m s .J H e a l t h E d u c
Initiative:the Tennesseeexperience. J HealthCare Poor Underserved l99l:22:233-9.
1997;8:292-9. FerdinandKC. Lessonslearnedfrom the Healthy Heart Community
t3 Hatch JW, CunninghamAC, Woods WW, SnipesFC. The Fitness PreventionProject in reachingthe African American population.J
ThroughChurchesprojecc descriptionof a community-based cardio- HealthCare Poor Underserved1997;8:366-72.
v a s c u l ahr e a l t hp r o m o t i o ni n t e r v e n t i o nH. y g i e1 9 8 6 ; 5 : 9 - 1 2 . 29 SandersECl.New insightsand interventions:churchesunitingto reach
14. Campbell MK, Demark-WahnefriedW, Symons M, KalsbeekWD, the African American communitywith health information.J Health
DoddsJ, CowanA, et al. Fruit and vegetableconsumptionand preven- Care Poor Underserved1997;8:373-6.
tion of cancer:The BlackChurchesUnited for Better Health Project. 30. F l y n nM A , N o l p hG B , B a k e r A SK , r a u s eG . A g i n gi n h u m a n sa: c o n t i n u -
A m J P u b l i cH e a l t h1 9 9 9 ; 8 9 : 1 3 9 0 - 6 . ous 20-yearstudy of physiologicand dietary parameters.j Am Coll
t5 LincolnCE, MamiyaLH. The BlackChurch in the African American N u t r 1 9 9 2 ;l : 6 6 0 - 7 2 .
Experience. Durham(NC): Duke UniversityPress;1990. Congressof National Black Churches.Congressof National Black
G a l l u p ,G . R e p o r co n R e l i g i o ni n A m e r i c a , 1 9 9 6 .P r i n c e t o n( N J ) : Churches Website, home page. Availablefrom: URL: http://
PrincetonReligionResearchCenter; 1996. www,cnbc.org Washington(DC): Lastupdated,January6, 200| . I

PUBLIC HEALTH REPORTS . 2001 SUPPLEMENT I . VOLUME