Infant muscle activity is modified by inclined environments during different styles of rolling

Abstract

Rolling is an important part of infant motor development and is largely affected by experiences and surroundings. The purpose of this study was to determine how inclined mechanical environments affect infant muscle activation during previously defined coordinated rolling movements. The results demonstrate that infant muscle utilization (age: 6.5 ± 0.7 months; 23 M/15 F) differs while achieving the same rolling movements on a flat surface compared to four inclined environments representing a range of inclines to which infants are commonly exposed. Thus, rolling milestone achievement on a firm flat surface is not indictive of the same ability to roll in inclined seated environments.

1. Introduction

Rolling is part of infant cognitive and motor development through interactions with others and by exploring their environment (Lobo et al., 2013). To achieve a roll, specific body and limb coordination is required to move supine-to-prone or prone-to-supine (Richter et al., 1989). Supine-to-prone rolling typically begins on firm flat surfaces between four to six months as gross motor control develops (Altmann & Hill, 2019; Colvin et al., 2014), however, infants begin learning skills needed to roll as soon as they are born (McGraw, 1941). Rolling encourages continued motor development leading to sitting, crawling, and walking (Altmann & Hill, 2019). Infants are constantly growing, causing muscle and fat redistribution as well as muscle strength and elasticity changes, altering motor milestone achievement daily (Van Dam et al., 2011). Additionally, infant musculoskeletal and motor development is affected by experiences, including environment (Hadders-Algra, 2018), which varies from being held in arms, lying on a firm flat surface, to spending time in various nursery products. Therefore, it is important to fully understand how infant rolling achievement throughout development is affected by their different environments.

Infants in the U.S. spend an average of 5.7 hours (range 0–16 hours) per day in inclined commercial infant products such as car seats and rockers (Little et al., 2019; Callahan, 1997). Most of these products are designed for use for no more than 30–60 minutes while supervised (Michalska et al., 2023). However, infants are often not continuously attended and/or restrained in these products for prolonged periods of time (Callahan, 1997; Freyne et al., 2014; Batra et al., 2015; Bamber et al., 2014). In addition, caregivers sometimes utilize inclined products, or independently incline their infants, in an effort to improve sleep (Paul et al., 2023; Bellaïche et al., 2017; American Academy of Pediatrics, 2020). The American Academy of Pediatrics set recommendations for safe sleep including a supine position, a firm flat separate sleep surface, and the avoidance of soft bedding, as environments with an incline or soft materials may increase the risk of sudden infant death syndrome (American Academy of Pediatrics, 2020). However, no recommendations have been made related to body position within inclined products for when an infant is awake, therefore it is important to understand how an infant’s movements, muscle activation, and motor development may be affected by an incline to determine what motor development risk factors and safety concerns may exist during awake hours.

Previous research found that excessive time in some nursery products can cause a disturbance in motor development (Michalska et al., 2023; Anitha, 2019) and certain product design features may allow infants to achieve a roll before they are able to do the same maneuver consistently on a firm flat surface (Wang et al., 2021). For infants rolling on a flat surface, Kobayashi et. al identified six coordinated movements based on the timing of stationary versus moving limbs (Fig. 1) (Kobayashi et al., 2016), and we have previously reported the corresponding muscle activation patterns (Siegel et al., 2023). Younger infants (5–7 months) are more likely to utilize coordinated movements with a greater number of stationary limbs compared to older infants (8–10 months) since younger infants require more support from the surface to achieve a roll (Kobayashi et al., 2016), indicating that the coordinated movements utilized to achieve a roll is based on motor development. Additionally, previous research from McGraw et al. defined four stages of infant rolling development on a flat surface, showing how rolling coordination changes with age (McGraw, 1941).

Fig. 1.

Six previously established coordinated movements of rolling as defined by Kobayashi et. al. (2016), figures and captions adapted and reproduced with permission. Illustrated rolls are to the infants left where the limbs are codified as ipsilateral arm (IA) and leg (IL) and contralateral arm (CA) and leg (CL) in the direction of the roll. The limbs marked with an asterisk (*) are stationary throughout the roll, the dark gray limbs initiate the roll, and the light gray limbs follow after the roll has already been initiated.

Our previous study found that muscle activation during rolling changes based on the inclined mechanical environment. At inclines of just 10°, the erector spinae muscle activation significantly increased, and the abdominal muscle activation significantly decreased when compared to the flat surface regardless of the coordinated movement utilized. However, the coordinated movements infants used were significantly different in the number of times each was chosen in the inclined environment compared to the flat surface (Siegel et al., 2024a). Additionally, a new coordinated movement, labeled movement O, only observed in the inclined environment and not on the flat surface alone was described where infants press their feet and shoulders into the device resulting in pelvis thrust and hyperextension of the thoracic spine to achieve a roll (Fig. 2) (Siegel et al., 2024a). This further indicates that motor milestone development on flat surface, particularly for rolling, does not directly translate to different environments that infants spend time in. While the muscle utilization in the different inclines has been explored for rolling achievement, no studies have examined how the muscle utilization required during the specific coordinated movements (Kobayashi et al., 2016) are impacted by an incline, which is needed to further our understanding of how infant products impact safety throughout different stages of motor development.

Fig. 2.

Depiction of coordinated movement O that was only observed in the device configurations. For movement O, infants press their feet and shoulders into the seat of the device, resulting in pelvic thrust and hyperextension of the thoracic spine. Once the trunk hyperextension occurs, infants lift their contralateral limbs to rotate, and then return the ipsilateral side of the body to the surface as they achieve a roll.

The purpose of this study was to determine how an infant’s muscle activation is affected by different device configurations for each coordinated movement of infant rolling via a secondary analysis on our existing data set. We hypothesized that each coordinated movement would have unique changes in muscle activation levels, dependent on the device configuration.

2. Material & methods

2.1. Participants

Thirty-eight healthy infants (6.5 ± 0.7 months; 23 M/15 F) were recruited as part of a previous study, where we performed a secondary data analysis on the data set provided (Siegel et al., 2024a). For this previous study, a two-sample a priori power analysis was performed based on normalized mean electromyography data of healthy infants in different positions (Siddicky et al., 2020a) and determined that n = 9 participants per test condition would be sufficient to produce significant results (1 −β=0.8; α=0.05). Therefore, infants were enrolled until this minimum threshold of n = 9 per testing condition was met for the previous study. Twenty-nine healthy infants (6.7 ± 0.6 months; 16 M/13 F) achieved at least one rolling movement, defined as the supine-to-lateral rotation of the roll (Siegel et al., 2023), and were included in our secondary analysis. Guardians provided informed assent before testing began.

2.2. Experimental procedure

Using a data set collected as part of a larger study (Siegel et al., 2024a), we performed a secondary analysis to compare the muscle activation of four muscle groups (erector spinae, hamstrings, abdominal muscles, and quadriceps) while infants rolled in five different mechanical environments, a firm foam playmat and four device configurations (Fig. 3). In this larger study, surface electromyography (EMG) electrodes (Delsys Trigno, Natick, MA) recorded bilateral muscle activity (2000 Hz) from the erector spinae (ES), hamstrings (HAM), abdominal muscles (AB), and quadriceps (QUAD). The electrode placement was determined based on adult surface EMG placement and previous studies using EMG electrodes on an infant population (Mannen et al., 2019; Siddicky et al., 2020b; Wang et al., 2020; Siddicky et al., 2021; Cram et al., 1998). The ES sensors were placed mid-back just below the distal ends of the scapula on either side of the spine, a finger-width from the spinous process. The QUAD sensors were placed midway between the knee and the anterior superior iliac spine aligned with the femur. The HAM sensors were placed directly posterior to the QUAD sensors. Finally, the AB sensors were placed approximately two finger widths apart just above the belly button. All sensors were wrapped in self-adherent bandage to ensure placement throughout testing. A GoPro camera recorded video of each trial to determine which previously defined coordinated movement was utilized (Siegel et al., 2024b). A medical grade pulse oximeter attached to the infant’s right foot monitored oxygen saturation to ensure safety (Hunt et al., 1999; Vold et al., 2012).

Fig. 3.

Depicts the device with the testing conditions of seatback angles of 0°, 10°, 18° or 28° where the 0° condition had a base angle of 0°, and the other three configurations had a base angle of 15°.

To evaluate the effects of different mechanical environments, infants were placed in a supine position on a firm, foam commercially available playmat and then in four different device configurations unrestrained, each for 5 minutes or until they achieved a complete supine-to-prone roll. The four device configurations featured firm support with thin soft-goods, designed to represent a range or inclines infants are exposed to. The device had seatback angles of 0°, 10°, 18° or 28° where the 0° condition had a base angle of 0°, and the other three configurations had a base angle of 15°. All configurations had small, angled side walls (Fig. 3). The configurations and design of the device chosen was based on common commercial infant products like bouncers, rockers, swings, and car seats (Mannen et al., 2023). In each condition, infants were encouraged to roll by their guardians and researchers through verbal cues or by offering them toys or food.

2.3. Data analysis

Raw EMG waveforms were analyzed using previously established methodology (Siegel et al., 2024a). In this methodology, the EMG was extracted and truncated to only include each rolling movement, defined as the supine-to-lateral portion of the roll (Siegel et al., 2023). For each infant a five-second resting period was also extracted during still lying in a supine position on the playmat for normalization since maximum voluntary isometric contractions, which are normally used in EMG analysis, are impossible to obtain in an infant population (Siegel et al., 2023). Using MATLAB, raw EMG waveforms were first assessed using a power-spectral analysis to exclude corrupted or missing data. Contamination from movement artifacts, electrocardiogram signals, and high frequency noise was reduced using a band-pass 4th order Butterworth filter between 35 and 500 Hz. To eliminate interference from nearby electronic sources, the EMG was notch-filtered at 60 Hz using a 4th order Butterworth filter. The EMG waveforms were then full-waved rectified, baselined, and subjected to a low-pass 4th order Butterworth filter (cutoff 50 Hz) to obtain the EMG envelope (Wang et al., 2021; Siegel et al., 2024a; Siddicky et al., 2020b; Wang et al., 2020; Siddicky et al., 2021). This filtering process was completed for each rolling movement and five-second resting period and then the muscle activation levels during the rolling movement were divided by the mean EMG of the resting period to compare between subjects. The mean magnitudes for each muscle group were then represented as a percent of the maximum values for the flat surface to compare between device configurations and determine the percent muscle activation level (Siegel et al., 2024a). If fewer than three of the previously defined coordinated movements were exhibited in a condition, that category was excluded from further analysis as the data set was not considered substantial. The mean muscle activation was then taken for each muscle group over the entirety of the rolling movement to determine the muscle utilization level to be low (<50 %), moderate (≥50 to <75 %), high (≥75 to ≤100 %), or elevated (>100 %). Elevated represented muscle utilization levels above what was exhibited on the flat surface (Siegel et al., 2024a).

2.4. Statistical analysis

Using the SPSS statistical package (SPSS Inc., version 26; Chicago, IL), differences between mean muscle activation of each device configuration were compared to the flat surface for movements A-F using a repeated-measures ANOVA with a Dunnett’s T3 post-hoc test (p < 0.05) for each muscle group and coordinated movement. Since coordinated movement O is only observed in the device configurations, a repeated-measures ANOVA with a Dunnett’s T3 post-hoc test (p < 0.05) compared each device configuration to each other for each muscle group. Overall, this resulted in twenty-four separate ANOVAs being performed, one for each combination of muscle group and coordinated movement type.

3. Results

3.1. Coordinated movements

Two-hundred and thirty-four rolling movements were analyzed by coordinated movement and testing configuration (Fig. 4). Fewer than 3 coordinated movements were identified for movements B and O in the 0° configuration, movement B in the 10° configuration, movements B and D in the 18° configuration, and movements A, B, C, and D in the 28° configuration and therefore excluded from further analysis.

Fig. 4.

Number and percent of each coordinated movement achieved in the different testing conditions.

3.2. Muscle activation

The erector spinae muscles exhibited differing trends in muscle activation for each coordinated movement when comparing the flat surface to the different seatback configurations (Table 1). Movement A exhibited a significant (Table 2) increase of overall erector spinae muscle activation into the elevated region for both the 10° (p = 0.037) and 18° (p = 0.001) seatback condition compared to the flat surface (low). Movement C exhibited moderate erector spinae muscle activation for the 0° and 10° configurations, while the 18° configuration and the flat surface was considered low. Movement D only showed an increase in erector spinae muscle activation to the high region for the 10° configuration with the 0° condition and flat surface remaining in the moderate range. Movement E exhibited a decrease in erector spinae muscle activation from the flat surface to the 0° and 10° configurations with the muscle activation moving from moderate to low. However, for the 18° and 28° configuration, the erector spinae muscle activation increased to the elevated range. The erector spinae muscle activation for movement F was low for all conditions except for the 18° configuration which was considered high. Movement O exhibited no significantly different changes between the different configurations (Table 2), and all muscle activations were elevated.

Table 1.

Mean percent muscle activation with for each coordinated movement, muscle group, and testing condition. Green represents low muscle activation, yellow is moderate, red is high, and purple is a region of elevated muscle activation above what was achieved on the flat surface. Grey represents groups that were excluded from analysis due to a sample size less than 3.

^†p < 0.05 for coordinated movements A-F compared to the flat surface for the muscle group and coordinated movement. No statistical significance was found for any muscle group when comparing the different conditions for coordinated movement O. The rolling movements depicted are described in depth in Fig. 1.

Table 2.

Statistical analysis results for each coordinated movement and muscle group comparing each device configuration to the flat surface (A-F) or each device configuration to each other (O) using a repeated-measures ANOVA with a Dunnett’s TS post-hoc test. Bold font represents when statistical significance p < 0.05 was found. If no statistical analysis was found for the ANOVA, a post-hoc test was not performed.

Roll Type	Comparison	Erector Spinae	Hamstrings	Abdominal Muscles	Quadriceps
A	ANOVA	p = 0.001	p = 0.177	p = 0.131	p = 0.197
	Flat - 0°	p = 0.994	-	-	-
	Flat - 10°	p = 0.037	-	-	-
	Flat - 18°	p = 0.001	-	-	-
C	ANOVA	p = 0.463	p = 0.009	p = 0.023	p = 0.110
	Flat - 0°	-	p = 0.007	p = 1.000	-
	Flat - 10°	-	p = 0.999	p = 0.022	-
	Flat - 18°	-	p = 1.000	p = 0.039	-
D	ANOVA	p = 0.275	p = 0.993	p < 0.001	p = 0.063
	Flat - 0°	-	-	p = 0.003	-
	Flat - 10°	-	-	p < 0.001	-
E	ANOVA	p = 0.123	p = 0.018	p = 0.006	p = 0.180
	Flat - 0°	-	p = 0.982	p = 0.338	-
	Flat - 10°	-	p = 1.000	p = 1.000	-
	Flat - 18°	-	p = 0.061	p = 0.986	-
	Flat - 28°	-	p = 1.000	p = 0.087	-
F	ANOVA	p = 0.003	p = 0.002	p = 0.007	p = 0.029
	Flat - 0°	p = 1.000	p = 1.000	p = 0.649	p = 0.430
	Flat - 10°	p = 0.881	p < 0.001	p = 0.001	p = 0.991
	Flat - 18°	p = 0.078	p = 0.389	p = 1.000	p = 0.163
	Flat - 28°	p = 0.976	p = 0.009	p < 0.001	p = 1.000
O	ANOVA	p = 0.378	p = 0.124	p = 0.652	p = 0.937
	10° - 18°	-	-	-	-
	10° - 28°	-	-	-	-
	18° - 28°	-	-	-	-

The hamstring muscle activation presented fewer deviations from the flat surface throughout the different coordinated movements (Table 1). For movement A, the hamstring muscle activation on the flat surface and 10° configuration were moderate while the 0° and 18° configuration were high and elevated, respectfully. Most conditions for movement C exhibited low hamstring muscle activation apart from the 0° configuration which was moderate and exhibited a significant difference (Table 2) from the flat surface (p = 0.007). For movement D, all three conditions analyzed had hamstring muscle activation in the moderate region. For movement E, the flat surface and 10° configuration exhibited low hamstring muscle activation, the 0° and 28° configurations were moderate, and the 18° configuration was elevated. Most conditions for movement F presented low hamstring muscle activation levels with the 28° configuration being significantly different (p = 0.009) compared to the flat surface (Table 2). The 18° configuration exhibited moderate hamstring muscle activation that was significantly different compared to the flat surface (p < 0.001). Movement O had elevated hamstring muscle activation for the 10° and 18° configuration and high levels for the 28° configuration, but no significant differences were found (Table 2).

The abdominal muscle activation remained similar to the flat surface or decreased as the seatback angle increased for all coordinated movements (Table 1). For movements A, D, E, and O, the abdominal muscle activation was low for all testing conditions with no significant differences found in the various testing conditions for movements A, E, or O (Table 2). For movement D, the 0° and 10° configurations exhibited abdominal muscle activation significantly different compared to the flat surface (p = 0.003 and p < 0.001, respectively) even though all activation levels were considered low (Table 2). For movement C, the flat surface and 0° configurations had high abdominal muscle activation and significantly decreased to the low range for the 10° (p = 0.022) and 18° (p = 0.039) configurations compared to the flat surface (Table 2). For movement F, the abdominal muscle activations varied throughout the different conditions with the flat surface and 18° configuration being moderate, the 0° seatback being high, and the 10° and 28° configurations being low. The 10° (p = 0.001) and 28° (p < 0.001) configurations had significantly lower activation levels compared to the flat surface (Table 2).

The quadricep muscle activation showed a variety of muscle activation trends throughout the different coordinated movements and device configurations (Table 1). For movement A, the flat surface and 0° configuration exhibited moderate quadricep muscle activation while the 10° and 18° configuration were considered low. For movement C, all conditions analyzed had moderate quadriceps muscle activation apart from the 0° configuration which was high. For movement D, there was an increase in quadricep muscle activation from the flat surface (moderate) to the 0° configuration (high), but a decrease to the 10° seatback (low). Movement E saw a decrease in quadricep muscle activation with the flat surface and 0° configuration exhibiting a moderate activation level and the 10° and 28° configuration being in the low region. However, the 18° configuration exhibited elevated levels of quadricep muscle activation. The quadricep muscle activation varied for movement F with the flat surface, 10° and 28° configurations exhibiting moderate activations while the 0° and 18° configurations were high. Movement O exhibited moderate quadricep muscle activation throughout all device configurations analyzed. No statistical differences were found for the quadricep muscles (Table 2).

4. Discussion

The purpose of this study was to determine how an infant’s muscle activation is affected by different device configurations for each coordinated movement of infant rolling. We hypothesized that each coordinated movement would have unique changes in muscle activation levels dependent on the device configuration.

4.1. EMG analysis

Looking at each movement, the abdominal muscles had the lowest and erector spinae had the highest range of activation in the inclined configurations. Our previous research, focused solely on the seatback angle while rolling without accounting for the different types of rolling movements, found a similar trend with significantly lower abdominal muscle activation and significantly higher erector spinae muscle activation at an incline compared to the flat surface (Siegel et al., 2024a). This indicates that incline angle has a large impact on these muscle groups independent of the coordinated movement utilized.

For the 0° configuration, coordinated movement types that utilized both legs to achieve the roll (C, D, and F) exhibited increased quadricep muscle activation into the high range compared to the flat surface. This increase in quadricep muscle activation reveals that infants are interacting with the device with their legs independent of the incline. A similar trend was observed when looking solely at the differences in seatback incline where we found significantly higher hamstring and quadricep muscle activation overall in the 0° configuration compared to the flat surface (Siegel et al., 2024a). The device used had sidewalls, meant to replicate a design feature that is present on many different commercial infant products. Infants were utilizing the sidewalls to alter their movements and muscle utilization to achieve a roll differently than they would on the flat surface alone when using specific leg focused coordinated movements. To fully understand the biomechanical impact of placing infants in different products, other design features should be considered. Currently, research on infant motor development when applied to the design and evaluation of consumer products is sparse (Ridenour, 1997), and our study suggests that many factors may have an impact on an infant’s motor development, movement, and safety.

Movement F, a movement utilizing trunk and hip flexion, exhibited unique muscle activation patterns for all muscle groups. In the 18° configuration, higher muscle activation levels were achieved compared to the 10° and 28° configurations. This spike in muscle activation indicates that an infant’s initial body position in a product changes their muscle utilization during rolling. Previous research has shown that body position can affect an infant’s muscle utilization and movements while lying on an inclined surface (Wang et al., 2021, 2020) and in various seated infant products (Siddicky et al., 2020b, 2021; Mannen et al., 2023). The changes in muscle utilization and movements while in different mechanical environments may be caused by a difference in the infant’s center of mass altering their movement capabilities. In the device at the 10° configuration, an infant’s head is slightly lower than their feet (Fig. 5). For the 28° configuration, the infant is in a reclined sitting position with their upper and lower body closer together where less range of trunk and hip flexion is required to achieve the same body position needed for movement F on a flat surface, making it easier to achieve this movement in this configuration. However, in the 18° configuration, the infants upper and lower body are positioned at nearly the same angles and infants are required to use higher muscle activation to achieve the trunk and hip flexion needed to roll.

Fig. 5.

Schematics of infant body position in the 10°, 18°, and 28° configurations. The asterisks (*) represent the center of mass estimation of the torso and lower extremities. The line shows the difference between the two based on the device configuration.

Movement O, a roll type only observed in the inclined device, exhibited no significant changes in muscle utilization for the different incline configurations. While other muscle groups that were not analyzed in this study may be impacted, our results indicate that this roll type is not dependent on specific seatback incline angles to be achieved and other design features may impact this movement. The nature of this movement requires infants to press their feet into the device and the angle of the base may have a greater impact than the seatback incline alone. We were unable to compare roll type O to the flat surface condition since no infants used this method on the firm flat surface. In our study, all seatback conditions were tested with a 15° base angle so this may explain why no differences in muscle activation were observed. Future work could consider how differences in the base angle may impact the muscle utilization for this movement type.

While not many significant differences were found, most likely due to the small sample size in our comparison groups, movement C exhibited a decreasing trend in muscle activity for all muscle groups, and a significant decrease for the abdominal muscles, as the seatback angle increased. This indicates that movement C is achieved more easily at higher seatback angles. While not every coordinated movement exhibits a clear muscle activation trend with an increase in seatback incline angle, it is evident that the differing coordinated movements require unique muscle activations which change based on device configuration, supporting our hypothesis. Since, different sets of coordinated movements are utilized more frequently at different ages (McGraw, 1941; Kobayashi et al., 2016), product configurations which allow for coordinated movements to be achieved at a lower level of muscle activation when compared to the flat surface, may result in infants achieving a roll more easily or earlier in life. Additionally, for coordinated movements like A and E where we see an increasing trend of muscle activation, particularly for the erector spinae, rolling may be harder to achieve using these maneuvers in inclined seated environments. This may cause an infant to utilize a different movement that they normally would not have used on a flat surface, like coordinated movement O in our study.

Infants utilize different coordinated movements based on their motor development (McGraw, 1941; Kobayashi et al., 2016) and their environment (Siegel et al., 2024a). The unique muscle activation trends exhibited by each coordinated movement in the different mechanical environments, as shown in this study, relates how rolling ability and ease of movement may be impacted by the environment. The different muscle activation levels exhibited in the different seatback configurations, similar to commercially available infant seated products, suggests that the maneuvers an infant utilizes on a flat surface may be achieved at different developmental stages in certain product configurations. If an infant achieves a roll in a product before doing so on the flat surface, they are often at an increased risk for injury or death for two main reasons: (1) since they are likely unfamiliar with the prone position within the unique mechanical environment, self-correction may be more challenging and certainly unpracticed, and (2) the concavity, soft-goods, and lack of a firm support present in most inclined seated products introduces unique suffocation-relation hazards due to mouth-nose interactions with soft-goods that infants do not experience while prone on a firm flat environment (Mannen et al., 2019). While fewer infants achieved a roll as the seatback incline increased, the way they achieved these rolls may be more concerning. Additionally, many warning and hazard labels associated with seated products relate to motor milestone achievement for safety, and this research shows that is not appropriate since rolling milestone achievement on a flat surface does not directly translate to inclined seated products. Proper use of the harness system available in inclined seated products should be used to theoretically decrease the chances of an infant achieving an unexpected roll, but future work is still needed to expand on how harness designs and use influence movement and on other product design features that may impact safety while rolling or prone in inclined seated products.

4.2. Considerations and limitations

The data collection occurred throughout the COVID-19 pandemic (October 2020-June 2021), meaning some infants were more uncomfortable outside of their homes or with strangers, altering their movements in the laboratory setting. Inherent limitations also exist with surface EMG sensors. However, the sensors chosen were successfully used in other infant studies (Wang et al., 2021; Siddicky et al., 2020b; Wang et al., 2020; Siddicky et al., 2021; Solopova et al., 2019; Price et al., 2018) and our experimental design allowed infants to serve as their own controls, limiting the variability sometimes associated with EMG placement. Additionally, the muscle groups explored were not comprehensive but selected based on the previously defined coordinated movements during rolling on a flat surface (Kobayashi et al., 2016) and their corresponding muscle activation patterns (Siegel et al., 2023). Our results showed that many design features may influence the way infants achieve a roll and future research could be expanded to features beyond an incline.

4.3. Conclusion

This study is the first to explore the muscle activation levels of the different coordinated movements of infant rolling in a range of inclined mechanical environments. All previously defined coordinated movements exhibit different trends of muscle utilization with an increase in seatback incline angle. Movement O, a coordinated movement utilizing a push-off mechanism exhibited only in the device, exhibited no significant change in muscle activation regardless of the device configuration, demonstrating that it can be achieved at a variety of inclines and it may rely on design features other than the seatback incline angle alone. Movements that utilize both legs to achieve a roll (C, D, and F), exhibited the largest change from the flat surface to the 0° configuration suggesting product interaction independent of the incline, though statistical significance was not achieved for movement F. Movement C utilizing trunk and hip flexion exhibited a decrease in muscle activation at increased seatback inclines, meaning this movement may be achieved more easily at higher seatback angles. For movement F, the higher muscle activation seen at the 18° seatback indicates that an infant’s initial body position caused by the device impacts their rolling ability. The changes in muscle activation patterns between the different coordinated movements show that an infant’s ability to achieve a roll in different mechanical environments may be completed differently at certain stages of motor and musculoskeletal development when compared to a firm flat surface.

Data availability

Data will be made available on request.

Data statement

The data that support the findings of this study are available from the corresponding author, EMM, upon reasonable request and in accordance with Institutional Review Board guidelines.