University of Wollongong

Research Online

Illawarra Health and Medical Research Institute Faculty of Science, Medicine and Health

1-1-2017

Correlates of tummy time in infants aged 0-12 months old: A systematic

review
Lyndel Hewitt
University of Wollongong, [email protected]

Rebecca M. Stanley
University of Wollongong, [email protected]

Anthony D. Okely
University of Wollongong, [email protected]

Follow this and additional works at: https://ro.uow.edu.au/ihmri

Part of the Medicine and Health Sciences Commons

Recommended Citation
Hewitt, Lyndel; Stanley, Rebecca M.; and Okely, Anthony D., "Correlates of tummy time in infants aged 0-12
months old: A systematic review" (2017). Illawarra Health and Medical Research Institute. 1143.
https://ro.uow.edu.au/ihmri/1143

Research Online is the open access institutional repository for the University of Wollongong. For further information
contact the UOW Library: [email protected]
Correlates of tummy time in infants aged 0-12 months old: A systematic review

Abstract
Background: Tummy time, defined as an infant being placed on their stomach whilst they are awake and
supervised, has been shown to have a positive effect on infant development and head shape. Tummy
time can be influenced by a number of factors. Using a social ecological model, categories of potential
variables can be examined to determine their influence on behaviours such as tummy time. The purpose
of this systematic review was to examine potential correlates of tummy time in infants from birth to 12
months old.

Methods: Electronic databases were originally searched between March to December 2016. Included
studies needed to be peer-reviewed, written in English, and meet a priori study criteria. The population
was apparently healthy infants aged from birth to 12 months old. The article needed to contain an
objective or subjective measure of tummy time as a dependent variable and examine the association
between a demographic, psychological, behavioral, and/or environmental variable and tummy time. For
this study, tummy time could include the ability of the infant to move whilst being positioned on their
stomach, for example, the infant's ability to roll from back to front, or lift their head when lying on their
stomach (prone positioning ability), or the capacity, time spent, age started, or parent attitudes/
behaviours regarding the infant being placed on their stomach. The outcomes were the relationships
between potential correlates and tummy time. Risk of bias was assessed at the individual study level
using the Cochrane risk of bias assessment for observational studies.

Results: 15 articles representing 2372 unique participants from 7 countries were included. Correlates that
were positively correlated with tummy time were age, prone sleeping, spending greater than 15 minutes
whilst awake in tummy time when 2 months old, amount of time in the bath, order of achievement of
prone extension and prone on elbow positions and parents/carers setting aside time for tummy time. Risk
of bias of the included studies ranged from low to high.

Conclusions: Specific demographic, environmental and behavioral variables were found to be positively
and negatively associated with tummy time. This evidence could assist future research regarding
interventions to promote tummy time, enhance motor development, increase infant physical activity and
contribute to future tummy time recommendations for parents and health care providers.

Disciplines
Medicine and Health Sciences

Publication Details
Hewitt, L., Stanley, R. M. & Okely, A. D. (2017). Correlates of tummy time in infants aged 0-12 months old:
A systematic review. Infant Behavior and Development, 49 310-321.

This journal article is available at Research Online: https://ro.uow.edu.au/ihmri/1143

 1 Correlates of tummy time in infants aged 0 to 12 months old: A systematic review

2 Lyndel Hewitt1, Rebecca M. Stanley1, Anthony D. Okely1

3 Main Affiliations
1
4 Early Start, Faculty of Social Sciences and Illawarra Health and Medical Research Institute,
5 University of Wollongong, Wollongong, New South Wales, 2522, Australia. Email: Lyndel
6 Hewitt, [email protected]; Dr Rebecca Stanley, [email protected]; Professor
7 Anthony Okely, [email protected]
8
9 Corresponding Author
10 Lyndel Hewitt
11 Early Start
12 Faculty of Social Sciences,
13 University of Wollongong
14 Wollongong, NSW, Australia, 2500
15 E-mail: [email protected]
16
17 PROSPERO 2017: CRD42016036931. Available from:
18 http://www.crd.york.ac.uk/PROSPERO
19
20 Word count
21 3887
22
23
24 Abstract

25 Background: Tummy time, defined as an infant being placed on their stomach whilst they are

26 awake and supervised, has been shown to have a positive effect on infant development and

27 head shape. Tummy time can be influenced by a number of factors. Using a social

28 ecological model, categories of potential variables can be examined to determine their

29 influence on behaviours such as tummy time. The purpose of this systematic review was to

30 examine potential correlates of tummy time in infants from birth to 12 months old.

31 Methods: Electronic databases were originally searched between March to December 2016.

32 Included studies needed to be peer-reviewed, written in English, and meet a priori study

33 criteria. The population was apparently healthy infants aged from birth to 12 months old. The

1
34 article needed to contain an objective or subjective measure of tummy time as a dependent

35 variable and examine the association between a demographic, psychological, behavioural,

36 and/or environmental variable and tummy time. For this study, tummy time could include the

37 ability of the infant to move whilst being positioned on their stomach, for example, the

38 infant’s ability to roll from back to front, or lift their head when lying on their stomach (prone

39 positioning ability), or the capacity, time spent, age started, or parent attitudes/behaviours

40 regarding the infant being placed on their stomach. The outcomes were the relationships

41 between potential correlates and tummy time. Risk of bias was assessed at the individual

42 study level using the Cochrane risk of bias assessment for observational studies.

43 Results: 15 articles representing 2372 unique participants from 7 countries were included.

44 Correlates that were positively correlated with tummy time were age, prone sleeping,

45 spending greater than 15 minutes whilst awake in tummy time when 2 months old, amount of

46 time in the bath, order of achievement of prone extension and prone on elbow positions and

47 parents/carers setting aside time for tummy time. Risk of bias of the included studies ranged

48 from low to high.

49 Conclusions: Specific demographic, environmental and behavioural variables were found to

50 be positively and negatively associated with tummy time. This evidence could assist future

51 research regarding interventions to promote tummy time, enhance motor development,

52 increase infant physical activity and contribute to future tummy time recommendations for

53 parents and health care providers.

54

55 Keywords: tummy time, prone positioning, infant, correlate, behavioural, demographic,

56 environmental, variable, physical activity, motor development

57

2
58 Background

59 Tummy time, defined as awake and supervised positioning on the stomach, is included in the

60 National Academy of Medicine (IOM, 2011) and both the Canadian (Tremblay et al., 2012)

61 and Australian Early Years (Australian Government Department of Health, 2014) physical

62 activity recommendations for infants. As tummy time has been included in these

63 recommendations it can be assumed that it is an important component of physical and motor

64 development in infancy. These recommendations suggest that tummy time should be

65 provided daily to an infant less than 6 months of age. Identifying factors that influence

66 tummy time is therefore important in assisting parents/carers, health professionals, and early

67 childhood educators meet these guidelines

68

69 Tummy time provides an opportunity for the infant to stimulate and enhance their motor

70 development. Infants can be placed on their tummy from birth for short periods of supervised

71 play. When an infant is on their tummy they are given the opportunity to practice lifting up

72 their head, lifting up and turning their head, moving their legs and pushing up with their

73 arms. Tummy time strengthens the infant’s head, neck, shoulder and trunk muscles they

74 will need to master motor skills such as rolling, sitting, crawling and pushing up to sit.

75 There are some studies that have demonstrated a positive effect between tummy time and

76 motor development (Russell et al., 2009, Salls et al., 2002b, Majnemer and Barr,

77 2005, Monson et al., 2003, Dudek-Shriber and Zelazny, 2007, Salls et al., 2002a).

78 However, studies that have explored factors that influence tummy time are limited. Some

79 potential examples of tummy time correlates may be age, sex, sleeping position, type of

80 positioning and handling from carer, home set up, amount of time placed prone, low birth

81 weight, gestational age, mental health issue of the carer and tolerance by the infant. In

82 addition, studies that investigate an infant’s ability to move when on their stomach (prone

3
 83 positioning ability) have not been systematically reviewed. This could include the ability to

84 roll from front to back, ability to lift their head, ability to push up with their arms, and ability

85 to move their arms and/or legs, Combining tummy time and prone positioning ability in the

86 search strategy will be important to ensure as many studies as possible are captured. A study

87 using the combination of these terms is yet to be conducted. As such, both the infant’s ability

88 to move in prone (prone positioning ability) and the infant’s capacity, time spent, age started,

89 or parent attitudes/behaviours regarding the infant being placed on their stomach will be

90 defined in this study as ‘tummy time’. A number of systematic reviews have been

91 conducted addressing the correlates of pre-school-aged children's physical activity

92 (Hinkley et al., 2008) and sedentary behaviour ( H i n k l e y e t a l . , 2 0 1 0 ) . In contrast,

93 reviews investigating correlates of infant behaviour or positioning practices are limited.

94 Identifying what influences tummy time will be important for the development of

95 evidence-based interventions. In addition, it will also highlight how these correlates relate

96 to infant health indicators. Therefore, the purpose of this systematic review is to examine the

97 correlates of objectively and subjectively measured tummy time in infants (aged 0 to 12

98 months) across observational study designs.

99

100 Methods

101 Protocol and Registration

102 This review was registered with the international prospective register of systematic reviews

103 PROSPERO network (http://www.crd.york.ac.uk/prospero/): Registration no.

104 CRD42016036931. This review followed the PRISMA statement for reporting systematic

105 reviews and meta-analyses (Moher et al., 2009).

106 Inclusion and Exclusion Criteria

4
107 For an article to be included in this review, it had to be peer-reviewed, published or in press,

108 written in English, and meet a priori determined population, intervention/exposure,

109 comparator/control, and outcome (PICO) study criteria (Schardt et al., 2007) from the

110 Grading of Recommendations Assessment, Development, and Evaluation (GRADE)

111 framework (Guyatt et al., 2011a, Guyatt et al., 2011b). Conference abstracts, book chapters,

112 and dissertations were excluded.

113 Population: The population was apparently healthy (i.e., general population, including

114 overweight/obese, but not studies that only included infants with a diagnosed medical

115 condition with the exception of studies relating to prematurity, sudden infant death syndrome

116 or low birth weight) infants from the ages of 0 to12 months. For studies using a longitudinal

117 design, the age criterion applied to at least one measurement time point during the study.

118 Observational studies and only the control group (i.e., not experienced any form of

119 intervention) from experimental studies were reviewed and were required to have a minimum

120 sample size of 20 participants. An article was included if it: (1) included human infants aged

121 from birth to 12 months old; (2) contained quantitative research and had been published in an

122 English-language, peer-reviewed journal; (3) contained a measure of tummy time and/or

123 prone positioning ability as a dependent variable (all defined in this study as tummy time);

124 (4) examined the association between a demographic, psychological, behavioral, and/or

125 environmental variable and tummy time.

126 Intervention (exposure): Tummy time could be measured objectively (e.g., direct

127 observation, validated measurement tool) or subjectively (e.g., proxy-report, questionnaire).

128 Comparator: Various levels of demographic (e.g., Age, gender), behavioral (e.g., Sleeping

129 position, type of positioning and handling from carer, tolerance by infant), environmental

5
130 (e.g., Home set up, amount of time placed prone), or psychological factors (e.g., Depression

131 or mental health issue of carer).

132 Outcomes (indicators): The outcomes were subjectively or objectively measured amount of

133 time spent prone or tummy time or stomach or abdomen or front or belly or position*, age at

134 which started tummy time and/or ability to move whilst on the stomach.

135 Information Sources and Search Strategy

136 Computerised searches were completed in April 2016 using MEDLINE, CINAHL, Scopus

137 and PsycINFO. A search top-up was conducted in April 2017 to capture any articles that

138 were not yet indexed in the search engines in April 2016. The following search terms were

139 used: “tummy time” OR “prone” OR “position*” OR “abdomen” OR “stomach” OR “belly”

140 OR “front” AND “correlate*” OR “determin*” OR “predictor*” OR “relationship*” OR

141 “associate*” OR “difference*” AND “infant* OR “baby” OR “babies” OR “newborn”. In

142 addition, studies from the author’s own libraries were also assessed for possible inclusion.

143 After duplicates were removed, two researchers independently reviewed the titles of the

144 articles to determine if they met the criteria for the systematic review. Abstract and full-text

145 articles were then referred to clarify and confirm eligibility. Any differences in articles

146 selected by the two researchers were discussed to reach a decision regarding inclusion.

147 Discrepancies that could not be resolved by the two independent reviewers were resolved by

148 discussions with a third reviewer. Reference lists of relevant reviews identified during

149 screening were also checked for relevant studies. To capture registered clinical trials, two

150 trial registries (https://clinicaltrials.gov/ and http://www.who.int/ictrp/en/) were searched in

151 May 2017 using search terms for tummy time and the infant age group.

152 Data Extraction

6
153 The data extracted included; authors name, publication year, country, study design, sample

154 size, characteristics of participants, tummy time measure and/or prone positioning ability

155 measure, the correlate and type of correlate and the risk of bias. A finding was deemed to be

156 statistically significant if p<0.05 was reported even if statistical significance was defined

157 differently in the article. One reviewer completed data extraction for each included article

158 and a second reviewer checked all data.

159 Quality Assessment

160 Risk of bias was assessed at the individual study level using the Cochrane risk of bias

161 assessment for observational studies (Higgins, 2011). Selection bias, performance bias,

162 selective reporting bias, detection bias, attrition bias, and other biases (e.g., inadequate

163 control for key confounders) were assessed (Guyatt et al., 2011c). For all studies, risk of bias

164 was assessed by one reviewer and checked by a second reviewer. Overall quality of evidence

165 was evaluated by one reviewer and verified by the larger review team.

166 Results

167 Description of studies

168 After de-duplication, 1840 titles, 466 abstracts and 41 full-text articles were screened (see

169 Figure 1). It was determined that 15 articles met the inclusion criteria. Reasons for excluding

170 articles are summarized in Figure 1.

171 The 15 articles involved 2372 participants from seven different countries. An experimental

172 study design was used in two articles; this included a randomized controlled trial (n=1) and a

173 non-randomized intervention (n=1). An observational study design was used in the remaining

174 13 articles, including longitudinal (n=6), prospective cross-sectional (1), prospective cohort

175 (1) and cross-sectional (n=5).

7
176 Time spent, tolerance of, age when first experienced and parent attitudes/knowledge of

177 tummy time was not measured objectively in any articles and subjectively in nine articles,

178 primarily by proxy-report questionnaire, log, or interview (Carmeli et al., 2009, Davis et al.,

179 1998, Hesketh et al., 2015, Jennings et al., 2005, Moir et al., 2016, Ricard and Metz, 2014,

180 Salls et al., 2002a, van Vlimmeren et al., 2007, Zachry and Kitzmann, 2011). The ability of

181 the infant to move whilst on the stomach was only measured objectively in seven articles,

182 primarily by validated assessment tools (e.g., prone AIMS scale, Chailey level of abilities

183 scale, prone position) (Bartlett and Fanning, 2003, Bell and Darling, 1965, Bridgewater and

184 Sullivan, 1999, Majnemer and Barr, 2006, Rocha and Tudella, 2008, Salls et al., 2002a) and

185 direct observation (Horowitz and Sharby, 1988). Further information on the study design,

186 sample size, tummy time outcome measure and correlates identified from each study are

187 summarized in Table 1. Rules for classifying the strength of the correlate to tummy time are

188 reported in Table 2. All correlates that are reported to have a positive or negative association

189 with tummy time were statistically significant (p<0.05) and are reported in Table 3.

190 Demographic variables

191 There were four demographic variables that correlated with tummy time from 10 articles

192 (Table 3). Age had a positive correlation with tummy time from six studies (Rocha and

193 Tudella, 2008, Majnemer and Barr, 2006, Hesketh et al., 2015, Carmeli et al., 2009,

194 Bridgewater and Sullivan, 1999, Salls et al., 2002a) and an unclear association in two studies

195 (Davis et al., 1998, Moir et al., 2016). Older parents and low parent education level was

196 found to have a negative correlation (van Vlimmeren et al., 2007, Majnemer and Barr, 2006).

197 One third of the studies investigating a demographic variable had a high risk of bias (Table

198 4).

199 Behavioral variables

8
200 There were 16 behavioral variables that correlated with tummy time from 10 articles (Table

201 3). Prone sleeping (Majnemer and Barr, 2006, Davis et al., 1998, Salls et al., 2002a), the

202 order of achievement of prone extension and prone on elbows position (Horowitz and

203 Sharby, 1988) and parents setting aside time for tummy time (Ricard and Metz, 2014) were

204 all positively correlated with tummy time. Interestingly, knowledge, a fearful attitude (Ricard

205 and Metz, 2014) and receiving information from a pediatrician (Jennings et al., 2005) about

206 tummy time had no effect. The frequency and duration of hand-mouth behaviors decreased as

207 the ability to move whilst on the stomach improved (Rocha and Tudella, 2008). Despite these

208 findings, almost half of the studies that had a behavioral variable had a high risk of bias

209 (Table 4).

210 Environmental variables

211 There were 15 environmental variables that correlated with tummy time from four studies

212 (Table 3). Among these studies, spending greater than 15 minutes in tummy time at two

213 months of age (Salls et al., 2002a) and amount of time in the bath (Bridgewater and Sullivan,

214 1999) was positively correlated with tummy time. Amount of time spent awake supine

215 (Bridgewater and Sullivan, 1999) was negatively correlated. Equipment and minutes spent

216 exercising had no effect (Bridgewater and Sullivan, 1999, Bartlett and Fanning, 2003).

217 Interestingly, time spent in tummy time (minutes per day) at 4 and 6 months had an

218 indeterminate effect on the ability to move whilst on the stomach, with one study reporting a

219 significant positive effect (Majnemer and Barr, 2006) and the other reporting no effect (Salls

220 et al., 2002a). Both were longitudinal studies with less than 100 participants. However they

221 used different assessment tools (AIMS prone subscale vs Denver II Gross Motor Sector) and

222 overall the risk of bias for Salls, Silverman et al. 2002 was high whereas it was moderate for

223 Majnemer and Barr 2006. Approximately, almost half of the studies with an environmental

224 variable had a high risk of bias (Table 4).

9
225

226 Discussion

227 In this systematic review, evidence from 15 articles were synthesized to examine the

228 correlates of tummy time in infants aged from birth to 12 months old. From these

229 observational studies and control groups from experimental studies, age and prone sleeping

230 has the strongest positive correlation with tummy time. It is not unexpected that as an infant

231 grows older, their tummy time practices improve. However, this information can provide

232 evidence for health professionals and encouragement to parents who report that their infant

233 does not enjoy tummy time when they first begin to experience it. Fifty percent of parents

234 from the study completed by Ricard and Metz 2014 reported that their infant cried,

235 rolled/squirmed or appeared frustrated during tummy time. Anecdotally, health professionals

236 assist parents to provide tummy time to their infants a few minutes at a time and gradually

237 increase the demand and duration. The knowledge that tummy time improves, as the baby

238 gets older can be a powerful tool in the early stages to persevere. Prone sleepers also had a

239 positive correlation with tummy time. There was no indication from studies that had the

240 correlate of prone sleeping as to why parents were not complying with the back to sleep

241 recommendations. One study, even gave parents brochures and advised them to adhere to

242 supine sleep positioning according to the American Academy of Pediatrics recommendation

243 prior to entry into the study (Davis et al., 1998). However, as the sample size for the prone

244 sleeping groups was smaller than the supine sleeping group it can be suggested that the

245 majority of those enrolled in these studies were complying with the recommendations. The

246 number of parents in this sample of participants who did not follow the sleep

247 recommendations was consistent with other studies that found that approximately one third of

248 parents who are aware of the recommendations continue to put their babies prone to sleep

249 (Taylor and Davis, 1996, Rainey and Lawless, 1994). Despite this, parents should be

10
250 encouraged to adhere to the ‘back to sleep’ campaign recommendation (AAP, 1992). For

251 safety, increasing the amount of time prone whilst the infant is awake and supervised would

252 be recommended rather than changing the infant sleeping position. This view is supported by

253 Pin, Eldridge et al. 2007 who reported that it is important to educate parents to continue

254 placing their baby to sleep supine but to change their position during play time when they are

255 awake (Pin et al., 2007).

256

257 The frequency and duration of hand-mouth behaviors decreased as the ability to move when

258 on the stomach improved (Rocha and Tudella, 2008). To explain this, Rocha and Tudella

259 2008 suggest that as infants start to use their arms for support in prone they begin to visually

260 explore their environment around them rather than exploring just their own body. Order of

261 achievement of a prone extension position was reported by one study to be head extension,

262 then leg extension, then arm extension and the prone on elbows position to be head extension

263 then leg extension (Horowitz and Sharby, 1988). This correlated positively with the infant’s

264 ability to move when on the stomach. This information could be helpful to Physiotherapists

265 and Occupational therapists assisting infants with motor development delay. Motor

266 development interventions could be structured to achieving head extension, then leg

267 extension and then arm extension. This is not to say that therapists cannot progress until the

268 first one is achieved, but that motor development training could be ordered and progressed as

269 tolerated by the infant. This information could also be helpful to parents. Being aware of the

270 stages of achieving tummy time may assist in relieving the pressure of achieving “text book”

271 tummy time (i.e. Head up, arms extended, happy baby on tummy) immediately or in the first

272 few attempts.

273

11
274 Amount of time in the bath (Bridgewater and Sullivan, 1999) and setting aside time for

275 tummy time (Ricard and Metz, 2014) were also positively correlated to tummy time. To our

276 knowledge, this is the first link between bath time and tummy time. This may be a result of

277 bath time promoting positive interactions between parents and infants, however this finding

278 requires further investigation. The infant’s position in the bath was not mentioned however

279 increased time in the bath was associated with more mature responses from the infant of

280 being able to lift their head in prone (p<0.0001) (Bridgewater and Sullivan, 1999). Having

281 the attitude of “setting aside time for my baby to spend on his/her tummy is important”

282 correlated with setting aside time for awake prone positioning (p<0.01) (Ricard and Metz,

283 2014). The most common factor influencing this decision for parents in this study was

284 “helping their infants develop”. Ricard and Metz 2014 suggest providing education to parents

285 on how, when, and the significance of setting aside time for tummy time may improve daily

286 practice of tummy time.

287

288 Variables that were detrimental to tummy time were found to be amount of time spent supine

289 whilst awake (Bridgewater and Sullivan, 1999), having older parents (Majnemer and Barr,

290 2006), having less educated parents (van Vlimmeren et al., 2007) and spending less than 15

291 minutes per day at 2 months of age in awake and supervised tummy time (Salls et al., 2002a).

292 From this information, health professionals could be extra vigilant in their tummy time

293 instructions with parents who are older and/or less educated. Parents could also be informed

294 that a minimum of 30 minutes per day spent in awake and supervised tummy time when the

295 infant is two months old is a positive step towards starting tummy time. In addition, avoiding

296 prolonged supine positioning could be beneficial to the infant’s motor development. Some of

297 the risk factors for deformational plagiocephaly at 7 weeks of age is experiencing tummy

298 time less than 3 times per day and slow achievement of motor milestones (van Vlimmeren et

12
299 al., 2007). This information could be combined with the results from this study to assist

300 parents with a more prescriptive approach to tummy time. For example, when your baby is

301 two months old, they could be receiving tummy time more than 30 minutes per day. This

302 could be broken up into small amounts (for example, more than 3 times per day) adding up to

303 more than 30 minutes. Proclaiming a specific goal to reach could be helpful to assist parents

304 to determine if their baby is getting enough tummy time and exposes tools or equipment that

305 would assist to meet it. For example, using a timer or diary to record their sessions, having a

306 space, play mat and suitable toys ready.

307

308 A number of research gaps and limitations to address in future research also warrant

309 attention. For instance, as all included articles were observational studies, they lack the rigor

310 of a randomized controlled trial and will all score high on risk of bias. The final outcomes

311 found in this study (positive and negative correlates described in Table 3) are drawn from

312 only 11 studies. As such, findings from this review should be interpreted with caution. In

313 addition, there were no objective measures of the time spent in tummy time. All were based

314 on parent questionnaires or position logs. Future research into objective measures of tummy

315 time using real time measurement devices is yet to be conducted. The majority of studies

316 (75%) had a high selection bias. As such, information from these studies may not be

317 generalizable to other cultures and/or socio economic groups. As there were no psychological

318 variables found, further research could be conducted examining the effect of depression or

319 mental health issue of the parent or carer with the aim to further target populations more in

320 need of specific interventions. As tummy time and prone positioning ability were used as

321 combined terms in this study, it is important to note that the correlates found are from studies

322 investigating tummy time and/or prone positioning ability, further analysis would be required

323 to separate out these two terms. Lastly, having English language limits for feasibility was

13
324 also a limitation as it is possible that studies published in other languages may have provided

325 additional correlates not discovered by this review.

326

327 Conclusions

328 This review synthesized low quality evidence from 15 studies on the correlates of tummy

329 time. Age, prone sleeping, greater than 15 minutes daily of tummy time at two months,

330 amount of time in the bath, order of achievement of prone extension and prone on elbows

331 position, parent education level and setting aside time for tummy time were all positively

332 correlated. Time spent supine, age of the parent and duration and frequency of hand mouth

333 behaviors were all negatively correlated to tummy time. This information could be used to

334 assist health professionals target intervention groups and specify intervention techniques.

335 Good quality studies would be beneficial to strengthen the evidence base and inform future

336 research aimed at improving motor development and physical activity for infants.

14
337 List of Abbreviations

338 IOM: Institute of Medicine; PROSPERO: International Prospective Register of Systematic

339 Reviews; PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses;

340 PICO: Population Intervention Control Outcome; GRADE: Grading of Recommendations,

341 Assessment, Development, and Evaluation; MEDLINE: Medical Literature Analysis and

342 Retrieval System Online; CINAHL: Cumulative Index of Nursing and Allied Health

343 Literature; Scopus: Bibliographic database for academic journal articles; PsycINFO:

344 Psychological Information Database; HRE: Head righting - extension; AWBTS: Active

345 weight bearing through shoulders; AIMS: Alberta Infant Motor Scale; AAP: American

346 Academy of Pediatrics.

347

348 Declarations

349 Ethics Approval and Consent to Participate: Not applicable.

350 Consent for Publication: Not applicable.

351 Availability of Data and Material: Not applicable.

352 Competing Interests: The authors declare that they have no competing interests.

353 Funding: This research has been conducted with the support of the Australian Government

354 Research Training Program Scholarship awarded to LH.

355 Authors’ Contributions: LH led the design and coordination of the review. ADO conceived

356 of the study. ADO, RS and LH conducted the screening of the records, and appraised the

357 quality of evidence. LH extracted the data, analyzed, interpreted the data and drafted the

358 manuscript. All authors were responsible for revising the manuscript critically for important

359 intellectual content. All authors read and approved the final manuscript.

15
360 Acknowledgements: The authors would like to thank the University of Wollongong library

361 staff for assistance with EndNote, database and search strategy education.

16
Table 1. Descriptive information of included studies (Ordered alphabetically)

Author Country Sample Sex Age at Design Tummy time outcome Correlates of tummy time Type of correlate (Social
size (B, G) measurement measure identified Ecological Framework
Domain Association)
(Bartlett and Canada 60 28, 32 8.08 months Observational Prone AIMS subscale - Amount of equipment use - Environmental
Fanning, 2003) (corrected) (jolly jumper, walker,
exersaucer, seat, swing,
backpack, carried, other, total
equipment), n.s

(Bell and Darling, USA 75 41, 34 Birth to 4 days Observational Prone Head Reaction (PHR) - Sex, n.s - Demographic
1965) old by an 11 point scale - Method of feeding (Breast or - Behavioural
bottle fed)

(Bridgewater and Australia 26 13, 13 14 to 18 weeks Observational Movement Assessment of - Age, p<0.001 - Demographic
Sullivan, 1999) Infants (MAI): Head righting - Bath time, p<0.001 - Environmental
(Extension) and active - Amount of time spent supine, - Environmental
weight bearing through negative correlation, p<0.05 - Environmental
shoulders - Exercise, n.s. - Environmental
- Capsule/cuddle, n.s - Environmental
(Carmeli et al., Israel 80 80, 0 Birth to 26 Longitudinal Position log completed by - Age, p=0.03 - Demographic
2009) weeks parents - AIMS percentile, prone - Behavioural
subscale, n.s
(Davis et al., 1998) USA 400 49%, 51% 1 week to 6 Longitudinal Position log completed by - Prone sleeping (p <0.003) - Behavioural
months parents - Age, no p value given - Demographic
(Hesketh et al., Australia 542 285, 257 4 and 9 months Longitudinal Questionnaire to the Mother - Age, p<0.001 - Demographic
2015)
(Horowitz and USA 20 10, 6 8 to 28 weeks Longitudinal Direct observation, prone - Order of achievement of - Behavioural
Sharby, 1988) (every 2 weeks) positioning ability prone extension posture (head
and limb positions) was head,
lower extremity, upper
extremity (p<0.001)
- Order to achieve prone on
elbows position was head and
lower extremity (p<0.01)
- Upper extremity extension not
required to achieve prone on
elbows position (n.s)
- Prone on hands position not
correlated with head, lower or
upper extremity extension (n.s)
(Jennings et al., USA 27 Not given 6 months Non- Parent survey - Parent receiving positioning - Behavioural
2005) *control randomized information from the
group intervention paediatricians office, no p value
only

17
(Majnemer and Canada 72 32, 40 6 months Longitudinal - Prone AIMS subscale - Prone sleeping, p<0.005 - Behavioural
Barr, 2006) - Motor milestones (AIMS) - Prone sleeping, p<0.02 - Behavioural
(% achieved): Rolling prone
to supine
- Prone AIMS subscale - Mean daily exposure to prone - Environmental
position (minutes/day), p<0.01

(Majnemer and Canada 83 42, 41 4 months Longitudinal Prone AIMS subscale - Prone sleeping, p<0.002 - Behavioural
Barr, 2006) - Mean daily exposure to prone - Environmental
position (minutes/day), p<0.05
- Older parents, negative, - Demographic
p<0.01
- Age, p<0.0001 - Demographic

(Moir et al., 2016) New Zealand 209 98, 111 4 and 6 months Randomized Parent questionnaire - Age, no p value - Demographic
*control controlled trial
group
only
(Ricard and Metz, USA 87 Not 3 months Observational Parent questionnaire - Knowledge of prone - Behavioural
2014) provided positioning, p>0.05
- Fearful attitude towards prone - Behavioural
position, p>0.05
- Setting aside time for prone - Behavioural
positioning, p<0.01
(Rocha and Brazil 40 16, 24 Newborn, 1, 2, Prospective Chailey level of abilities - Frequency of hand-mouth - Behavioural
Tudella, 2008) 3 and 4 months cross-sectional scale, prone position behaviour, negative, p<0.001
study - Duration of hand-mouth - Behavioural
behaviour, negative, p=0.005
- Age, p<0.001 - Demographic
(Salls et al., 2002a) USA 66 Not 2, 4 and 6 Longitudinal Parent questionnaire - Age, no p value, unknown if - Demographic
provided months significant

Denver II Gross Motor - Awake time in prone >15mins - Environmental

Sector, (head up 45deg, head at 2 months old, p<0.05
up 90deg, chest up-arm - Awake time in prone <15 - Environmental
support minutes at 2 months old,
p<0.05
- Awake time in prone < or > - Environmental
15 minutes, 4 and 6 months, p>
0.05
- Sleeping position at 2 months - Behavioural
- Sleeping position at 4 months - Behavioural
(van Vlimmeren et Netherlands 380 178, 202 7 weeks Prospective Parent questionnaire, gave - Low education level, - Demographic
al., 2007) cohort study their infant tummy time for negative, significant but no p
the first time at >=3 weeks of value given
age
(Zachry and USA 205 42%, 52% 2 weeks to 24 Observational Parent questionnaire - Tolerance of tummy time in - Behavioural

18
Kitzmann, 2011) months (range) minutes, ?, no p value
- Intolerance of tummy time in - Behavioural
minutes, ? no p value
- Caregiver awareness of - Behavioural
tummy time, ? no p value

19
Table 2. Rules for classifying variables regarding strength of association with tummy time

Studies supporting association (%) Summary code Explanation of code

0-33 0 No association
34-59 ? Indeterminate/inconclusive association
60-100 + Positive association
60-100 – Negative association
Note: When an outcome was found four or more times, it was coded as: 00 (no association); ?? (indeterminate); ++ (positive association); or – – (negative association) (Tonge et al., 2016)

20
Table 3. Summary of reported correlates – tummy time

Correlate Investigated an association Association (0, ?, – Summary coding for Summary code for
with tummy time or +) studies with an association
(reference) association (n/N; %) (– /+)
Demographic variables
Age (Rocha and Tudella, 2008) +, p<0.001 6/9 (67%) ++
(Majnemer and Barr, 2006) +, p<0.0001
(Hesketh et al., 2015) +, p<0.001
(Davis et al., 1998) ?, no p value given
(Carmeli et al., 2009) +, p=0.03
(Bridgewater and Sullivan, 1999) +, p<0.001 (HRE)
(Bridgewater and Sullivan, 1999) +, p<0.01 (AWBTS)
(Moir et al., 2016) ?, no p value given
(Salls et al., 2002a) ?, no p value given,
unknown if significant
Male infant (Bell and Darling, 1965) 0, n.s but no p value given 1/2 (50%) ?
(Bell and Darling, 1965) +, p<0.01
Low parent education level (van Vlimmeren et al., 2007) -, significant but no p value 1/1(100%) –
given
Older parents (Majnemer and Barr, 2006) -, p<0.01 1/1, (100%) –

Behavioural variables
Method of feeding (breast or (Bell and Darling, 1965) ? 1/1 (100%) ?
bottle)

AIMS percentile, prone subscale (Carmeli et al., 2009) 0, n.s 0/1, (0%) 0
Prone sleeping (Majnemer and Barr, 2006) +, p<0.002 5/6 (83%) ++
(Majnemer and Barr, 2006) +, p<0.02
(Majnemer and Barr, 2006) +, p<0.02
(Davis et al., 1998) +, p<0.003
(Salls et al., 2002a) +, p<0.05, 2 months old
(Salls et al., 2002a) 0, p>0.05, 4 months old
Order of achievement of prone Horowitz and Sharby 1998 +, p<0.001 1/1 (100%) +
extension posture (head and limb
positions) was head, lower
extremity, upper extremity

Order to achieve prone on elbows Horowitz and Sharby 1998 +, p<0.01 1/1 (100%) +
position was head and lower
extremity

Upper extremity extension not Horowitz and Sharby 1998 0, n.s 0/1 (0%) 0
required to achieve prone on
elbows position

21
Prone on hands position not Horowitz and Sharby 1998 0, n.s 0/1 (0%) 0
correlated with head, lower or
upper extremity extension

Parent receiving positioning (Jennings et al., 2005) 0, no p value 0/1 (0%) 0

information from the
paediatricians office
Knowledge of prone positioning (Ricard and Metz, 2014) 0, p>0.05 0/1 (0%) 0
Fearful attitude towards prone (Ricard and Metz, 2014) 0, p>0.05 0/1 (0%) 0
position
Setting aside time for prone (Ricard and Metz, 2014) +, p<0.01 1/1 (100%) +
positioning
Frequency of hand-mouth (Rocha and Tudella, 2008) -, p<0.001 1/1 (100%) –
behaviour, negative

Duration of hand-mouth (Rocha and Tudella, 2008) -, p=0.005 1/1 (100%) –

behaviour

Tolerance of tummy time in (Zachry and Kitzmann, 2011) ?, no p value 1/1 (100%) ?
minutes

Intolerance of tummy time in (Zachry and Kitzmann, 2011) ?, no p value 1/1 (100%) ?
minutes

Caregiver awareness of tummy (Zachry and Kitzmann, 2011) ?, no p value 1/1 (100%) ?
time

Environmental variables
Awake time in prone >15mins at 2 (Salls et al., 2002a) +, p<0.05 1/1 (100%) +
months old

Mean daily exposure to prone (Majnemer and Barr, 2006) +, p<0.05 2/4 (50%) ?
position (minutes/day), 4 months

Mean daily exposure to prone (Majnemer and Barr, 2006) +, p<0.01

position (minutes/day), 6 months

Awake time in prone (< or > (Salls et al., 2002a) 0, p>0.05

15minutes per day), 4 months

Awake time in prone (< or > (Salls et al., 2002a) 0, p>0.05

15minutes per day), 6 months

Amount of time spent supine (Bridgewater and Sullivan, 1999) -, p<0.05 (AWBTS) 1/1 (100%) –

Amount of time in the bath (Bridgewater and Sullivan, 1999) +, p<0.001 1/1 (100%) +

22
Minutes spent experiencing (Bridgewater and Sullivan, 1999) 0, n.s 0/1 (0%) 0
exercise

Amount of time in capsule/cuddle (Bridgewater and Sullivan, 1999) 0, n.s 0/1 (0%) 0

Amount of time in jolly jumper (Bartlett and Fanning, 2003) 0, n.s 0/1 (0%) 0

Amount of time in walker (Bartlett and Fanning, 2003) 0, n.s 0/1 (0%) 0

Amount of time in exersaucer (Bartlett and Fanning, 2003) 0, n.s 0/1 (0%) 0

Amount of time in seat (e.g., (Bartlett and Fanning, 2003) 0, n.s 0/1 (0%) 0
highchair, infant seat, bouncer
seat, car seat – other than for
meals)

Amount of time in swing (Bartlett and Fanning, 2003) 0, p=0.24 *excludes outlier 0/1 (100%) 0

Amount of time in backpack (Bartlett and Fanning, 2003) 0, n.s 0/1 (0%) 0

Amount of time carried (Bartlett and Fanning, 2003) 0, n.s 0/1 (0%) 0

Amount of time in other (Bartlett and Fanning, 2003) 0, n.s 0/1 (0%) 0
equipment not mentioned above
Amount of time use total (Bartlett and Fanning, 2003) 0, n.s 0/1 (0%) 0
equipment
n.s: not significant

23
Table 4. Risk of bias of included studies

Author Were the participants Did an adequate Did the study report Did the study report Did the study have Other sources of
likely to be proportion of those the sources and details the sources and details incomplete or absent bias
representative of the consenting to of the type of tummy of the type of correlate reporting of some
chosen population? participate in the time measurement tool measurement tool used outcomes and not
(Selection bias) study have complete used in the study? in the study? AND did others on the basis of
data (i.e. no more than AND did the study the study report the results? (Selective
20% of data missing report adequate adequate reliability reporting bias)
from a cross sectional reliability and/or and/or validity of this
study and no more validity of this measurement tool used
than 30% for a measurement tool used in the study
longitudinal study) in the study (Detection (Performance bias)
(Attrition bias) bias)

(Bartlett and Fanning, High Low High Low Low Low

2003)

(Bell and Darling, High High Low Low High Low

1965)

(Bridgewater and High Low Low High Low Low

Sullivan, 1999)

(Carmeli et al., 2009) High Low Low High Low High

(Davis et al., 1998) High Low High Low Low Low

(Gajewska and High Low Low Low Low Low

Sobieska, 2015)

(Hesketh et al., 2015) Low Low Low Low Unclear Low

(Horowitz and Sharby, High Low Low Low Low Low

1988)

(Jennings et al., 2005) High High High Low Low Low

(Majnemer and Barr, High Low Low High Low Low

2006)

(Moir et al., 2016) Low Low High Low Low Low

(Ricard and Metz, High High High High Low Low

2014)

(Rocha and Tudella, Unclear Low Low Low Low Unclear

24
2008)

(Salls et al., 2002a) High High High Low High High

(van Vlimmeren et al., High Low High Low Low Unclear

2007)

(Wen et al., 2011) Low Low High Low Low High

(Zachry and High High High High High High

Kitzmann, 2011)

25
References

AAP 1992. American Academy of Pediatrics Task Force on infant positioning and SIDS.
Pediatrics, 89, 1120-1126.
Australian Government Department of Health 2014. National Physical Activity
Recommendations for Children 0-5 years. Australian Government Department of
Health, viewed 14 October 2017,
http://www.health.gov.au/internet/main/publishing.nsf/content/health-pubhlth-
strateg-phys-act-guidelines - npa05.
Bartlett, D. J. & Fanning, J. E. K. 2003. Relationships of equipment use and play positions
to motor development at eight months corrected age of infants born preterm.
Pediatric Physical Therapy, 15, 8-15 8p.
Bell, R. Q. & Darling, J. F. 1965. The prone head reaction in the human neonate: relation
with sex and tactile sensitivity. Child Development, 36, 943-949.
Bridgewater, K. J. & Sullivan, M. J. 1999. Wakeful positioning and movement control in
young infants: a pilot study. Australian Journal of Physiotherapy, 45, 259-266 8p.
Carmeli, E., Marmur, R., Cohen, A. & Tirosh, E. 2009. Preferred sleep position and gross
motor achievement in early infancy. European Journal Of Pediatrics, 168, 711-
715.
Davis, B. E., Moon, R. Y., Sachs, H. C. & Ottolini, M. C. 1998. Effects of sleep position on
infant motor development. Pediatrics, 102, 1135-1140.
Dudek-Shriber, L. & Zelazny, S. 2007. The effects of prone positioning on the quality and
acquisition of developmental milestones in four-month-old infants. Pediatric
Physical Therapy, 19, 48-55.
Gajewska, E. & Sobieska, M. 2015. Qualitative elements of early motor development that
influence reaching of the erect posture. A prospective cohort study. Infant
Behavior and Development, 39, 124-130.
Guyatt, G., Oxman, A. D., Akl, E. A., Kunz, R., Vist, G., Brozek, J., Norris, S., Falck-Ytter, Y.,
Glasziou, P. & Jaeschke, R. 2011a. GRADE guidelines: 1. Introduction—GRADE
evidence profiles and summary of findings tables. Journal of clinical
epidemiology, 64, 383-394.
Guyatt, G., Oxman, A. D., Sultan, S., Glasziou, P., Akl, E. A., Alonso-Coello, P., Atkins, D.,
Kunz, R., Brozek, J. & Montori, V. 2011b. GRADE guidelines: 9. Rating up the
quality of evidence. Journal of clinical epidemiology, 64, 1311-1316.
Guyatt, G., Oxman, A. D., Vist, G., Kunz, R., Brozek, J., Alonso-Coello, P., Montori, V., Akl, E.
A., Djulbegovic, B. & Falck-Ytter, Y. 2011c. GRADE guidelines: 4. Rating the
quality of evidence—study limitations (risk of bias). Journal of clinical
epidemiology, 64, 407-415.
Hesketh, K. D., Crawford, D. A., Abbott, G., Campbell, K. J. & Salmon, J. 2015. Prevalence
and stability of active play, restricted movement and television viewing in
infants. Early Child Development and Care, 185, 883-894.
Higgins, J., Green S 2011. Cochrane handbook for systematic review of interventions, West
Sussex, England, John Wiley & Sons.
Hinkley, T., Crawford, D., Salmon, J., Okely, A. D. & Hesketh, K. 2008. Preschool children
and physical activity: a review of correlates. American journal of preventive
medicine, 34, 435-441. e7.
Hinkley, T., Salmon, J., Okely, A. D. & Trost, S. G. 2010. Correlates of sedentary
behaviours in preschool children: a review. International Journal of Behavioral
Nutrition and Physical Activity, 7, 1.

26
Horowitz, L. & Sharby, N. 1988. Development of prone extension postures in healthy
infants. Physical Therapy, 68, 32-36 5p.
IOM 2011. Early Childhood Obesity Prevention Policies. Goals, Recommendations, and
Potential Actions. Institute of Medicine.
Jennings, J. T., Sarbaugh, B. G. & Payne, N. S. 2005. Conveying the message about optimal
infant positions. Physical & Occupational Therapy in Pediatrics, 25, 3-18 16p.
Majnemer, A. & Barr, R. G. 2005. Influence of supine sleep positioning on early motor
milestone acquisition. Developmental medicine & child neurology, 47, 370-376.
Majnemer, A. & Barr, R. G. 2006. Association between sleep position and early motor
development. Journal of Pediatrics, 149, 623-629 7p.
Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G. & Group, P. 2009. Preferred reporting
items for systematic reviews and meta-analyses: the PRISMA statement. PLoS
med, 6, e1000097.
Moir, C., Meredith-Jones, K., Taylor, B. J., Gray, A., Heath, A. L. M., Dale, K., Galland, B.,
Lawrence, J., Sayers, R. M. & Taylor, R. W. 2016. Early Intervention to Encourage
Physical Activity in Infants and Toddlers: A Randomized Controlled Trial.
Medicine and Science in Sports and Exercise, 48, 2446-2453.
Monson, R. M., Deitz, J. & Kartin, D. 2003. The relationship between awake positioning
and motor performance among infants who slept supine. Pediatric physical
therapy, 15, 196-203.
Pin, T., Eldridge, B. & Galea, M. P. 2007. A review of the effects of sleep position, play
position, and equipment use on motor development in infants. Developmental
Medicine And Child Neurology, 49, 858-867.
Rainey, D. Y. & Lawless, M. R. 1994. Infant Positioning and SIDS: Acceptance of the
Nonprone Position Among Clinic Mothers. Clinical Pediatrics, 33, 322-324.
Ricard, A. & Metz, A. E. 2014. Caregivers' Knowledge, Attitudes, and Implementation of
Awake Infant Prone Positioning. Journal of Occupational Therapy, Schools, and
Early Intervention, 7, 16-28.
Rocha, N. A. C. F. & Tudella, E. 2008. The influence of lying positions and postural
control on hand-mouth and hand-hand behaviors in 0-4-month-old infants.
Infant Behavior and Development, 31, 107-114.
Russell, D. C., Kriel, H., Joubert, G. & Goosen, Y. 2009. Prone positioning and motor
development in the first 6 weeks of life. South African Journal of Occupational
Therapy, 39, 11-14.
Salls, J. S., Silverman, L. N. & Gatty, C. M. 2002. The relationship of infant sleep and play
positioning to motor milestone achievement. American Journal of Occupational
Therapy, 56, 577-580.
Schardt, C., Adams, M. B., Owens, T., Keitz, S. & Fontelo, P. 2007. Utilization of the PICO
framework to improve searching PubMed for clinical questions. BMC medical
informatics and decision making, 7, 16.
Taylor, J. A. & Davis, R. L. 1996. Risk factors for the infant prone sleep position. Archives
of Pediatrics and Adolescent Medicine, 150, 834-837.
Tonge, K. L., Jones, R. A. & Okely, A. D. 2016. Correlates of children's objectively
measured physical activity and sedentary behavior in early childhood education
and care services: A systematic review. Preventive Medicine, 89, 129-139.
Tremblay, M. S., LeBlanc, A. G., Carson, V., Choquette, L., Gorber, S. C., Dillman, C.,
Duggan, M., Gordon, M. J., Hicks, A., Janssen, I., Kho, M. E., Latimer-Cheung, A. E.,
LeBlanc, C., Murumets, K., Okely, A. D., Reilly, J. J., Spence, J. C., Stearns, J. A. &

27
 Timmons, B. W. 2012. Canadian physical activity guidelines for the early years
(aged 0-4 years). Applied Physiology, Nutrition and Metabolism, 37, 345-356.
van Vlimmeren, L. A., van der Graaf, Y., Boere-Boonekamp, M. M., L'Hoir, M. P., Helders,
P. J. & Engelbert, R. H. 2007. Risk factors for deformational plagiocephaly at birth
and at 7 weeks of age: a prospective cohort study. Pediatrics, 119, e408-e418.
Wen, L. M., Baur, L. A., Simpson, J. M., Rissel, C. & Flood, V. M. 2011. Effectiveness of an
early intervention on infant feeding practices and 'tummy time': a randomized
controlled trial. Archives of Pediatrics & Adolescent Medicine, 165, 701-707 7p.
Zachry, A. H. & Kitzmann, K. M. 2011. Caregiver awareness of prone play
recommendations. American Journal of Occupational Therapy, 65, 101-105 5p.

28