How are sex-gender differences in chair-and-desk-based postural variability explained? A scoping review

Abstract

Background

Desk-work-related musculoskeletal pain is more prevalent among female workers than male workers. This may be contributed to by sex and/or gender differences in postural variability however, the mechanisms underpinning these differences are poorly understood. This review investigates whether desk-based postural variability studies investigate sex-gender differences and, how they explain the mechanisms behind these differences.

Methods

A scoping review was conducted with four databases (PubMed, Embase, Scopus and ProQuest) searched in June and July 2023. Studies investigating postural variability among desk-based workers were included and a narrative approach used to synthesise results.

Results

15 studies were included. Only four reported on sex-gender differences. None collected psychological or social information to explore reasons for sex-gender differences.

Conclusion

The mechanisms behind postural variability differences between sexes and genders are complex and multifactorial. Studies largely do not consider sex and gender and do not collect the information necessary to explain their results.

PRACTITIONER SUMMARY

This scoping review investigates desk-based postural variability, exploring sex-gender differences and underlying mechanisms. Among 15 studies, four address sex-gender disparities, while none consider psychological or social factors. Failing to recognize these differences leads to ineffective, generalized interventions .Tailored approaches, encompassing biomechanical, psychological, and social aspects, are crucial for effective interventions.

Keywords:

• Postural variability
• sex
• gender
• desk-based work
• work-related pain

Introduction

Musculoskeletal (MSk) pain is common among desk-based workers worldwide, particularly at the head, spine and upper limbs (Janwantanakul et al. 2008; Putsa et al. 2022; James et al. 2018), and is more prevalent among female workers (Paksaichol et al. 2012; Hush et al. 2009; Oha et al. 2014; Soria-Oliver et al. 2021; Holzgreve et al. 2021; Collins and O’Sullivan 2015; Karlqvist et al. 2002). A number of interventions have been recommended to manage or prevent work-related MSk pain development including physical exercise (e.g. strengthening, stretching), use of sit-stand desks, and increased work-break frequency, however, their effectiveness is inconsistent and effect sizes are often small (Luger et al. 2019; Hoe et al. 2018; Tersa-Miralles et al. 2022; Putsa et al. 2022; Reliquias and Kuebler 2019). This may be due, at least in part, to the multifactorial nature of work-related MSk pain with no single contributor sufficient to independently cause pain, and no single intervention sufficient to prevent or manage it (Bontrup et al. 2019; Petit et al. 2018). A wide range of biological, psychological and social contributing factors have been associated with MSk pain among desk-based workers, many of which interact with postural variability, pain, the physical environment, and with each other, frequently in bidirectional relationships (Soria-Oliver et al. 2021). Understanding sex-gender differences in desk-work-related MSk pain requires understanding of sex-gender differences in these underlying contributing factors. Although all contributors to work-related pain are worthy of exploration, this paper focuses on potential mechanisms underlying sex-gender differences in postural variability.

Postural and movement variability refers to the diversity of biomechanical strategies a person uses to complete a given task (Srinivasan and Mathiassen 2012; Hamill, Palmer, and Van Emmerik 2012; van Emmerik et al. 2016). While the ability to coordinate a range of different patterns to achieve the same task (high variability) allows for flexibility and adaptation to task demands and environmental constraints, lower variability is theorised to cause sustained or repeated load distribution to a limited group of active and passive structures increasing the risk of fatigue, overload, tissue damage and pain development (Jun et al. 2017; Putsa et al. 2022; Mingels et al. 2021; Srinivasan and Mathiassen 2012; Hamill, Palmer, and Van Emmerik 2012). In addition to an overload-related nociceptive pain pathway, lower variability may contribute to nociplastic pain pathways through persistent nociceptive signalling from chronically overloaded structures (Nijs et al. 2021). MSk pain among desk-based workers has been associated with low work-task variation (Jun et al. 2017), prolonged sitting time and infrequent breaks (Putsa et al. 2022; Ranasinghe et al. 2011; Daneshmandi et al. 2017), with authors often attributing these relationships to the effects of low biomechanical variability. In contrast, desk-based workers who make more frequent postural changes (greater postural variability) are less likely to report low back pain (Bontrup et al. 2019; Akkarakittichoke and Janwantanakul 2017; Zemp et al. 2016).

Female participant groups have been reported to be less variable in their postural (Mason-Mackay 2023), movement (Pollard et al. 2005) and muscle activation patterns, (Srinivasan and Mathiassen 2012), potentially contributing to the reported sex-gender differences in MSk pain prevalence. Differences in variability between sexes-genders are likely due to differences in the underlying factors which create limitations on the movement or position of any part of the body. Sex-gender differences in biological factors such as strength, flexibility and endurance are plausible contributors (e.g. a worker with lower endurance of the thoracic extensors may lose the option of using an upright sitting posture after a short period of time) however, differences in psychological and social experience may also play a role.

Postural variability and psychological experience

Psychological factors such as stress (Putsa et al. 2022; Hush et al. 2009; Alhakami et al. 2022), mental workload (composite measure including stress, fatigue and job-demand measures) (Soria-Oliver et al. 2021) and discomfort/dissatisfaction within the workplace environment (e.g. poor lighting, temperature, air quality, acoustic conditions) (Robertson, Huang, and Larson 2016; Jun et al. 2017) have been associated with MSk pain among desk-based workers. While these likely contribute to nociplastic pain through changes to central and peripheral pain processing (Bułdyś et al. 2023; Fitzcharles et al. 2021), they may also contribute to nociceptive pain and reduced variability through changes to posture and muscle tension. Several studies have shown a direct relationship between posture and psychological distress, with depression, anxiety, and lower levels of self-efficacy and self-esteem associated with flexed spinal postures (Mingels et al. 2021; Dehcheshmeh, Majelan, and Maleki 2023; O’Sullivan et al. 2011) and elevated (Dehcheshmeh, Majelan, and Maleki 2023; Mork et al. 2018) and protracted (Meier, Vrana, and Schweinhardt 2019) scapulae. Anxiety and depression have also been associated with reduced postural control (Meier, Vrana, and Schweinhardt 2019) and a recent study found that stress and anxiety were associated with lower postural variability (Mingels et al. 2021). These results suggest that physical responses to prolonged psychological distress have the potential to contribute to reduced postural variability over time.

Postural changes may also be caused or exacerbated by interactions between stress and muscle tension. Increases in muscle activity have been associated with psychological stress during computer work (Eijckelhof et al. 2013; Wixted, O’Riordan, and O’Sullivan 2018), potentially as a result of changes to work style (e.g. increased work pace, higher forces on the keyboard and mouse), increased arousal (Eijckelhof et al. 2013), or stress-induced reduced inhibition of upper trapezius muscle activation by the central nervous system (Marker, Campeau, and Maluf 2016). Authors have suggested that sustained muscle-activation (low-variability in muscle activation patterns) may lead to tissue overload and pain through insufficient rest (Larsman, Kadefors, and Sandsjö 2013). Furthermore, sustained muscle contraction may contribute to low postural variability by maintaining a joint in a particular position (e.g. scapular elevation). The relationship between stress, muscle tension and pain is supported by a recent systematic review which reported that self-perceived muscle tension among office workers is associated with higher stress levels and predicts the development of neck pain (Jun et al. 2017).

These relationships between psychological distress, muscle tension and posture may also contribute to sex-gender differences in MSk pain as intersecting social and organisational factors within some workplaces create conditions in which women are more likely to experience higher levels of work-related stress. Women are more likely than men to experience lower and unequal pay (LeanIn.Org and McKinsey & Company 2022), fewer work opportunities (Starmarski and Hing 2015), being overworked (LeanIn.Org and McKinsey & Company 2022), and sexism (Starmarski and Hing 2015; LeanIn.Org and McKinsey & Company 2022). These experiences have all been associated with negative psychological outcomes including higher stress levels, anxiety and depression (King et al. 2018; Starmarski and Hing 2015; Zakerian and Subramaniam 2009; Ranasinghe et al. 2011; Triana et al. 2019) potentially contributing to both lower postural variability and greater work-related pain prevalence among women.

Postural variability and workplace social environment

Other aspects of the workplace social and organisational environment may also affect postural variability. There is evidence that workers are less likely to take breaks from their desk if they believe that colleagues or managers will criticise or penalise them for doing so (Oliver et al. 2021; Ojo, Bailey, Brierley, et al. 2019; Ojo, Bailey, Hewson, et al. 2019) and less likely to make use of standing desks if others perceive it as ‘abnormal’ or as indicating a lack of productivity (Hall et al. 2019). These factors may affect sexes-genders differently as a result of sex-gender-related power dynamics (Mason-Mackay 2023). Because most gender systems grant greater power to men (Cislaghi and Heise 2020) and, because men often hold higher positions in workplace hierarchies (Huang et al. 2019; Shannon et al. 2019; Messing et al. 2003; LeanIn.Org and McKinsey & Company 2022) and have greater workplace autonomy (Messing et al. 2003) men may feel freer to move, stand or take breaks from their desks without concern for social or professional consequences (Messing et al. 2003; Mason-Mackay 2023). As this effectively results in men being socially permitted to move more at work (allowing greater biomechanical variability) than women, these social and organisational differences may play an important role in sex-gender differences in work-related MSk pain. This is consistent with research linking a lack of autonomy (Wahlström et al. 2004; Ranasinghe et al. 2011; Sihawong et al. 2016) and poor social support from colleagues and managers (Sihawong et al. 2016; Ranasinghe et al. 2011; Robertson, Huang, and Larson 2016) with MSk pain among desk-based workers. Furthermore, Larsman, Kadefors and Sandsjö (2013) reported that while having a greater level of influence at work did not lower stress levels, it did lower perceived muscle tension. They theorised that while stress levels remained high, greater influence at work might increase worker ability to plan their work in a way which allows for mechanical variation and small breaks, allowing muscles to rest more often and reducing perceived tension.

Postural and movement factors also influence perceptions of ‘femininity’ and ‘masculinity’ with stillness (D’Argenio et al. 2020) and closed, contracted postures associated with and often expected of women (Ellemers 2018; Mason-Mackay 2023). Studies have reported that women who adopt dynamic and expansive postures are viewed negatively due to failure to conform with gender-based expectations (D’Argenio, Finisguerra, and Urgesi 2020; Bailey, LaFrance, and Dovidio 2017). Although gender norms create postural constraints for both men and women it has been suggested that the combination of gendered power dynamics and gendered posture and movement expectations result in fewer postural options and lower biomechanical variability for women than for men, at least in some contexts (Mason-Mackay 2023).

Multifactorial mechanisms

In addition to directly affecting postural variability, biological, psychological and social factors can together influence outcomes such as pain and sleep which themselves become factors with potential to influence variability (Mingels et al. 2021; Meier, Vrana, and Schweinhardt 2019; van Dieën et al. 2019). These factors (like biomechanical variability itself) could be considered composite biopsychosocial factors in the sense that they cannot easily be placed in biological, psychological or social categories but are created by complex interactions between all three. Understanding these factors and the relationships between them is necessary in order to design effective interventions which target the particular combination of pain drivers for specific workers. Studies therefore need to analyse sex and gender groups separately and consider a range of biological, psychological, social and composite factors in order to understand the mechanisms operating behind any differences found.

Therefore, this review aims to identify how current research explains the mechanisms behind differences in postural variability between sexes-genders.

Sex and gender

The terms ‘sex’ and ‘gender’ refer to different, inter-related concepts but are frequently used interchangeably in biomechanical research. ‘Sex’ refers to biological characteristics including chromosomes, hormones and anatomy while ‘gender’ is a social construct which refers to social relations and expectations, self-expression and identity (Mason-Mackay 2023). Physical factors which could contribute to variability differences (e.g. strength, endurance) may be related to sex differences, while social factors (e.g. social expectations) may be related to gender. Throughout this review we use the terms sex and gender as described here and the term sex-gender when both are being referred to. However, it is important to note that the studies we reference often do not differentiate between these terms. When referring to these papers we use the terminology of the original authors but recognise these terms are being used imprecisely.

Methods

A scoping review was conducted according to the five-step approach recommended by Levac, Colquhoun, and O’Brien (2010): (1) identifying the research question, (2) identifying relevant studies, (3) selecting the studies, (4) charting the data and (5) collating, summarising and reporting the results. The results are reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA) guidelines.

Identifying the research question

To answer the question ‘how does the current literature explain the mechanisms behind differences in seated workplace postural variability between sexes-genders’ the review objectives were to:

• Identify whether studies investigating seated postural variability analyse sex-gender groups separately

• Identify the biological, psychological, social and composite factors these studies use to understand postural variability differences between sexes-genders

• Identify the underpinning theories used to explain sex-gender differences in postural variability

Identifying relevant studies

In line with the purpose of scoping reviews, our approach was broad, with an emphasis on studies that investigated postural variability in the office setting.

Search strategy 

A comprehensive literature search for potentially eligible publications was conducted by electronic searching. Four databases (PubMed, Embase, Scopus and ProQuest) were searched for relevant studies and researchers independently reviewed studies for eligibility and inclusion. A combination of the following search terms was used in all databases with the appropriate Boolean operators and truncations: (((sitting) OR (seat*)) OR (‘working posture’)) OR (‘computer work’) AND ((((‘postural variability’) OR (ergonomic*)) OR (‘postural behaviour’)) OR (‘spine mechanics’)) OR (‘trunk movement’)))) NOT standing. Additionally, reference list screening and citation tracking were performed.

Eligibility criteria 

All original research studies objectively investigating seated postural variability, studies including healthy men and women, above 18 years of age were included. Intervention studies were included if they captured baseline data which could be used to determine pre-intervention variability differences between sexes-genders. Studies were excluded if they reported on one gender only. Further, to retrieve the latest evidence a date restriction was applied; studies were sought from 2005 to 2023 and only English articles were included. Full-text articles were retrieved for studies that appeared to meet the inclusion criteria, and for those in which insufficient information was presented in the title, abstract and key words to determine eligibility.

Selecting the studies

Two reviewers independently screened the titles and abstracts of identified studies. A third reviewer checked the results for accuracy. Results of the initial screening were compared, and full-text records were obtained for articles deemed eligible by at least one reviewer. Two reviewers independently screened the full texts using the eligibility criteria. Any discrepancies were resolved by discussion with a third reviewer.

Data charting

Four researchers extracted, collated and summarised relevant data into a purpose-built Microsoft Excel database from the included studies. The following study components were extracted: general study information, including author and year of publication; study design and settings, population characteristics, postural variability measure and all other measures collected and the theoretical underpinnings used to explain study results. In case of missing data, the authors of the studies were contacted for additional data via the information provided in their publications.

Data synthesis

A narrative synthesis of the findings was conducted, highlighting the scope of studies collecting office-based postural variability data and analysing gender/sex separately. Further to this we summarised the nature of measures collected to describe postural variability differences reported between genders.

Results

A total of 2116 records were identified and 349 abstracts were screened. After reviewing 150 full-text papers, a total of 15 full texts were included in this review. A PRISMA flow diagram summarised the study selection process (Figure 1).

Figure 1. PRISMA diagram.

Study characteristics

A summary of the characteristics of the studies included in the scoping review can be seen in Table 1. The number of participants varied from 12 to 46. Five of the studies were conducted in Canada and by the same research group. Three studies used a pressure mapping system to measure postural variability, another three used accelerometers and the remaining nine studies used various motion analysis systems. Five studies included only basic demographic biological information (e.g. height, age) and postural variability. Three studies included additional biological factors, one included psychological factors, and ten included composite biopsychosocial factors. None of the included studies’ primary aim was to explore the reasons behind sex-gender differences in postural variability.

Table 1. Study characteristics.

	Author	Country	Study main aim/Question	Population	Methodology	Postural variation measurement tool	Other factors recorded
1.	Akkarakittichoke and Janwantanakul (2017)	Thailand	To investigate seat pressure distribution characteristics, i.e., average pressure, peak pressure ratio, frequency of postural shift, and body perceived discomfort (BPD), during 1 hour of sitting among office workers with and without chronic LBP.	Participants: Office workers Age: Control group: 29.6 (±5.1) males and females combined Sample Size: F: n = 36 M: n = 10 Total: n = 46	Task: Standardised task of typing text passage at a computer work station for 1 hour. Workstation set-up: Standard office chair with a pressure map	Seat pressure mat device	Composite factor (pain): Borg scale - Ratings of perceived discomfort (RPD)
2.	Arippa et al. (2022)	USA	To investigate the postural strategies in a cohort of sedentary office workers during a prolonged sitting bout.	Participants: University staff and students Age: 33.2 (±9.4) males and females combined Sample Size: F: n = 23 M: n = 5 Total: n = 28	Task: Normal work in a seated position for up to 3 hours. Workstation set-up: Own standardised adjustable desk and chairs.	Seat pressure mat device	Composite factor (regular activities): Baseline survey- daily habits, physical activity, commuting preferences, gym subscriptions etc. Biological factors: Baseline survey - chronic health conditions and demographics.
3.	Barbieri et al. (2019)	Brazil	To what extent does variability in neck, trunk and right upper arm postures change when computer work is performed both sitting and standing at a sit-stand station, compared to only sitting?	Participants: Office workers Age: 41.3 (±8.8) males and females combined Sample Size: F: n = 16 M: n = 8 Total: n = 24	Task: Typical office work for 2 hours. Workstation set-up: adjustable sit-stand table	Digital tri-axel accelerometers and sensors	none
4.	De Carvalho and Callaghan (2022a)*	Canada	To investigate the effect of a high velocity low amplitude (HVLA) lumbar spinal manipulation on trunk muscle activation, spine posture and movements during prolonged office sitting.	Participants: University population Age: F: 22.0 (±2.7) M: 24.5 (±6.2) Sample Size: F: n = 10 M: n = 10 Total: n = 20	Task: Standardised 2 hour typing task on at a computer workstation. Workstation set-up: seated computer workstation	Accelerometers	Biological factor (muscle activity): Indwelling electromyographic (EMG) data of bilateral multifidus muscles at L4/L5. Surface EMG of bilateral thoracic erector spinae, lumbar erector spinae, lumbar multifidus and glutaeus medius. Composite factor (pain): Visual Analogue Scale (VAS) - Perceived ratings of transient pain
5.	De Carvalho and Callaghan (2022b)*	Canada	To compare these chair design features, with the same configurable chair, in a prolonged trial of computer work.	Participants: University population Age: F:24.1 (±3.2) M: 24.1 (±4.2) Sample Size: F: n = 16 M: n = 15 Total: n = 31	Task: Computer-based controlled data entry task at a computer workstation for four, 30 minute blocks, sitting in a block randomised order in one of four seat conditions. Workstation set-up: An ergonomic computer desk and keyboard tray was adjusted such that each participant initially sat with 90 degrees of knee flexion with their feet flat on the floor, 90 degrees at the hips and 90 degrees flexion at the elbows with neutral wrist posture and relaxed shoulders. A footrest was used when necessary.	Accelerometers	Composite factor (pain): Visual Analogue Scale (VAS) - Perceived ratings of transient pain
6.	Gregory, Dunk, and Callaghan (2006)*	Canada	To evaluate differences between sitting on a stability ball and sitting on an office chair with respect to back and abdominal muscle activation levels and lumbar spine postures.	Participants: University population Age: F: 22.3 (±0.9) M: 25.4 (±5.4) Sample Size: F: n = 7 M: n = 7 Total: n = 14	Task: Four controlled computer workstation tasks while sitting on a stability ball and armless office chair for 1 hr each. Workstation set-up: Armless office chair and a stability ball	Motion analysis system (ISOTRACK 3space) Seat pressure mapping device	Composite factor (pain): Borg scale - Ratings of perceived discomfort (RPD)
7.	Groenesteijn et al. (2012)	The Netherlands	What are the effects of different office tasks on postures and movements of body parts and inclinations of parts of the chair?	Participants: Office workers Age: F: 36.1 (±7.8) M: 35.3 (±7.5) Sample Size: F: n = 8 M: n = 4 Total: n = 12	Task: Typical tasks of office work over 4.5 hours. Workstation set-up: Four different office chairs with various adjustability features	Movement sensors and accelerometers (CUELA system)	Composite factor (discomfort): Questionnaire - Comfort experience in relation to body parts, chair elements and work
8.	Gruevski and Callaghan (2019)	Canada	To determine the effect of age on lumped passive spine stiffness, lumbar postures and discomfort during prolonged seated exposures.	Participants: Community dwellers Age: ‘Young’: 23 (±5); ‘Mature’: 63.7 (±3.9) males and females combined Sample Size: Total: n = 34 males and females combined	Task: Standardised transcribing task on a desktop computer over 90 minutes split into 15 minutes intervals. Workstation set-up: Seated workstation	Motion capture system (Optotrak)	Composite factor (discomfort): VAS-scale -Musculoskeletal discomfort Biological factor (passive spine stiffness) - side-lying passive stiffness test
9.	Hauck et al. (2018)	Germany	To obtain a detailed kinematic analysis of movements that are considered treatment-related or unspecific but nonetheless essential for maintaining everyday work flow.	Participants: Orthodontic residents and orthodontists currently working at German University medical centres Age: 31.5 (±3.8) males and females combined Sample Size: F: n = 13 M: n = 8 Total: n = 21	Task: Main activities of orthodontists: treatment, office and other activities. Workstation set-up: Usual workstation set-up	Movement sensors and accelerometers (CUELA system)	none
10.	Jun et al. (2017)	Australia	To evaluate the reliability of continuous (for 1 hr) measurement of postural behaviour between different periods of office work using quantified technical measures (with wearable motion sensors) of upper body posture (head, thorax, neck, and arm).	Participants: Office workers Age: 33.9 (±11.8) males and females combined Sample Size: F: n = 26 M: n = 5 Total: n = 31	Task: Typical tasks of office work, 1-hr sessions (two morning sessions and one afternoon session). Workstation set-up: Usual workstation set-up.	Axial wireless motion sensors	none
11.	Le and Marras (2016)*	USA	To explore the biomechanics of how different postures induced by different workstations may affect spinal loads and discomfort in relation to movement.	Participants: Community dwellers Age: 26.5 (±8.5) males and females combined Sample Size: F: n = 10 M: n = 10 Total: n = 10	Task: Continuous typing task for 1 h per workstation condition Workstation set-up: three different levels of workstations: seated, standing, and perching	motion capture (OptiTrak) EMG	Composite factor (discomfort): VAS-scale -Musculoskeletal discomfort; Heart Rate Variability (HRV) Biological factor (physical loading): Biologically-driven, EMG-assisted spine model included compression, anterior/posterior shear, and lateral shear loads for the superior and inferior endplate levels from T12 to S1
12.	Mingels et al. (2021)	Belgium & The Nether-lands	To explore if biopsychosocial variables such as pain, psychosocial, and lifestyle characteristics significantly relate to spinal postural variability in patients with cervicogenic headache compared with healthy controls.	Participants: General public with and without cervicogenic headache Age: Control group: 39.2 (±13.1) males and females combined Sample Size: Total n = 36 males and females combined	Task: 30-min-laptop-task includes Workstation set-up: Seated laptop set-up	Motion analysis system (Vicon)	Psychological factor (depression and anxiety): Dutch Depression Anxiety Stress Scale-21 (DASS-21): Degree of depression, anxiety and/or stress Composite factor (pain): Electronic pressure algometer (Somedic AB, Stockholm, Sweden) - Pressure Pain Thresholds (PPTs) (kPa/cm2/s) of the bilateral suboccipitals (cephalic), erector spine at L1 (extra-cephalic), and tibialis anterior (extra-cephalic) Dutch Central Sensitisation Inventory (CSI)- Symptoms of central sensitisation Visual Analogue Scale (VAS)- headache characteristics The Numeric Pain Rating Scale (NPRS) - headache-intensity pre and post the 30-min-laptop-task Composite factor (quality of life): Dutch Headache Impact Test-6 (HIT-6) - Impact of headache on quality of life Composite factor (sleep): Dutch Pittsburgh Sleep Quality Index (PSQI) - Sleep quality
13.	Mingels et al. (2021)	Belgium & The Nether-lands	To evaluate if spinal postural variability differed between patients with episodic cervicogenic headache and an asymptomatic matched control-group during a laptop-task.	Participants: General public with and without cervicogenic headache Age: 39.2 (±13.1) males and females combined Sample Size: Total n = 36 males and females combined	Task: 30-min-laptop-task (recording, data acquisition, storage, gap felling). Workstation set-up: Seated laptop set-up	Motion analysis system (Vicon)	Composite factor (pain): Numeric Pain Rating Scale (NPRS) - Headache-intensity pre and post laptop-task
14.	Tahernejad et al. (2021)	Iran	To determine the sitting behaviour of office workers at ergonomically adjusted workstations, which use appropriate ergonomic features and adjustments.	Participants: Office workers Age: 35.1 (±3.6) males and females combined Sample Size: F: n = 13 M: n = 13 Total n = 26	Task: Typical office work for 1 hour. Workstation set-up: Usual workstation set-up	Posture classification system	none
15.	Waongenngarm et al. (2021)	Thailand	To identify the characteristics of perceived discomfort and different magnitudes of postural shifts during a 4-h sitting period and to examine the association between perceived discomfort and number of postural shifts at different magnitudes.	Participants: Office workers Age: 29 (±3.9) males and females combined Sample Size: F: n = 11 M: n = 29 Total n = 40	Task: Type a standardised text passage at their own normal pace and access the internet to do their work for 4 hours. Workstation set-up: Adjustable office chair with backrest and armrest	A seat pressure mat device	Composite factor (pain): Borg scale - Perceived discomfort measured with

Data synthesis: sex-gender analysis

Although all 15 included papers recruited both men and women as participants in their studies, only four studies (27%) reported on sex-gender separately (De Carvalho and Callaghan 2022a, 2022b; Dunk and Callaghan 2005; Gregory, Dunk, and Callaghan 2006), as shown in Figure 2. None of these studies differentiated between the terms ‘sex’ and ‘gender’

Figure 2. Studies analysing genders separately.

Data synthesis: Factors with potential to explain sex-gender differences in variability.

None of the four studies that analysed sex-gender separately collected psychological or social information to explore the mechanisms operating behind any reported differences in variability between sexes-genders. Factors collected were either biological or composite (pain). However, none of these were used to explain the differences in postural variability between sexes-genders (Table 2).

Table 2. Potential contributing factors captured and theoretical underpinnings used to explain their effect of variability and on sex-gender differences.

Author(s)	Biological factors	Psychological factors	Social factors	Composite factors	Physical environment factors	Theoretical underpinning for effect of factor on variability	Theoretical underpinning for effect of factor on sex/gender differences
De Carvalho and Callaghan (2022a)	Age Height Weight Muscle activity	none	none	Perceived Pain	none	Not provided	Finding: Males moved significantly more than females at baseline. Underpinning: Not provided
De Carvalho and Callaghan (2022b)	Age Height Weight Muscle activity	none	none	Perceived Discomfort	Workstation	Pain Finding: Postural variability increased among pain developers Underpinning: People who develop pain while sitting adjust their posture more often in order to relieve the pain Workstation Finding: No difference in variability between chair conditions Underpinning: n/a Other factors Underpinning not provided	Finding: No difference in fidgets and shifts between male and female participant groups Underpinning: n/a
Gregory, Dunk, and Callaghan (2006)	Age; Muscle activity	none	none	Perceived pain	Workstation	Workstation Finding: No difference in variability between sitting on a chair and on a stability ball Underpinning: n/a Other factors Underpinning not provided	Finding: Males moved significantly more than females Underpinning: Not provided
Le and Marras (2016)	Age, Weight, Height Lumbar spine compression and shear loading	none	none	Perceived discomfort; Discomfort was recorded as a subjective report and as a function of heart rate variability (HRV) over each hour of testing.	Workstation	Discomfort Finding: Postural variability increased among discomfort developers. Underpinning: People who develop discomfort while sitting adjust their posture more often in order to relieve the discomfort Workstation Finding: Variability was lowest at the sitting workstation followed by the perching and then standing. Standing had the highest reports of discomfort and sitting the least Underpinnings: (1). Sitting was more comfortable than perching or standing resulting in reduced postural variability (2) Different workstations create different physical limitations on postural variability Other factors Underpinning not provided	Finding: Women displayed fewer postural shifts than men Underpinning: Not provided

Discussion

Fifteen studies investigating postural variability included male and female participant groups (table 1) however, eleven did not report on sex-gender group differences (table 2). Three of these studies (Hauck et al. 2018; Mingels et al. 2021) stated that because age, height and weight were sufficiently similar across male and female groups all participants could be analysed together as a single group. Collapsing sex-gender groups together based on broad biological similarities invisiblises the relationship between sex-gender and worker health. These studies fail to recognise the ways that contributing factors operate differently for different sexes and genders and, contribute to the assumption that the causes of and interventions for desk-work-related MSk pain can be generalised and standardised across the workforce (Messing and Stellman 2006; Messing et al. 2015).

Of the four studies which reported on sex-gender groups separately three (De Carvalho and Callaghan 2022a; Gregory, Dunk, and Callaghan 2006; Le and Marras 2016) found differences between sex-gender groups (figure 2) but none attempted to explain these findings. It is worth noting that the primary aim of these studies was not the investigation of postural variability differences between sexes-genders and it is therefore unsurprising that they did not closely analyse differences between male and female participant groups. However, reporting sex or gender differences without considering the underlying mechanisms and contributing factors risks attributing outcomes to sex or gender themselves rather than the underlying causes (e.g. attributing greater likelihood of developing pain to ‘being female’ rather than to inequitable treatment at work) (DuBois and Shattuck-Heidorn 2021; Messing and Stellman 2006). Biomechanical and occupational health studies often list female gender as a (non-modifiable) ‘risk’ for a given outcome (Messing and Stellman 2006; Dunk and Callaghan 2005) suggesting that this risk is inherent to being female and cannot be addressed. This can also become an issue of equity with potential to be used as a justification for inequitable hiring and work distribution practices (Messing and Stellman 2006).

Biomechanical studies often justify the conclusion that being female directly affects biomechanical outcomes by attributing sex-gender differences to common biological differences between sexes such as anthropometrics, strength, and fatigue resistance (Côté 2012; Endo et al. 2012; Janwantanakul et al. 2008; Messing and Stellman 2006). In addition to ignoring potential psychological and social explanations, this attribution assumes there is little-to-no physiological overlap between sexes and that physiological differences can be reliably assumed without needing to be measured (e.g. using sex-gender as a proxy for height rather than measuring and making comparisons between participants of different heights) (Messing and Stellman 2006). If a biological parameter such as height or strength is theorised to be the cause of important biomechanical outcomes, making comparisons on the basis of height or strength may be more appropriate than making comparisons on the basis of sex. Alternatively, if the intention is to explain pain differences between sexes and/or genders then grouping by sex or gender may be appropriate.However, underlying mechanisms need to be made explicit and their related factors (including psychological, social and composite factors) need to be explored (Heidari et al. 2016). Failure to consider mechanisms and the multitude of interacting contributing factors is likely to lead to poorly designed interventions which do not address the causes of pain (Messing and Stellman 2006).

Postural variability and pain

Although all four studies which reported on sex-gender differences included a measure of pain or discomfort, only two (Le and Marras 2016; De Carvalho and Callaghan 2022b) explored a potential link between discomfort and postural variability. It is believed that musculoskeletal discomfort can be a precursor to pain at work and may be able to predict future musculoskeletal pain (Hamberg-van Reenen et al. 2008). Le and Marras (2016) and De Carvalho and Callaghan (2022b) suggested that participants increased their postural variability in order to redistribute load and reduce discomfort because maintenance of a single posture over time becomes painful. Although this is a plausible explanation for increasing variability in response to short-term discomfort during a single work session it does not explore the relationship between patterns of variability across multiple work sessions and the development of persistent pain. Furthermore, although one of these two studies (Le and Marras 2016) reported lower postural variability among women they did not consider whether this was due to male and female participants experiencing or being affected by pain in different ways.

Persistent pain, fear and anticipation of pain, and pain-catastrophizing have been associated with motor control changes in people with persistent low back pain, including increased trunk stiffness and reduced postural variability which may persist even after the pain eases (Meier, Vrana, and Schweinhardt 2019; van Dieën et al. 2019). Theories regarding this include persistent pain and fear of pain causing patients to limit their movement (reduced variability) in an effort to avoid provoking or exacerbating symptoms (van Dieën et al. 2019; Bontrup et al. 2019). Others have suggested that pain-related cortical neuroplastic changes and altered trunk proprioception may lead to reduced trunk control and greater trunk muscle co-contraction (Meier, Vrana, and Schweinhardt 2019), contributing to reductions in variability which may itself impact control as reduced movement reduces proprioceptive input (Meier, Vrana, and Schweinhardt 2019). Regardless of how the pain or fear began, resultant reductions in variability may become an additional driver of pain with potential to continue even after the initial cause has stopped. In regard to sex-gender differences, there is evidence of pain processing differences between sexes due to differences in sex hormones, genetics and neuroimmune pain processing which may contribute to greater pain prevalence and/or pain sensitivity among women, though the exact mechanisms involved and extent of sex differences remains unclear (Mogil 2020). Furthermore, there is small body of evidence suggesting that gender may affect the pain experience, with ‘masculinity’ and ‘femininity’ affecting understanding and constructions of meaning regarding pain (Boerner et al. 2018). There is therefore potential for the process and experience of pain to influence posture differently for different sexes and genders.

Pain (particularly chronic pain) may also cause or exacerbate existing elevated stress levels, anxiety and depression, both in relation to the pain itself and through negative impacts on cognition, ability to concentrate and social relationships (Cohen, Vase, and Hooten 2021; Fitzcharles et al. 2021). This may exacerbate higher stress levels already experienced by some women in the workplace, magnifying the effect of stress on both pain and postural variability. Pain also interferes with sleep quality which in turn can exacerbate pain levels, stress, cognitive and concentration difficulties, and pain sensitivity, feeding back into the cycle (Fitzcharles et al. 2021). Furthermore, poor sleep quality has itself been associated with reduced postural variability (Mingels et al. 2021), potentially due it’s impact on fatigue (Fitzcharles et al. 2021) and is more common among women than among men (Decker, Fischer, and Gunn 2022). Sex-gender sleep differences are reportedly related to differing social roles (e.g. family responsibilities) and work conditions with experiences of marginalisation amplifying this difference (Decker, Fischer, and Gunn 2022) and potentially further contributing to sex-gender-related differences in pain and posture.

Postural variability and physical environment

Three studies investigated the effect of working position and workstation design (De Carvalho and Callaghan 2022b; Gregory, Dunk, and Callaghan 2006; Le and Marras 2016) with one (Le and Marras 2016) reporting an effect on postural variability. This study compared workstations with different working positions (sitting, perching and standing) and suggested that some created more postural constraints than others. However, despite finding that women had lower postural variability than men they did not consider whether male and female participants were affected by workstation differently.

Interventions to prevent or reduce pain and to increase postural variability often involve changes to the physical environment. Standing desks have been trialled in several studies and are consistently found to reduce workplace sitting time and increase transitions between sitting and standing. However, results for pain reductions are conflicting (Reliquias and Kuebler 2019; Chambers, Robertson, and Baker 2019). One possible factor behind the inconsistent results is that a gross positional change between sitting and standing does not necessarily change spine or upper limb alignment (e.g. someone sitting in a lordotic posture may also stand in a lordotic posture) resulting ongoing loading of the same structures despite a change in overall position. A study by Yu et al. (2018) found no differences in neck and upper limb alignment between sitting and standing while working at a desktop computer, suggesting that variability at these joints was unaffected by the change in position. The potential for position changes to influence posture is likely different for different workers. While standing intermittently at work may increase postural variability for some workers, a female worker experiencing sex-gender-related work inequities whose persistent scapular elevation is driven by stress may continue to hold scapular elevation in standing, effectively leaving scapular variability unchanged. Similarly, a female worker who keeps her upper limbs close to her sides (contracted posture) in order to conform with gender norms is likely to maintain this posture when standing. Alternatively, a worker whose persistent scapular elevation or contracted upper limb posture is related to the configuration of their workstation may find they are able to change joint alignments with a change position and/or workstation configuration. The potential for standing desks and other workstation alterations to address pain prevalence may therefore depend on intersecting psychological and social factors that affect men and women differently. These factors may therefore play a role in posture and desk-work-related MSk pain prevalence differences between male and female workers.

Multi-directional interactions

Although it appears that the research included in our review has not considered the mechanisms operating behind sex-gender differences in their results (Table 2), we have outlined how aspects of individual psychological experience, social/organisational environment, pain and workstation design could potentially affect variability and, how these might be influenced by sex-gender. Many of the relationships discussed are likely bidirectional and have co-occurring interactions with each other. It is also conceivable that reduced postural variability may itself contribute to social and psychological distress. Movement and postural restrictions, whether physically, socially or psychologically mediated, create limitations on free self-expression (e.g. the expression of femininity or masculinity) (Mason-Mackay 2023). The inability to be and express oneself at work has potential implications for psychological and social wellbeing which may feed back into pathways contributing to pain. Regardless of which factors initiate pain, once pain has started (and especially when it becomes chronic) other factors may become affected and subsequently contribute to, perpetuate or become the primary drivers of pain. While effect sizes for the impact of each factor on pain are often small (Bontrup et al. 2019; Soria-Oliver et al. 2021) this may be reflective of the multifactorial nature of desk-work-related MSk pain rather than a suggestion that any given contributing factor should be thought of as trivial. Studies seeking to understand the mechanisms behind reduced postural variability, the effects reduced postural variability has, or to create effective interventions to address it, need to consider and illuminate the mechanisms and interactions operating behind their results.

Designing prevention and management strategies

The constellation of factors and the mechanisms by which they contribute to postural variability and to pain are likely different for each worker and as such, the interventions which will be effective are likely worker-specific (Petit et al. 2018). Given this it is unsurprising that studies investigating management and prevention approaches often report conflicting results and small effect sizes (Luger et al. 2019; Hoe et al. 2018; Tersa-Miralles et al. 2022; Putsa et al. 2022; Reliquias and Kuebler 2019). Furthermore, authors have noted that when interventions are found to be effective their mechanisms are often unclear, with studies not collecting the information needed to explain their results (Putsa et al. 2022; Zakerian and Subramaniam 2009). The combination of interacting factors contributing to lower postural variability is likely different for different workers (even in the same work environment) and interventions therefore need to be tailored and worker-specific. Workers with less variability due to limitations in flexibility, endurance or poor neuromuscular control may benefit from specific exercise prescription while those with constraining workstations may benefit from a workstation re-design or introduction of a standing desk if standing facilitates postural changes for them. Workers struggling with stress and poor sleep quality might benefit management strategies such as physical activity (Cohen, Vase, and Hooten 2021), breathing, relaxation and mindfulness techniques (Hülsheger, Feinholdt, and Nübold 2015; Wolever et al. 2012) perhaps until such time as the sources of stress and sleep difficulties can be addressed directly. Workers with muscle tension may also benefit from breathing or relaxation interventions (Wixted, O’Riordan, and O’Sullivan 2018) or therapeutic touch and manual therapy interventions (Field 2019; Bialosky et al. 2009). People working under difficult, high-demand or inequitable social and organisational conditions may benefit from organisational and social interventions such as equity and unconscious bias training (Messing, Chadoin, and Blanchette-Luong 2022; Mousa et al. 2021), changes to organisational processes and policies (Mousa et al. 2021) and the facilitation of conversations to improve workplace social dynamics (Maslach and Leiter 2017; Mousa et al. 2021). Finally, for workers with chronic or nociplastic pain, interventions to increase postural variability may need to begin with multi-modal treatment of the pain itself (Cohen, Vase, and Hooten 2021).

Some interventions may be short-term approaches until more fundamental contributors can be addressed or, serve as ongoing management strategies when other factors cannot be changed (e.g. a worker in a stressful environment who is unable to change or leave the job). Although consideration of social and psychological contributors and interventions may be thought to fall outside of usual practice for some physiotherapy clinicians, there may be room for a degree of expansion. When socially and/or psychologically-mediated biomechanical patterns are contributing to pain, social and psychological interventions can be considered as inherent components of biomechanical training. Furthermore, it has recently been argued that ergonomics professionals who traditionally focus on the physical environment of the workplace (e.g. workstation design) can and should assess and make recommendations regarding organisational, cultural and social aspects of the workplace (Messing, Chadoin, and Blanchette-Luong 2022). Being aware of the social and psychological components of postural variability may also help clinicians to make appropriate referrals, discuss contributing factors with their patients, and improve the functioning of interdisciplinary teams.

Limitations and future research

This review has highlighted the paucity of research regarding sex-gender differences in postural variability. The frequency with which sex and gender differences are overlooked and/or under-examined in research has led to the creation of research guidelines regarding sex and gender analysis (Heidari et al. 2016). The guidelines suggest that studies should conduct sex and/or gender analyses regardless of whether the study aim includes sex or gender comparisons, ensure that this analysis includes biological, psychological and social explanations for any differences found, and differentiates between gender and sex. The guidelines further state that in the case that authors consider a sex and/or gender analysis to be inappropriate or unnecessary they should justify this decision.

Given the complexity of intersecting factors and underlying mechanisms involved in postural variability and related symptoms, researchers investigating intervention effectiveness could consider a case-series approach which allows for individualised interventions tailored to the specific contributors affecting each worker.

While researching for this review it was noted that studies occasionally used the term ‘postural variability’ to refer to global positional changes between sitting and standing (Davis and Kotowski 2014; Garrett et al. 2019). We contend that there is an important distinction between the two and recommend using the term ‘positional variability’to refer to changes in global position (e.g. sitting versus standing) and ‘postural variability’to refer to changes in alignment within a given position.

Conclusion

The factors affecting postural variability and contributing to differences between sexes and genders are likely multifactorial and include biological, psychological and social contributors with bidirectional relationships between them. However, studies largely to do not consider sex and gender in their analyses and do not collect the data necessary to explain the mechanisms behind their results. Authors often attribute differences in outcome between sex-gender groups to assumed physiological differences and do not consider psychological or social contributors. A lack of understanding of the potential mechanisms behind lower postural variability makes it difficult to design effective pain prevention and management strategies. Researchers investigating the biomechanical components of work-related musculoskeletal pain need to consider both the contributors to and the effects of lower postural variability across physiological, psychological and social domains. Clinicians intending to guide patients towards greater postural variability need to consider all potential influences on postural mechanics which may involve expanding into areas of practice they not usually consider to be within their role including psychological, social and organisational domains.

Authors contributions

RQ, LA, RA, JE equally contributed to writing the manuscript with guidance and editing from SvN; AMM; ML. RQ, LA, RA and JE contributed to literature searches, study selection and data extraction. Final adaptations and approval were given by SvN, AMM and ML.

Acknowledgements

Open Access funding provided by the Qatar National Library.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.