scieee Science in your language
[en] (orig)

Participant Experiences in the Environmental Determinants of Diabetes in the Young Study: Common Reasons for Withdrawing

Abstract

Background. To characterize participant reasons for withdrawing from a diabetes focused longitudinal clinical observational trial (TEDDY) during the first three study years. Methods. 8677 children were recruited into the TEDDY study. At participant withdrawal staff recorded any reason parents provided for withdrawal. Reasons were categorized into (1) family characteristics and (2) protocol reasons. Families who informed staff of their withdrawal were classified as active withdrawals (AW); families without a final contact were considered passive withdrawals (PW). Results. Withdrawal was highest during the first study year . Most families were AW (; 73.4%). PW was more common in the United States (; 37.8%) and among young mothers . The most frequent protocol characteristic was blood draw (55%) and the most common family reason was not having enough time (66%). The blood draw was more common among female participants; being too busy was more common among males. Both reasons were associated with study satisfaction. Conclusions. Results suggest that, for families of children genetically at risk for diabetes, procedures that can be painful/frightening should be used with caution. Study procedures must also be considered for the demands placed on participants. Study satisfaction should be regularly assessed as an indicator of risk for withdrawal.

Read accessible full text

Participant Experiences in the Environmental Determinants of Diabetes in the Young Study: Common Reasons for Withdrawing

Author: Lernmark, Barbro,Lynch, Kristian,Baxter, Judith,Hyöty, Heikki,Kurppa, Kalle
Year: 2016
Source: https://trepo.tuni.fi/bitstream/10024/99074/1/participant_experiences_in_the_2016.pdf
Resea ch A icle
Pa icipan Expe iences in he En i onmen al De e minan s o
Diabe es in he Young S udy: Common Reasons o Wi hd awing
Ba b o Le nma k,1K is ian Lynch,2Judi h Bax e ,3,4
Roswi h Ro h,5Tuula Simell,6Lau a Smi h,2Ul ica Swa ling,1
Suzanne Benne Johnson,7and TEDDY S udy G oup1,2,3,5,6,8,9
1Depa men o Clinical Sciences, Lund Uni e si y, CRC, Jan Waldens ¨
oms Ga a 35, Sk˚
ane Uni e si y Hospi al (SUS),
20502 Malm¨
o, Sweden
2Heal h In o ma ics Ins i u e, Uni e si y o Sou h Flo ida, 3650 Spec um Boule a d, Sui e 100, Tampa, FL 33612, USA
3Ba ba a Da is Cen e o Childhood Diabe es, School o Medicine, Uni e si y o Colo ado Den e -AMC, 1775 Au o a Cou ,
Au o a, CO 80045, USA
4Colo ado School o Public Heal h, Depa men o Communi y and Beha io al Heal h, Uni e si y o Colo ado Den e -AMC,
13001E.17 hPlace,Au o a,CO80045,USA
5Ins i u e o Diabe es Resea ch, Helmhol z Cen e Munich and Clinic on he igh o Isa , Technical Uni e si y Munich,
Resea ch G oup Diabe es e.V., Ingols aed e Lands asse 1, 85764 Neuhe be g, Ge many
6Depa men o Pedia ics, Uni e si y o Tu ku, Kiinamyllynka u 4-8, 20100 Tu ku, Finland
7Depa men o Medical Humani ies and Social Sciences, Flo ida S a e Uni e si y College o Medicine, 1115 Wes Call S ee ,
Tallahassee, FL 32306, USA
8Paci ic No hwes Diabe es Resea ch Ins i u e, 720 B oadway, Sea le, WA 98122, USA
9Cen e o Bio echnology and Genomic Medicine, Medical College o Geo gia, Geo gia Regen s Uni e si y, 1120 15 h S ee ,
CA-4123, Augus a, GA 30912, USA
Co espondence should be add essed o Lau a Smi h; lsmi h5@heal h.us .edu
Recei ed 23 Janua y 2015; Accep ed 2 May 2015
Academic Edi o : Ni in Gup a
Copy igh © 2016 Ba b o Le nma k e al. This is an open access a icle dis ibu ed unde he C ea i e Commons A ibu ion
License, which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly
ci ed.
Backg ound. To cha ac e ize pa icipan easons o wi hd awing om a diabe es ocused longi udinal clinical obse a ional ial
(TEDDY) du ing he i s h ee s udy yea s. Me hods. 8677 child en we e ec ui ed in o he TEDDY s udy. A pa icipan wi hd awal
s a eco ded any eason pa en s p o ided o wi hd awal. Reasons we e ca ego ized in o (1) amily cha ac e is ics and (2) p o ocol
easons. Families who in o med s a o hei wi hd awal we e classi ied as ac i e wi hd awals (AW); amilies wi hou a inal con ac
we e conside ed passi e wi hd awals (PW). Resul s. Wi hd awal was highes du ing he i s s udy yea (𝑛 = 1220). Mos amilies
we e AW (𝑛 = 1549; 73.4%). PW was mo e common in he Uni ed S a es (𝑛 = 1001; 37.8%) and among young mo he s (𝑝 = 0.001).
The mos equen p o ocol cha ac e is ic was blood d aw (55%) and he mos common amily eason was no ha ing enough ime
(66%). The blood d aw was mo e common among emale pa icipan s; being oo busy was mo e common among males. Bo h
easons we e associa ed wi h s udy sa is ac ion. Conclusions. Resul s sugges ha , o amilies o child en gene ically a isk o
diabe es, p ocedu es ha can be pain ul/ igh ening should be used wi h cau ion. S udy p ocedu es mus also be conside ed o he
demands placed on pa icipan s. S udy sa is ac ion should be egula ly assessed as an indica o o isk o wi hd awal.
1. In oduc ion
The En i onmen al De e minan s o Diabe es in he Young
(TEDDY) s udy is a mul icen e longi udinal clinical obse -
a ional ial s udying he na u al his o y o he de elopmen
o ype 1 diabe es (T1DM) in child en. Soon a e bi h,
child en we e es ed o HLA con e ed gene ic isk o
T1DM. Child en wi h he highes gene ic isk we e in i ed o
pa icipa e in TEDDY. The pu pose o he TEDDY s udy is
o iden i y en i onmen al ac o s ha igge au oimmuni y
and TIDM [1].
Hindawi Publishing Co po a ion
Jou nal o Diabe es Resea ch
Volume 2016, A icle ID 2720650, 13 pages
h p://dx.doi.o g/10.1155/2016/2720650
2Jou nal o Diabe es Resea ch
The success o longi udinal esea ch s udies, in es iga ing
impo an ac o s ha con ibu e o T1DM, like TEDDY, is
dependen upon s udy e en ion. I is impo an o in es-
iga e wha s udy and psychosocial cha ac e is ics p omp
amilies o lea e a s udy in o de o (1) implemen possible
p e en i e ac ions o inc ease e en ion and (2) design u u e
longi udinal s udies o his a - isk T1DM popula ion in ways
ha enhance s udy e en ion. Al hough his opic is c i ical
o he success o longi udinal ials, he ex an li e a u e
is somewha spa se. In ac , only 55% pe cen o pedia ic
ials epo e usal o wi hd awal easons based on a ecen
li e a u e e iew [2]. P e ious wo k wi hin mo e a ied
pedia ic popula ions and in e en ional s udies (e.g., T1DM,
as hma, and obesi y in e en ion ials) has sugges ed ha a
numbe o sociodemog aphic ac o s (e.g., olde child age,
mino i y s a us, and lowe income) and psychological ac o s
(e.g., g ea e dep ession and lowe quali y o li e) we e ela ed
os udywi hd awal[3–6].Howe e , hese indings om
in e en ion s udies in ch onic illness popula ions may no be
ully applicable o he TEDDY a - isk o T1DM popula ion.
Fu he , p e ious s udies ha e ended o ocus on exis ing
cha ac e is ics o pa icipan s who do no comple e a s udy
a he han di ec ly asce aining easons o wi hd awal om
he pa icipan s hemsel es.
A longi udinal s udy simila o he TEDDY s udy
epo ed ha logis ical ma e s like blood sampling and
lack o ime we e he easons mos o en men ioned by
amilies who wi hd ew [7]. In he TEDDY popula ion, we
ha e epo ed ha cha ac e is ics o he s udy p o ocol, like
blood d aws, and amily ac o s, like being oo busy, we e he
p ima y easons amilies did no join he s udy [8]. S udy
en ollmen was associa ed wi h sociodemog aphic ac o s
such as whe he he child had a mo he , a he , and/o
sibling wi h T1DM ( i s deg ee ela i e (FDR)), had an olde
mo he , was a single on, o had a sibling al eady en olled
in he s udy. En ollmen a es di e ed be ween he TEDDY
coun ies, wi h a la ge p opo ion o pa en s ec ui ed om
he Eu opean coun ies [8].
In o he p io wo ks, we iden i ied p edic o s o wi h-
d awal du ing he i s yea (up o he 15 mon hs’ isi ) o
TEDDY among amilies om he gene al popula ion (GP)
who had no immedia e amily membe wi h T1DM [9]. S udy
wi hd awal was mo e common i he mo he was young, he
a he did no pa icipa e, o he s udy child was emale. Also,
mo he s o child en who wi hd ew we e mo e likely o epo
smoking du ing p egnancy, abs aining om alcohol, and
educing hei wo k hou s o no wo king a all du ing p eg-
nancy. Mo he s who wi hd ew we e also mo e likely o ail o
comple e i ems on s udy ques ionnai es and o unde es ima e
hei child’s TIDM isk. Among mo he s wi h accu a e pe -
cep ions o hei child’s T1DM isk, high ma e nal anxie y was
associa ed wi h s udy wi hd awal [9]. This in o ma ion was
used o iden i y amilies a high isk o lea ing he TEDDY
s udy in he i s yea ; hese amilies we e hen p o ided
wi h an in e en ion o p omo e e en ion [10]. While ac o s
associa ed wi h wi hd awal could be used o sc een o
amilies a isk o d opou , he e ec i eness o in e en ions
aimed a imp o ing he e en ion o pa icipan s can also be
in luenced by he amilies’ speci ic easons o wi hd awing.
Thus, in his s udy we examined da a om all amilies who le
TEDDY (bo h GP and FDR) du ing he i s h ee s udy yea s,
including sociodemog aphic and psychosocial a iables, and
hei easons o op ingou o hes udy.
2. Me hods
2.1. TEDDY S udy and Da a. The TEDDY s udy has cen e s
in ou coun ies (Finland, Ge many, Sweden, and he Uni ed
S a es) and is suppo ed by he Na ional Ins i u es o Heal h
(NIH). The s udy p o ocol includes s udy isi s e e y h ee
mon hs om 3-4 mon hs o age un il he child is ou yea s
old and biannually he ea e . The s udy p o ocol includes
blood d aws, nasal swabs, heigh and weigh measu emen s,
and pa en al in e iews whe e aspec s o he child’s heal h
a e eco ded oge he wi h di e en ypes o li e e en s.
A egula in e als TEDDY pa en s ill ou ques ionnai es
wi h demog aphic ques ions, heal h his o ies, li e e en s,
and pa en s’ wo ies and anxie y conce ning he child’s
inc eased gene ic isk o T1DM. Pa en s a e also eques ed
o comple e ood dia ies and collec s ool samples. Al oge he ,
he TEDDY s udy p o ocol is e y demanding in leng h and
in e ms o he equency and na u e o i s componen s [1].
The collec ion o co d blood o sc eening and possible
TEDDY en ollmen s a ed in Sep embe 1, 2004, and ended
Feb ua y 28, 2010. A o al o 424,788 child en we e sc eened
o inc eased HLA con e ed gene ic isk o T1DM and
21,589 we e HLA eligible [11]. The en ollmen a e o amilies
wi h child en a inc eased gene ic isk was 38.4% om he
gene al popula ion (GP) and 64.8% om FDR amilies. The
numbe and p opo ion o eligible child en as well as he
numbe o child en en olled di e ed bo h be ween coun ies
and be ween GP and FDR amilies [8].
The ea lie a amily wi hd ew om he s udy, he ewe
da a poin s we e a ailable. Fo all child en, demog aphic ac-
o s like coun y, gende , mon h o bi h, FDR/GP s a us, and
mo he ’s age we e ob ained in connec ion wi h he collec ion
o co d blood o de e mine eligibili y o TEDDY. Da a o
all child en en olled in TEDDY we e also a ailable om
he ques ionnai es ha each pa en comple ed in connec ion
wi h he i s TEDDY isi . These psychosocial measu es
con ained ques ions on pa en s’ iews on he child’s isk o
de eloping diabe es, hei wo ies abou ha possibili y, and
hei hough s on ha ing he child es ed o gene ic T1DM
isk.Themo he alsoanswe edques ionsonli es yle ac o s
du ing p egnancy (smoking and alcohol consump ion). Fo
amilies wi hd awing a e he i s s udy yea , he e we e
addi ional demog aphic da a collec ed du ing he i s yea o
TEDDY, including pa en s’ educa ion, child e hnic mino i y
s a us, only child s a us, and household c owding. Psychoso-
cial da a we e collec ed epea edly s a ing a he 3 and 6
mon hs’ isi s, a 15 mon hs, a 27 mon hs, and yea ly. In he
cu en s udy, he psychosocial da a collec ed immedia ely
be o e wi hd awal we e used. These da a a e wo ies (e.g.,
Howo endoyouwo y ha you childwillge diabe es:
ne e / e y o en), anxie y (sho e sion o he S a e-T ai
Anxie y In en o y (SAI)) [12],dep ession (B adley’s Wellbe-
ing Scale) [13], s udy sa is ac ion ( h ee co ela ed ques ions
summed in o a sa is ac ion sco e: O e all, how do you eel
Jou nal o Diabe es Resea ch 3
The e we e 2621 changes in he s udy
pa icipa ion o ms illed ou
No ac i e con ac o una ailable (PW)
N=2353
Change in pa icipa ion in ol ed ejoining he
s udy (n = 186)
Change in pa icipa ion in ol ed second
wi hd awal o loss o ollow-up a e amily had
ejoined (n=35)
Ac i e wi hd awal, ac i e con ac (AW) N=1452
N = 560
HLA ineligible de e mined a e sc eening
(n=76) o de eloped T1D sho ly a e las isi ,
N = 2318
N = 2109
N=2539 CSP o m comple ed by mis ake (n=6)
Wi hd ew a e 36 mon hs’ isi (n = 209)
Figu e 1: S udy popula ion.
abou ha ingyou childpa icipa ein heTEDDYs udy?Do
you hink you child’s pa icipa ion in he TEDDY s udy was
a good decision? Would you ecommend he TEDDY s udy o
a iend?), and isk pe cep ion (Compa ed o o he child en
do you hink you child’s isk o de elop diabe es is much
lowe /much highe ?).
2.2. Da a Collec ion. When a amily wi hd ew om TEDDY,
aChangeinS udyPa icipa ionFo m(CSPFo m)was illed
ou by TEDDY s a and any eason he pa en ga e o
lea ing he s udy was eco ded. Mo e han one eason could
be eco ded. I a amily did no gi e a eason o lea ing he
s udy his was also no ed. The las isi when any da a we e
collec ed om he amily was aken as he ime o wi hd awal
e en hough he CSP o m could be comple ed a a much
la e da e. Fo some amilies, he decision o wi hd aw om
TEDDY was di icul and could span o e a long pe iod o
ime wi h se e al cancelled isi s and no collec ion o da a
be o e a inal decision o lea e he s udy. I a amily did no
come o he clinic o mo e han a yea despi e scheduled
isi s and did no con ibu e any da a o e he cou se o one
yea , he amily was conside ed o be a Passi e Wi hd awal
(PW) and a CSP Fo m was comple ed. Families ha became
una ailable and impossible o each we e also PW and we e
conside ed TEDDY wi hd awals.
2.3. S a is ical Analysis. Di e ences in equencies be ween
ca ego ical g oups we e es ed by chi-squa e es s. Fo con in-
uous a iables, di e ences in means we e es ed by indepen-
den wo-sample 𝑡- es s. Mul iple linea eg ession was used
o examine he associa ion o demog aphic and psychosocial
a iables wi h speci ic common easons o wi hd awing.
Demog aphic a iables a ailable on all subjec s we e exam-
ined i s and la e psychosocial ac o s we e added. Da a
om he las ques ionnai e p io o wi hd awal we e used o
es ima e ma e nal anxie y, isk pe cep ion, wo y, and s udy
sa is ac ion. I o any eason he e we e missing da a, in o -
ma ion was aken om he las ques ionnai e a ailable. Uni-
a ia e and mul iple logis ic eg ession models we e used o
es o signi ican ac o s associa ed wi h ype o wi hd awal
(PW e sus AW). Analyses we e pe o med using SAS 9.2. 𝑝
alues less han 0.05 we e conside ed s a is ically signi ican .
3. Resul s
A o al o 8677 child en we e ec ui ed in o he TEDDY
s udy. F om Sep embe 1, 2004, un il July 31, 2012, he e we e
2621 CSP o ms illed ou . CSP o ms om all child en who
le hes udyono be o e he3-yea isi we eselec ed o
analysis.Figu e1gi esano e iewo how hes udycoho
wasc ea ed.Inall,512 o mswe eexcluded om hes udy o
easons ou lined in he igu e, esul ing in 2109 CSP o ms
desc ibing he i s ime wi hd awal o he amily om he
s udy. O hese, 1549 o ms came om amilies who old he
s a hey wan ed o wi hd aw om TEDDY (AW); 90% ga e
a leas one eason why hey op ed ou . A o al o 560 amilies
(26.6%) did no espond o epea ed scheduling a emp s
o mo e han a yea o became impossible o con ac and
we e wi hd awn by he TEDDY s a wi hou any u he
con ac (PW). In Figu e 2, he numbe o AW and PW
o he di e en coun ies is shown by isi . O e all, s udy
wi hd awal was highes du ing he i s yea o he s udy and
dec eased he ea e and AW was a mo e common han PW.
The Uni ed S a es had he highes equency o PW (𝑛 = 378;
4Jou nal o Diabe es Resea ch
Finland
150
100
50
0
To al numbe o wi hd awals (n)
Las isi
3,69,12 15,18 21,24 27, 30 33,36
Passi e wi hd awal (PW)
Ac i e wi hd awal (AW)
Ge many
To al numbe o wi hd awals (n)
Las isi
3,69,12 15,18 21,24 27, 30 33,36
50
40
30
20
10
0
Sweden
To al numbe o wi hd awals (n)
Las isi
3,69,12 15,18 21,24 27, 30 33,36
250
200
150
100
50
0
Passi e wi hd awal (PW)
Ac i e wi hd awal (AW)
US
To al numbe o wi hd awals (n)
Las isi
3,69,12 15,18 21,24 27, 30 33,36
500
400
300
200
100
0
Figu e 2: Numbe o wi hd awals by age o las isi and by coun y di ided in o passi e wi hd awals (PW) (checked) and ac i e wi hd awals
(AW) (whi e).
23.9%) and he ela i e p opo ion o PW inc eased o e
ime. Finland had he lowes PW a e (𝑛=34;8.9%)andi
emained low ac oss all s udy yea s. In Ge many 39 child en
(23.9%) and in Sweden 109 child en (19.0%) we e classi ied
as PW. The e was a signi ican ly inc easing end o PW
p opo ion o e he s udy pe iod in Ge many (𝑝 = 0.024)
and Sweden du ing he h ee yea s (𝑝 = 0.001)e en hough
he numbe o bo h PW and AW dec eased.
Uni a ia e and mul iple eg essions we e used o iden i y
ac o s ha di e en ia ed be ween AW and PW. PW we e
signi ican ly mo e common in o he coun ies compa ed o
Finland, among young moms, and in olde child en. In he
uni a ia e models, high anxie y, ma e nal smoking du ing
p egnancy, and lack o a he pa icipa ion in TEDDY we e
associa ed wi h PW. Howe e , hese ac o s did no emain
signi ican in he mul i a ia e model (Table 1).
Table 2 depic s cha ac e is ics o he AW by age o he
childa he imeo wi hd awal.In he i s hal yea a e
en ollmen , 24.1% o he AW we e young mo he s (<25 yea s)
which was signi ican ly di e en compa ed o yea 3 when
14.8% we e young mo he s. The ea ly AW we e also mo e anx-
ious mo he s (<0.001) and mo he s mo e wo ied abou hei
child de eloping diabe es (𝑝 = 0.003)compa ed o hose
emaining in he s udy. The e we e no signi ican associa ions
be ween he accu acy o he mo he ’s T1DM isk pe cep ion
o s udy sa is ac ion and he child’s age a wi hd awal.
The equencies o di e en easons epo ed by he AW
dis ibu ed o e isi s du ing he h ee s udy yea s a e shown
Jou nal o Diabe es Resea ch 5
Table 1: Fac o s associa ed wi h passi e wi hd awal (PW) e sus ac i e wi hd awal (AW).
Fac o saNumbe
mean (SD)
Uni a ia e Mul i a ia e
%o PW OR 95%CI 𝑝 alue OR 95% CI 𝑝 alue
Coun y o esidence
Finland 383 8.9 1.00 Re . 1.00 Re .
Sweden 562 19.4 2.47 1.64–3.72 2.86 1.85–4.44
Ge many 163 23.9 3.22 1.95–5.34 3.73 2.16–6.43
US 1001 37.8 4.28 3.26–9.06 <0.001 6.57 4.35–9.93 <0.0011
Ma e nal age a child’s bi h (yea s)
Yea s 28.9 (5.7) 0.94 0.93–0.96 <0.001 0.95 0.93–0.97 <0.001
<29 yea s 1016 32.0
≥29 yea s 1093 21.5
Child’s age o wi hd awal (yea s)
Yea s 1.16 (0.80) 1.12 0.99–1.25 0.08 1.33 1.16–1.52 <0.001
<1.16 yea s 1231 25.8
≥1.16 yea s 878 27.6
Gende
Female 1094 25.9 1.00 Re . 1.00 Re .
Male 1015 27.3 1.08 0.88–1.31 0.46 1.10 0.88–1.36 0.40
FDRa
No 1957 26.8 1.00 Re . 1.00 Re .
Yes 152 23.7 0.85 0.57–1.25 0.41 0.95 0.62–1.47 0.81
Smoking du ing p egnancy
No 1556 25.2 1.00 Re . 1.00 Re .
Yes 415 30.4 1.30 1.02–1.64 0.03 1.31 1.00–1.72 0.05
Alcohol 3 d imes e
No 1713 26.1 1.00 Re . 1.00 Re .
Yes 289 27.3 1.07 0.81–1.41 0.66 1.16 0.85–1.60 0.35
Wo ked du ing p egnancy
No o educed hou s 1184 26.6 1.00 Re . 1.00 Re .
Yeso inc easedhou s 819 25.8 0.96 0.78–1.17 0.67 0.92 0.74–1.15 0.45
High anxie y (SAI >48)
Sco e 40.9 (10.7) 1.02 1.01–1.03 <0.001 1.00 0.99–1.01 0.80
<40.9 sco e 1059 22.6
≥40.9 sco e 921 29.4
Risk pe cep ion
Unde es ima e 901 27.3 1.00 Re . 1.00 Re .
Accu a e 1097 25.2 0.90 0.73–1.09 0.28 0.99 0.80–1.23 0.93
Fa he no ac i e (3 mo)b
No 1807 25.0 1.00 Re . 1.00 Re .
Yes 302 36.1 1.70 1.31–2.20 <0.001 1.27 0.90–1.78 0.17
aFDR = i s deg ee ela i e has ype 1 diabe es. bFa he did no answe 3-mon h ques ionnai es.

6Jou nal o Diabe es Resea ch
Table 2: Ac i e wi hd awals by child’s age o wi hd awal and demog aphic ac o s, ma e nal psychosocial ac o s, and ma e nal s udy
sa is ac ion.
Fac o s
All
3–36 m
(𝑁)
Yea 1
3–6 m
𝑁( ow %) o
mean (SD)
Yea 1
9–12 m
𝑁( ow %) o
mean (SD)
Yea 2
15–24 m
𝑁( ow %) o
mean (SD)
Yea 3
27–36 m
𝑁( ow %) o
mean (SD)
𝑝 alue
Numbe o ac i e wi hd awals 1549 584 (37.7) 324 (20.9) 431 (27.8) 210 (13.6)
Coun y o esidence
Finland 349 106 (30.4) 69 (19.8) 122 (35.0) 52 (14.9)
Sweden 453 167 (36.9) 92 (20.3) 130 (28.7) 64 (14.1)
Ge many 124 37 (29.8) 36 (29.0) 38 (30.6) 13 (10.5)
US 623 274 (44.0) 127 (20.4) 141 (22.6) 81 (13.0) <0.001
Gende
Female 811 323 (39.8) 159 (19.6) 232 (28.6) 97 (12.0)
Male 738 261 (35.4) 165 (22.4) 199 (27.0) 113 (15.3) 0.07
Fi s deg ee ela i e wi h T1D
No 1433 534 (37.3) 302 (21.1) 406 (28.3) 91 (13.3)
Yes 116 50 (43.1) 22 (19.0) 25 (21.6) 19 (16.4) 0.30
Ma e nal age a child’s bi h
yea s 29.4 (5.6) 28.9 (5.8) 29.0 (5.2) 29.6 (5.5) 30.9 (5.5) <0.001
Highlyanxiousa las isi (SAI>48)a
No 1241 413 (33.3) 259 (20.9) 375 (30.5) 191 (15.4)
Yes 229 116 (50.7) 55 (24.0) 45 (19.7) 13 (5.7) <0.001
Wo y abou diabe es a las isi a,b
Ne e o a ely 706 104 (14.7) 166 (23.5) 290 (41.4) 144 (20.4)
Some imes o e y o en 344 56 (16.3) 114 (33.1) 113 (33.4) 59 (17.2) 0.004
Risk pe cep ion a las isi a
Unde es ima e 692 247 (35.7) 147 (21.2) 207 (30.1) 90 (13.0)
Accu a e 783 283 (36.1) 170 (21.7) 213 (27.5) 115 (14.7) 0.63
S udy sa is ac ion a las isi a,b
Ve y sa is ied 276 38 (13.8) 77 (27.9) 108 (39.1) 53 (19.2)
Sa is ied 289 44 (15.2) 87 (30.1) 99 (34.3) 59 (20.4)
Somewha sa is ied 274 39 (14.2) 71 (25.9) 115 (42.0) 49 (17.9)
Neu al o dissa is ied 208 38 (18.3) 44 (21.2) 84 (40.4) 42 (20.2) 0.47
aI no measu e las isi , he second o las isi is aken i a ailable.
bS udy sa is ac ion and wo y abou diabe es a e no asked in he i s ques ionnai es so he e a e ewe a ailable answe s a 3 and 6 mon hs.
in Table 3. The di e en easons a e g ouped in o “p o ocol
cha ac e is ics” o “ amily ac o s.” The wo easons mos
equen ly gi en o lea ing TEDDY we e conce ns abou he
blood d aw (𝑛 = 359; 55% o all p o ocol cha ac e is ics)
and being oo busy/no ha ing enough ime (𝑛 = 587; 66.6%
o all amily ac o s).O he equen lymen ionedp o ocol
cha ac e is ics included he ollowing: he p o ocol is oo
demanding, anspo a ion di icul ies, and he equency o
isi s. Among he amily ac o s, eeling o e whelmed/being
oo s essed is he second mos common eason gi en (𝑛=
206, 23.4%). Ha ing conce ns abou he blood d aw was mo e
o en men ioned o olde child en (𝑝 = 0.039), no wan ing
o be eminded o he child’s isk was signi ican ly mo e o en
epo ed as a eason o wi hd awal du ing he i s isi s (𝑝=
0.003), and being oo busy/no ha ing enough ime was mo e
o en epo ed in he la e isi s (𝑝 = 0.037). Du ing he s udy
pe iod, 11% (𝑛=97) o he amilies ci ed mo ing ou o he
a ea as a eason o op ing ou . A o al o 165 amilies (10.7%)
did no wan o gi e a eason o lea ing o only wan ed o
wai and see wha migh happen.
The esul o mul iple eg essions examining demo-
g aphic ac o s in all AW subjec s o he wo mos o en
men ioned easons o wi hd awal (conce ns abou blood
d aw and being oo busy) is p esen ed in Table 4. Ge man
and US mo he s we e mo e likely o epo he blood d aw
as he eason o lea ing TEDDY compa ed o Finland
and Sweden. Also, he blood d aw was men ioned mo e
o enas hechildgo olde andi hechildwasagi l.
Being oo busy was gi en as he eason o lea ing TEDDY
mos o en among Swedish mo he s and leas o en among
Ge man mo he s. This eason was mo e common in amilies
wi h an olde TEDDY child and i he child was a boy.
Jou nal o Diabe es Resea ch 7
Table 3: F equency o common easons o wi hd awing om he TEDDY s udy du ing he i s h ee s udy yea s. Pe cen ages o easons
men ioned mo e equen ly a e shown.
Wi hd awals
Las isi be o e wi hd awal (mon h)
Yea 1 Yea 1 Yea 2 Yea 3 All
3–6 m 9–12 m 15–24 m 27–36 m 3–36 m
𝑁( ow %) 𝑁( ow %) 𝑁( ow %) 𝑁( ow %) 𝑁%
Ac i e wi hd awals (AW) 584 (37.7) 324 (20.9) 431 (27.8) 210 (13.6) 1549 73.4
Reasons o wi hd awal
P o ocol cha ac e is ics 244 (37.4) 128 (19.6) 195 (29.9) 86 (13.2) 653 100
Conce ns abou blood d aw 124 (34.5) 63 (17.5) 112 (31.2) 60 (16.7) 359 55.0a
P o ocol oo demanding 68 (40.0) 41 (24.1) 43 (25.3) 18 (10.6) 170 26.0a
T anspo a ion di icul ies 44 (41.1) 21 (19.6) 35 (32.7) 7 (6.5) 107 16.4
Conce ns abou equency o isi s 40 (39.2) 20 (19.6) 33 (32.4) 9 (8.8) 102 15.6
Conce ns abou s ool samples 14 9 13 7 43 6.6
Conce ns abou ques ionnai es 15 8 10 4 37 5.7
Food dia ies oo oublesome 12 1 2 1 34 5.2
Do no wan o be eminded o isk 19 6 5 1 31 4.7
Du a ion o s udy is oo long 1351101.5
No ea men o p e en o e ed 312171.1
Wo ied abou p i acy 11002—
Wo ied abou loss o insu ance 30003—
O he p o ocol cha ac e is ics 2352121.8
Family ac o s 279 (35.6) 162 (20.7) 231 (29.5) 112 (14.3) 784 100
Too busy/no enough ime 187 (31.9) 134 (22.8) 185 (31.5) 81 (13.8) 587 66.6a
Feeling o e whelmed/s essed 82 (39.8) 40 (19.4) 58 (28.2) 26 (12.6) 206 23.4
Mo ing ou o he s udy a ea 33 (30.4) 23 (23.7) 22 (22.7) 19 (19.6) 97 11.0
Child medical/beha io al p oblems 29 (40.3) 16 (22.2) 16 (22.7) 11 (19.6) 72 8.2
Family membe emo ional p oblems 15 9 18 10 52 5.9
Does no wan o be in esea ch 9385252.8
Family membe does no ag ee o pa icipa e 8054171.9
Family membe in ano he s udy 10203—
Subjec al eady in ano he s udy 01001—
Fam. heal h ca e p o ide no ecommended 10001—
Language ba ie 10001—
O he amily ac o s 4 2 4 3 13 1.5
Ac i e con ac made bu no eason gi en o
wan s o wai and see 70 (42.4) 28 (17.0) 42 (25.4) 25 (15.2) 165 10.7
aSigni ican di e ence be ween age g oups: conce ns abou blood d aw (𝑝 = 0.039),dono wan obe emindedo isk(𝑝 = 0.003), and oo busy/no enough
ime (𝑝 = 0.037).
Ma e nal smoking du ing p egnancy, mo he wo king du ing
p egnancy, ma e nal alcohol consump ion du ing p egnancy,
a he pa icipa ion in TEDDY, and whe he he TEDDY
child was a FDR o om he GP we e no associa ed wi h
ei he eason o lea ing TEDDY.
In Table 5, wo logis ic eg essions explo e ma e nal psy-
chosocial ac o s in ela ion o he wo mos equen ly
men ioned ac o s o wi hd awing, conce ns abou blood
d aw and being oo busy. The eg essions a e adjus ed o
coun y o esidence, ma e nal age, age o child a s udy
wi hd awal, and gende . The esul s show ha conce ns
abou blood d aw we e associa ed wi h he mo he ’s s udy
sa is ac ion bo h a 6 mon hs and a he las isi be o e
wi hd awal. A bo h 6 mon hs and he las isi , mo he s
8Jou nal o Diabe es Resea ch
Table 4: Mul iple logis ic eg ession examining demog aphic ac o s in ela ion o (a) conce ns abou blood o (b) being oo busy as a eason
o wi hd awing among hose who ac i ely wi hd ew.
Fac o s a en ollmen
Demog aphic measu es in ela ion o conce ns abou blood d aw and being oo busy (𝑛 = 1549)
(a) Conce ns abou blood d aw (b) Being oo busy
OR 95% CI 𝑝 alue OR 95% CI 𝑝 alue
Coun y o esidence
Finland 1.00 Re . 1.00 Re .
Sweden 1.33 0.91–1.93 1.47 1.10–1.96
US 1.72 1.21–2.43 0.83 0.63–1.09
Ge many 7.80 4.89–12.4 <0.001 0.48 0.30–0.77 <0.001
Ma e nal age a child’s bi h (y s)
Yea s 1.05 1.02–1.07 <0.001 1.00 0.98–1.01 0.62
Child’s age o wi hd awal (mon hs)
Mon hs 1.20 1.03–1.40 0.02 1.15 1.01–1.31 0.04
Gende
Male 1.00 Re . 1.00 Re .
Female 1.44 1.12–1.85 0.004 0.78 0.63–0.96 0.02
No e: FDR/GP s a us, smoking du ing p egnancy, wo king du ing p egnancy, alcohol consump ion du ing p egnancy, and dad’s pa icipa ion in TEDDY we e
no associa ed wi h ei he conce ns abou blood d aw o being oo busy.
who epo ed less sa is ac ion wi h TEDDY we e mo e likely
o epo conce ns abou he blood d aw as he eason o
lea ing TEDDY. The ela ionship be ween s udy sa is ac ion
and mo he s’ epo o being oo busy as he eason o
lea e TEDDY was less clea . No o he psychosocial ac o s
(ma e nal anxie y o mo he ’s isk pe cep ion) showed an
associa ion wi h he wo mos common easons o lea ing
he TEDDY s udy.
4. Discussion
The TEDDY s udy, which seeks o iden i y ac o s associa ed
wi h he de elopmen o T1DM, has a demanding p o ocol
o bo h he child en and hei pa en s. The s udy is also
longi udinal wi h ou isi s o a TEDDY clinic each yea
un il hechildis ou yea so ageandbiannually he ea e
un il he child is i een yea s o age. A e 8 yea s, 72.2%
o he ec ui ed child en a e s ill pa icipa ing in he s udy.
The majo i y o amilies who le ga e a eason o lea ing.
PW was mo e common among he US pa icipan s. Being a
la ge coun y wi h a di e se popula ion, i is mo e di icul
o ack people compa ed o he Eu opean coun ies. We
p e iously epo ed ha U.S. amilies o en ailed o espond
o phone messages o le e s in i ing hem o join TEDDY [8],
cons i u ing passi e e usal, which is simila o PW.
Finlandhad helowes numbe o PWandsimila num-
be s o e he yea s, while he p opo ion o PW in Ge many
and Sweden ended o inc ease du ing he h ee s udy yea s
e en hough he o al numbe o wi hd awals dec eased
signi ican ly. Some s udy si es may keep a TEDDY pa icipan
as “ac i e” in TEDDY despi e mul iple missed isi s. A e
ge ing o know TEDDY s a o e many mon hs some
amilies may ha e di icul y di ec ly elling s a ha hey a e
lea ing he s udy and may ins ead jus “no-show.” Di e ences
be ween s udy cen e s in how amilies a e managed migh
de elop o e ime and his is a weakness o he cu en esul s
epo ed.Howe e ,i isdi icul o ala ges udylikeTEDDY
o sys ema ically de ine how s a uni o mly manage s udy
amilies o e many yea s.
Sociodemog aphic ac o s also ela ed o s udy wi h-
d awal. PW mo he s we e younge and mo e likely o smoke
du ing p egnancy han hose e ained in he s udy. In he
uni a ia e analysis, lack o a he pa icipa ion in TEDDY
was associa ed wi h PW bu his e ec was no s a is ically
signi ican in he mul i a ia e model. In p e ious wo k, we
ound ha lacko a he pa icipa ionwasanimpo an
p edic o o s udywi hd awalin he i s yea o TEDDY
[9]. Fa he in ol emen in a s udy may be a mo e impo an
de e minan o whe he a amily s ays in a s udy o wi hd aws;
i may no p edic whe he he wi hd awal is ac i e o passi e.
Logis ical aspec s o he s udy we e ound o be common
easons o wi hd awal. E en hough TEDDY s a is e y
skilled, d awing blood om a small child can be e y
challenging and cause unpleasan expe iences bo h o he
child and he pa en . O he s udies ha e epo ed ha blood
d aws can be an obs acle o s udy pa icipa ion and a eason
o op ing ou [7]. Repo ing he blood d aw as a eason o
lea ing TEDDY was mo e common la e in he s udy han in
he ea ly phase. All blood d aws a e done a e applica ion o
de mal anes he ics so a baby migh eac less han a sligh ly
olde child who migh ha e lea ned o ea he blood d aw.
This obse a ion has been e i ied by Swedish TEDDY nu ses
who conduc ed a pa en su ey o he child’s eac ion o he
blood d aw. In ac , pa en s ended o a e s onge eac ions
in olde child en (pe sonal communica ion).
Being oo busy and no ha ing ime o do he TEDDY
asks was he mos equen ly men ioned eason o lea ing
he s udy. Being s essed and eeling o e whelmed was
Jou nal o Diabe es Resea ch 9
Table 5: Logis ic eg ession examining ma e nal psychosocial ac o s in ela ion o (a) conce ns abou blood d aw o (b) being oo busy as a
eason o wi hd awing a e adjus ing o coun y o esidence, ma e nal age, child’s age a wi hd awal, and gende (see Table 4).
Psychosocial ac o s
Psychosocial measu es in ela ion o conce ns abou blood and being oo busy a e adjus ing o
demog aphic ac o s (𝑛 ∼ 1031)
(a) Conce ns abou blood d aw (b) Being oo busy
OR 95% CI 𝑝 alue OR 95% CI 𝑝 alue
Ma e nal high anxie y (SAI >48)
Sco e 1.00 0.98–1.01 0.42 0.99 0.98–1.00 0.09
Mo he ’s isk pe cep ion
Unde es ima e 1.00 Re . 1.00 Re .
Accu a e 1.01 0.74–1.38 0.95 0.91 0.70–1.18 0.49
S udy sa is ac ion (6 mo)
Ve y sa is ied 1.00 Re . 1.00 Re .
Sa is ied 1.38 0.89–2.15 1.24 0.88–1.74
Somewha sa is ied 1.80 1.16–2.78 1.80 1.27–2.59
Neu al o dissa is ied 2.73 1.68–4.42 <0.001 1.11 0.73–1.69 0.007
S udy sa is ac ion (las isi )
Ve y sa is ied 1.00 Re . 1.00 Re .
Sa is ied 1.57 0.99–2.49 1.19 0.84–1.69
Somewha sa is ied 2.54 1.61–4.03 1.43 1.00–2.05
Neu al o dissa is ied 2.64 1.62–4.31 <0.001 1.33 0.89–1.97 0.252
ano he impo an eason o no pa icipa ing anymo e.
Being busy was signi ican ly mo e o en men ioned as a
eason o wi hd awal a he la e s udy isi s. I may be
ha when a baby is bo n, he mo he is o en home ca ing
o he baby and may no expe ience he TEDDY asks as
bu densome, compa ed o la e when she may e u n o wo k.
Also, some TEDDY asks a e easie o comple e when he
child is a baby like collec ing s ool samples o doing a 3-day
ood dia y. Also, being in he beginning o a s udy may gi e
pa icipan s a eeling o cu iosi y and en husiasm, some hing
ha may disappea as he s udy seems less no el o amilies.
Psychological easons also played a ole in wi hd awal
o some amilies, pa icula ly ea ly in he s udy. Mo he s
in amilies who wi hd ew ea ly, a e he i s o second
TEDDY isi ,appea ed obemo eanxiousandwo iedabou
hei child ge ing diabe es compa ed o mo he s in amilies
lea ing hes udya 9mon hso la e .Thisisunde sco edby
he obse a ion ha mo he s who epo ed ha hey did no
wan o be eminded o he child’s isk o T1DM as a eason
o wi hd awal o en le TEDDY a e he i s wo isi s.
The wo mos impo an ac o s men ioned as easons
o wi hd awal (blood d aw and being oo busy) we e each
analyzed in sepa a e eg ession models, i s in ela ion o
demog aphic ac o s and in a second model in ela ion
o ma e nal psychological ac o s and s udy sa is ac ion.
Repo ing he blood d aw as a eason o wi hd awal was
mo e common i he child was a gi l while s a ing ha he
amily was oo busy o pa icipa e was mo e common i he
child was a boy. A s udy on in an pain esponse ollowing
immuniza ion injec ion demons a ed ha pa en al beha io
has a key ole in in luencing how in an s espond o pain ul
p ocedu es wi h di e ences be ween emale and male in an s
[14].Wecanonlyspecula e ha pa en smaybemo esensi i e
o he possible discom o o he blood d aw in gi ls han in
boys; boys a e o en expec ed o be b a e han gi ls. Ano he
s udy ound ha gi ls’ pain h eshold is lowe han ha o
boys, a leas o sligh ly olde child en compa ed o he
TEDDY child en in ou s udy [15]. This could indica e ha
he eac ion o gi ls o he blood d aw migh be s onge han
ha o he boys and he e o e he pa en s migh be mo e
p one o op ou when he child is a gi l e en when he child
is younge .
Why being oo busy was mo e o en men ioned when
he child was a boy is ha de o explain. Some imes boys a e
mo e physically ac i e and his may c ea e mo e p oblems o
pa en s in collec ing TEDDY samples and da a in p epa a ion
o he isi . I may also mean ha boys a e mo e likely o
openly p o es going o he TEDDY clinic. All his aken
oge he migh gi e pa en s a eeling o no ha ing ime and
being oobusy.In hecu en s udy,pa en swhoga e he
blood d aw as a eason o lea ing o en exp essed lowe
sa is ac ion wi h he s udy bo h a 6 mon hs, when his
was i s assessed in TEDDY, and in he su ey comple ed
be o e op ing ou . I is likely ha di icul ies wi h he blood
d aw we e seen ea ly in he s udy, some imes con inued, and
esul ed in lowe sa is ac ion wi h he TEDDY expe ience and
ul ima ely wi hd awal om he s udy.
In a p io s udy explo ing easons o why pa en s s ay in
TEDDY, ha ing someone wa ching he child o de elopmen
o T1DM was he mos o en men ioned eason. Among
he mino i y o pa en s who had conside ed lea ing he
s udy, he blood d aw, being oo busy/no ha ing enough
ime, a demanding p o ocol, and ood dia ies we e he mos
equen ly epo ed easons o conside ing lea ing [16].