An electronic, narrative literature surveillance adopting the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) benchmarks was followed (Moher et al. 2009). The definitions were guided by the PRIMSA checklist for participants, interventions, comparisons, outcomes and study designs (PICOS) (Miller 2001). The participants were the research articles pertaining to the change in cervical posture amongst 10–14-year-old school children that carried backpack; the intervention was not necessarily a therapeutic intervention but is interpreted as an exposure, namely, the change in cervical posture of 10–14-year-old school children who carry backpacks. The outcomes of interest included (1) a cascade of kinematic events resulted in cervical postural deviations when school children carried backpacks, (2) specific percent mass of backpack loads that initiate change in craniovertebral angles and (3) gender-specific variations with regard to differing backpack mass loads relative to girls’ and boys’ body mass that can be carried without producing deviations in cervical posture.

Patient/Problem: Cervical postural changes amongst children aged 10–14 years old, who carry school backpacks.

Intervention: Change in CVA when carrying school backpacks.

Comparison: The change in children’s CVA when carrying backpacks as compared to when not carrying backpacks.

Outcome: Altered cervical posture manifested through reduced CVA, which is associated with cervical postural syndrome and thoracic kyphosis.

Research question: Does cervical posture change when children aged 10–14 years old carry school backpacks as compared to when they do not carry backpacks.

The study design of this review involved pre-test and post-test assessments

An electronic exploration of peer-reviewed literature using the Google Scholar, Science Direct, PubMed, EMBASE, AMED, CINAHL, OVID and Sabinet search engines was completed for papers published during the period 2009–2019 (Figure 1).

The primary keyword in the literature search included ‘schoolbag carriage’; then subsequent words such as ‘backpack’, ‘cervical posture’ and ‘children’ were added. The review and selection criterion for the documents was accomplished in three phases: title review, followed by abstract review and full text review. Literature search was conducted from December 2018 until August 2019, and the records were screened by the authors (M.S., T.J.E. and H.V.H.). Each of the authors completed the three phases, resulting in a list of studies to be synthesised into the commentary. Divergent views amongst the authors whether to include or exclude a study was resolved by holding a joint review and it was put to vote whether the study should be included, based on the application of the inclusion and exclusion criteria (majority vote dictated decision).

Participants were records pertaining to the impact of schoolbag backpack carriage on the student’s cervical posture. Participants of the studies had to be within the age strata of 10–14 years and included both genders. The types of studies that were included in this evaluation were empirical articles and randomised control studies. Relevant themes that emerged included altered CHA, CVA and SSA because of schoolbag carriage and rehabilitative exercises were formulated to resolve the altered cervical posture of children carrying heavy schoolbags.

Records and articles preceding to the period prior to 2009, relating to the altered cervical posture of adults carrying backpacks, and those of children older than 15 years were not included in the study. Similarly, backpack studies relating to non-cervical posture, electromyography studies (EMG), non-English papers, meta-analyses, systematic reviews and case reports were excluded, as the authors’ primary aim was to synthesise empirical articles pertaining to the aforementioned topic.

The value of each record was assessed by adopting a modified Downs and Black Appraisal Scale, which examines the merit of randomised controlled trials and non-randomised papers (Downs & Black 1998) (Table 1). A modified Downs and Black Appraisal Scale was applied as not all of the questions on the original checklist were related to this study, as underlined by Gorber et al. (2007). These practices were employed in order to avoid any researcher bias. The modified checklist comprises 13 questions, with a maximum of 13 points. A score of either 0 (no) or 1 (yes) was given for each answer. The questions adopted from the modified Downs and Black Appraisal Scale are 1, 2, 3, 4, 6, 10, 11, 12, 13, 14, 18, 20 and 27 (Table 1). These questions are classified into four sections, which evaluate the whole merit of each record (Table 1). The classification considered the reporting prowess (n = 6 questions), external validity (n = 3 questions), internal validity (n = 3 questions) and power of significance (n = 1 question) of each publication (Downs & Blacks 1998). The reporting sub-section reviews the studies’ aims, sample characteristics and the outcome measures. The external validity reviews the representativeness of findings, and whether they can be generalised from the population the subjects were recruited from (Downs & Black 1998). The internal validity reviews whether subjects were blind to interventions used, and whether the statistical analyses were appropriate. The power of significance reviews whether the statistical tests used were adequate to determine clinically important findings (Downs & Black 1998). In the event of any disagreements amongst the authors (M.S., T.J.E. and H.V.H.) regarding the score of the selected records or articles, the authors were able to query the scoring of each record and would then discuss the scores adopting the jointly accepted score. The cumulative score of each record was subsequently converted into a percentage, thereby appraising the overall merit of the individual records (Downs & Black 1998). The overall merits of the records were further classified into the following scale: less than 50% (weak), 50% – 69% (fair), 70% – 79% (good) and less than 80% (very good) (Li, Khoo & Adnan 2017). The mean rating of the selected papers was 76.3% (good).

The following data were extracted from 14 articles:

The authors completed a search in the Sabinet database under the categorisation Medicine and Health. The preliminary search word used was ‘schoolbag’ that yielded 40 records. Then the subsequent word ‘schoolbag carriage’ was entered that yielded nine records. These records were then reviewed for relevance with respect to title and year of publication and it yielded two records. A similar search strategy was completed with regard to the other search engines.

Descriptive statistical analyses including mean and percentages were performed. An inferential statistical paired T-test that compared the change in CVA during the unloaded versus loaded phases of the 14 studies was also completed (with the probability factor set at 0.05). The descriptive analyses involved calculating the sum of the participants in the 14 studies, then calculating their mean age, body mass, height and backpack mass. The backpack mass was then expressed as a percentage relative to the average body mass of the participants. The mean CVA of the participants when loaded (carrying traditional type backpacks) and unloaded (not carrying backpacks) was calculated and compared to determine changes in sagittal plane posture. The studies that employed an exercise intervention to combat the effect of heavy backpack post-test CVA results were omitted from the above calculation of the mean CVA and its subsequent comparison.

In order to completely comprehend the cascade of sagittal plane kinematics (deviated cervical posture) when carrying school backpacks, a number of terms will need to be defined. The CHA is created by drawing a horizontal line bisecting the tragus of ear and another line drawn from the tragus to the external canthus of the eye (Hande et al. 2012). Hande et al. (2012) further recommended that this is an approximation of the head on neck angle, which is established in relation to the upper cervical spine. The CVA is created at the juncture between the horizontal line drawn through the spinous process of cervical vertebra seven (C7) and a subsequent line drawn to the tragus of the ear. It is an approximation of neck on cervical vertebrae and head alignment in relation to the thoracic vertebrae. A small CVA is suggestive of a forward head posture (Hande et al. 2012). The SSA is the angle created at the juncture between the horizontal line drawn through C7 and a corresponding line drawn between the mid-point of the greater tuberosity of the humerus and posterior aspect of the acromion (Hande et al. 2012). Hande et al. (2012) reported that this angle identifies a forward shoulder position with a smaller angle, suggesting that the shoulder is positioned further anterior than C7 (rounded shoulder).

This article followed all ethical standards for research without direct contact with human or animal subjects.

The 14 studies comprised 13 experimental, observational and cross-sectional studies, without concurrent controls (Abrahams et al. 2011; Goswami, Sarkar & Mishra 2017; Hande et al. 2012; Hundekari et al. 2013; Khallaf et al. 2016; Kistner et al. 2013; Leman, Idris & Murdana 2013; Malik, Vinay & Pandey 2017; Mo et al. 2013; Mosaad & Abdel-aziem 2018; Pahwa 2013; Ramprasad, Alias & Raghuveer 2009; Vaghela et al. 2019) and one experimental study with a concurrent control (Misra, Nigm & Alagesan 2012). Whilst all studies (n = 14) reviewed the effects of backpack loading on cervical posture, four included interventions (Leman et al. 2013; Misra et al. 2012; Mo et al. 2013; Mosaad & Abdel-aziem 2018) (Table 2). One study reviewed the effects of exercise rehabilitation in order to resolve deviated cervical posture caused by backpack carriage (Misra et al. 2012), whilst another study compared traditional backpack loading with modified bag carriage (Leman et al. 2013). Mosaad and Abdel-aziem (2018) compared the impact of carrying a traditional style bag with that of a double-sided bag on children’s posture and proprioception. Mo et al. (2013) and Goswami et al. (2017) reported on the association between altered cervical posture, gait analyses and backpack loading. Of the 14 studies considered, eight studies were conducted in India (57.1%), two studies in Egypt (14.2%) and one in each of the following: South Africa (7.1%), United States of America (7.1%), Canada (7.1%) and Indonesia (7.1%).

The measures of the risk of bias assessment are described in Table 3. The Reporting sub-section mean was 5.07, whilst the External Validity sub-section mean was 1.85. The mean of the sub-sections of Internal Validity and Power of Significance was 2.0 and 1.0, respectively. The overall mean merit rating of the 14 records as a percentage was 76.3%, which was classified as good according to the Li et al. (2017) Scale.

A total of 1061 participants were recorded across the 14 studies, with an average age of 11.5 years (± 1.3), which yielded an average of 13.75% backpack mass relative to the child’s body mass. The average CVA was 53.9° ± 14.6°, whilst standing without carrying a backpack (unloaded phase) was reduced to 50.4° ± 16.4° when loaded (p < 0.05) (Table 4). Backpack loads carried by subjects varied from 5% to 30% of the participant’s average body mass and produced a mean CVA decline of 3.5°. The participants’ average body mass was 37.8 kilograms (kg) ± 6.6 kg, height was 1.41 metres (m) ± 0.05 m and backpack mass was 5.2 kg ± 0.9 kg.

The following themes evolved from the literature review:

Children’s sagittal plane posture was altered when they carried backpacks weighing between 5% – 20% of their relative body mass (Hande et al. 2012; Khallaf et al. 2016; Pahwa 2013; Vaghela et al. 2019). The deviation in posture was indicated by the change in CVA, CHA, and SSA. The CVA and SSA decreased progressively as backpack loads increased, whilst the CHA increased progressively (Hundekari et al. 2013). The CVA decreased in order to maintain balance within the anterior-posterior vertebral curves. The normal anterior-posture vertebral curves include marginal cervical lordosis, thoracic kyphosis and lumbar lordosis, which are responsible for aiding the vertebral column in supporting an upright posture (Mansfield & Neumann 2014). When the child carries a backpack load which is beyond the muscular strength of the erector spinae, a forward lean away from the medial-lateral axis is adopted (Kistner et al. 2013), resulting in the forward movement of their centre of gravity (anteroposterior index) (Mosaad & Abdel-aziem 2018), thereby increasing the risk of falling forward. In an attempt to avoid falling and simultaneously securing the backpack, the child compensates by hyper-extending their lumbar vertebrae (excessive lordosis), then hyper-flexing their thoracic vertebrae (excessive kyphosis) and anteriorly protruding their cervical vertebrae (diminished CVA, resulting in cervical postural syndrome). When the child’s CVA decreases, the child’s CHA increases so as to maintain the head in an upright position, producing altered kinetic chain affects which ripple down the lower vertebrae. This kinematic vertebral change decreases CVA and SSA, but simultaneously increases CHA (Hande et al. 2012; Hundekari et al. 2013). The spontaneous re-alignment of vertebrae in order to maintain balance and an upright standing posture is known as serial distortion of the kinetic chain (Prentice 2011). Habitual carrying of heavy backpacks produces a kypholordotic posture with cervical postural syndrome (decreased CVA and SSA, coupled with increased CHA). Furthermore, the posterior kyphosis produces an anterior sunken chest (pes cavus) that may impact the child’s ventilation. Clinical literature has confirmed that carrying hefty school backpacks reduces the subject’s lung volume (Alaa & Baiee 2016; Ramadan & Ali-Shayea 2013; Veirria & Ribiero 2014). The sunken chest produces a decrease in the intra-rib spacing, bringing the superior ribs closer to the inferior ribs (Hammill et al. 2017). This action asymmetrically strengthens the internal intercostal muscles (responsible for expiration), whilst simultaneously elongating the external intercostal muscles (responsible for inspiration) (Mansfield & Neumann 2009). The asymmetrical alteration of the resting length tension relationship of these force-couple muscles produces a negative impact on the child’s inspiration, producing chronic restrictive pulmonary disorder, thereby diminishing their forced vital capacity and inspiratory lung volumes (Mansfield & Neumann 2009; McArdle, Katch & Katch 2015).

However, Misra et al. (2012) reported that specific muscle strengthening and endurance conditioning help to resolve habitual cervical postural syndrome amongst children carrying heavy backpacks. Similarly, Prentice (2011) advocated that therapeutic resistance strengthening of the thoracic erector spinae muscles can reduce the presence of kyphosis, whilst simultaneously stretching the anterior chest muscles (pectoralis major and minor, serratus anterior and the internal intercostals). The impact of the improved muscle strength, endurance and posture garnered from the exercise therapy in association with its influence on the child’s pulmonary functioning has however not being measured. It is recommended that this gap in the literature should be measured with specific investigations.

Pahwa (2013), Khallaf et al. (2016) and Goswami et al. (2017) noted significantly decreased CVA once backpack loads exceeded 9%, 10% and 12% of the boys’ relative body mass, respectively. The mean age of the cohort in these studies was 10.7–14.1 years, referring to a cervical postural, change-specific age strata of 10–14 years. These findings suggest that a critical safe backpack load might be set at 8% relative to boys’ body mass for boys aged 10–14 years (if one adopts the lower percent load relative to the child’s body mass). The critical limitations of these studies were the small sample size: Pahwa (2013) (n = 10 boys), Khallaf et al. (2016) (n = 50 boys) and Goswami et al. (2017) (n = 6 boys). A larger sample is needed in order to validate their findings. The empirical evidence indicates that CVA changes occur when a child carries a backpack, but they differ in opinions as to what percent mass of the backpack load produces a significant CVA change. Chansirinukor et al. (2001) reported that backpack loads from 15% produce significant CVA changes, whilst Pahwa (2013) reported that loads exceeding 8% produce altered CVA. Pahwa (2013) concurs with Ramprasad et al.’s (2009) findings. These findings are conflicting in their precise percentage load value. Therefore, the authors recommend that further empirical investigations should be conducted to determine the precise percentage of backpack load that produces a significantly altered CVA, CHA and SSA, which are also associated with neuro-musculoskeletal discomfort and pain.

Gender variations were also identified. Khallaf et al. (2016) reported that girls’ CVA started to change with loads of 5% relative to their body mass. They suggest that girls’ cervical posture begins to change earlier in order to accommodate carrying heavy backpacks well before the prescribed guidelines adopted by the American Occupational Therapy Association (15%) and the American Academy of Paediatrics (10%). Khallaf et al.’s (2016) findings concur with Hammill et al.’s (2017) recommendation that boys and girls should have different safe mass loads per age strata in so far as their muscle strength and endurance differ. Boys and girls during their anatomical and physiological development possess different muscle strength and endurance capacities, which will impact their ability to carry different relative, percent body mass backpack loads, as well as being able to maintain anatomically correct and neuro-musculoskeletal discomfort-free, and/or pain-free posture. Males are generally stronger than females within their age-specific strata (Hammill et al. 2017; Khallaf et al. 2016) that may allow them to carry relatively greater/heavier backpack loads in relation to their body mass and adopting pain-free postures. This aforementioned evidence warrants the international paediatric health associations to prescribe independent gender-specific safe backpack mass loads for boys and girls.

It is a common practice to adopt Mill’s Canons (Dishman, Heath & Lee 2013) to determine the vigour of the evidence supporting casual inferences. As such, the authors embraced Mill’s Canons in order to establish the strength of evidence supporting the causal inference that carrying heavy backpacks produces cervical posture deviation amongst children: