Impact of handstand on airway resistance and pulmonary diffusing capacity in healthy humans

The upright posture has long been a central theme in philosophical reflections on human nature, as well as the subject of scientific investigation and basis for practical measures aimed at reinforcing social norms [1]. It has been argued that it is through discourse on the upright body that humanity has modelled its understanding of itself [2]. Thus, humans have long experimented with acrobatic endeavours, among which the handstand occupies a distinctive place. The earliest known reference to the handstand posture is that practised in yoga traditions dating back to the 10^th century [3]. Popularised in the late 19^th century through the rise of circus performance, as well as artistic and competitive gymnastics, the handstand arguably epitomises inversion of the upright posture in humankind’s quest for meaning [2].

In gymnastics, the handstand represents an inherently unstable balance position and constitutes one of the fundamental elements within numerous movement structures, often serving as both the initial and final position [4]. Its successful execution depends on the integration of inputs from the vestibular apparatus, visual cues, and proprioceptive information, as well as core strength and muscular endurance [4, 5]. However, an often overlooked factor, readily recognised by practising gymnasts as a limitation on handstand duration, is the impact on breathing.

There is currently a notable paucity of literature addressing postural influences on the respiratory system in gymnasts [6]. One earlier study reported that ventilation during handstand becomes more difficult owing to abdominal contents exerting increased pressure on the diaphragm [7]. Nevertheless, the precise impact of handstand on the mechanics of breathing and pulmonary gas exchange remains entirely unknown. Unsurprisingly, measuring pulmonary gas exchange in an inverted position is challenging; however, recent advances in portable equipment now allow for investigation across various postures. In the present study, we aimed to investigate the impact of posture on respiratory mechanics and pulmonary gas exchange in healthy individuals, comparing airway resistance and pulmonary diffusing capacity obtained during a handstand with those in upright, supine, and quadruped postures.

MATERIALS AND METHODS

Ethical approval

This project was approved by the Regional Ethical Committee of the Capital Region of Denmark (H-24047914) and conducted in accordance with the most recent guidelines of the Declaration of Helsinki. All participants provided oral and written informed consent.

Study design and setting

Twelve participants were enrolled in the study conducted at the Department of Clinical Physiology and Nuclear Medicine and Centre for Physical Activity Research, Rigshospitalet, Denmark, from November 25, 2024, to December 31, 2024. The study consisted of two study days, denoted as visit 1 and visit 2 (Figure 1), conducted within a four-week period with at least 24 hours in between, due to limitations on the number of measurements allowed on a single day. Before each study day, participants were instructed to refrain from alcohol, vigorous exercise, smoking and caffeine.

Participants

The inclusion criteria consisted of gymnasts who were able to perform a two-minute handstand against a wall and were aged 18-40 years, regardless of sex. Exclusion criteria included any known heart, lung, or joint disease, pregnancy, and symptoms of disease within the two weeks preceding participation.

Measurements

Medical examination

The medical examination consisted of auscultation of the heart and lungs, a review of the medical history, measurement of blood pressure and heart rate, measurement of height (nearest cm) and weight (nearest 0.1kg), and calculation of the BMI (kg/m²).

Lung function testing

Lung function testing, including dynamic spirometry and whole-body plethysmography, was performed on a Jaeger MasterScreen PFT pro System (CareFusion, Höchberg, Germany) by trained personnel following standardised protocols [8, 9]. Forced expiratory volume in one second (FEV₁), forced vital capacity (FVC), total lung capacity (TLC) and residual volume (RV) were determined.

Impulse oscillometry and blood pressure measurements

Impulse oscillometry measurements were performed using tremoFlo C-100 (THORASYS, Thoracic Medical Systems, Quebec, Canada). The participant was placed in the given posture immediately before the measurements and equipped with a nose clip. A 20-second measurement period with normal tidal breathing was conducted. For upright, supine, and quadruped postures, mean values of three acceptable manoeuvres were analysed, accepting a coefficient of variation (CV) for resistance at 5 Hz (R₅) up to 15% [10]. In handstand, three manoeuvres were conducted: if CV≤15% all measurements were analysed; if >15%, the two most similar measurements were analysed. Reported values included R₅, frequency dependence of resistance at 5-20 Hz (R_5-20), reactance area (AX), and tidal volume (V_t).

Additionally, blood pressure was measured (Microlife, Widnau, Switzerland) during handstand on a relaxed leg and during quadruped posture on a relaxed arm. Height correction of blood pressure was made (cm × 0.7).

Single-breath pulmonary diffusing capacity

Single-breath diffusing capacity measurements were performed using EasyOne Pro (Zürich, Switzerland). Before each measurement, the flow and gas analysers were zeroed, while the participant was placed in the given posture. The participant started normal tidal breathing, followed by maximal unforced expiration to RV. Subsequently, the participant inhaled rapidly to TLC, held the breath for 8 to 10 seconds, and exhaled to RV. During the manoeuvre, standardised gas fractions (0.3% CO, 10% He, 21% O₂, and balance N₂) were inhaled via a dedicated gas analyser system (CiTicel, City Technology, Nuremberg, Germany). Haemoglobin (Hb) levels were measured (to the nearest 0.1 mmol/L) by HemoCue Hb 201+ (HemoCue, Brønshøj, Denmark). The measurement was repeated following a four-minute wash-out period with the participant standing upright. All tests were assessed for acceptability and repeatability criteria and graded according to predefined criteria [11, 12]. Additionally, saturation, and heart rate were measured. Reported values included D_L,COc, carbon monoxide transfer coefficient (K_CO), V_A, inspiratory volume (V_IN), V_IN% of maximal V_IN and breath-hold-time.

Key outcome measures

Between posture differences in R₅, R_5-20, AX, and D_L,COc.

Potential sources of bias

We expect measurement bias to be low, as all measurements were performed according to standardised guidelines with trained personnel. However, during quadruped, supine and handstand postures, achieving test acceptability and repeatability for pulmonary diffusing capacity was limited. Tests not fulfilling these criteria were reported according to established criteria [13].

Sample size

In an exploratory study like this, a sample size of 12 participants is recommended as a general guideline, based on feasibility and precision regarding the mean and variance [14].

Statistical methods

Statistical analysis was handled in RStudio, version 4.3.3 [15]. The statistical code for the analysis will be publicly available on GitHub [16].

Analysis used a linear mixed model (LMM), with the LMMstar package [17]. Assuming an unstructured covariance pattern to account for repeated measures in the same subject. Sex (assigned at birth, male or female), height (numerical), and posture (categorical: upright, handstand, supine, quadruped posture) were included as fixed effects. Missing data were handled implicitly by restricted maximum likelihood estimation in the LMM. Models were assessed for goodness of fit visually by assessing the qq-plot and the residual vs. predicted values. Outcomes were not adjusted for multiple testing.

Results

Participant characteristics

Six men and six women, with a mean age of 25.3 (5.31) years, a BMI of 23.7 (2.4) kg/m² and normal lung function, were included (Table 1).

Impulse oscillometry

Data are represented in Table 2. R₅ increased from upright standing to supine (p < 0.001) and handstand (p < 0.001), whereas R_5-20 increased from upright standing to handstand (p = 0.019). AX increased from upright standing to handstand (p < 0.001). V_t was similar between all postures. Measurement order did not significantly affect outcomes.

Diffusing capacity

Data are represented in Figure 2 and Table 2. D_L,COc (Figure 2A) and K_CO (Figure 2B) increased from upright standing to supine (D_L,COc, p < 0.001; K_CO, p < 0.001), quadruped posture (D_L,COc, p = 0.028; K_CO, p = 0.018), and handstand (D_L,COc, p < 0.001; K_CO, p < 0.001). A decrease in V_A (Figure 2C) and V_IN was found from upright standing to handstand (V_A, p = 0.040; V_IN, p < 0.001). Measurement order did not significantly affect outcomes.

DISCUSSION

The findings indicate that handstand has an impact on pulmonary mechanics and airway resistance. Compared to upright standing, R₅ increased to a similar extent as observed in supine posture. However, R_5-20 and AX increased only during the handstand. With respect to pulmonary gas exchange, both D_L,COc increased relative to upright standing, similarly to changes seen in supine posture.

The finding that R₅ increased from upright standing in both supine posture and handstand, while R_5-20 rose only in the latter, indicates that overall airway resistance increased in both postures; however, small airway resistance was also affected during the handstand. This may reflect a more pronounced compression of small airways and lung tissue, likely due to greater cranial displacement of the heart and abdominal contents, which also accounts for reductions in V_A during handstands [6]. Furthermore, an increase in AX during handstand suggests a concomitant decrease in pulmonary compliance, rendering the lungs less expandable, which is likely attributable to a reduction in compliance of the basal region of the lungs.

Transition from upright standing to handstand increases pulmonary diffusing capacity to levels similar to those in supine and quadruped postures, with a reduction in V_A during the handstand. The effects of supine posture on diffusing capacity are well documented, reflecting central mobilisation of blood volume with concomitant recruitment and distension of pulmonary capillaries, enhancing interface for gas exchange and improving ventilation-perfusion matching, and this also applies to the prone posture, where the direction of the gravitational vector is similar to that of quadruped [18-20]. It may be speculated that greater capillary recruitment and distension, as well as better ventilation-perfusion matching, underlie the diffusing capacity observed in handstand despite a lower V_A. Greater centralisation of blood volume during a handstand likely reflects the gravitational vector directing blood from both lower extremities and splanchnic circulation towards the thoracic cavity.

In terms of ventilation-perfusion matching, it is essential to consider that in the upright posture the human lung resembles a triangular-based pyramid, conical in the transverse plane with its apex oriented superiorly [21]. As only a small apical area is in contact with the thoracic wall, this anatomical configuration gives rise to a pleural pressure gradient that increases intrapleural pressure from the apex to the base, directing regional ventilation in an apical-to-basal manner. Additionally, a perfusion gradient further contributes to ventilation-perfusion mismatch even in healthy lungs [18]. When the gravitational relationship is inverted, due to the pyramidal shape of the lung, the organ effectively hangs from its base at the diaphragm. In this configuration, a greater volume of lung tissue becomes fixated at the diaphragmatic surface, which theoretically reduces the pleural pressure gradient. Moreover, the geometry of the lung in this posture would be expected to diminish the basal-to-apical perfusion gradient, thereby decreasing ventilation-perfusion mismatch. Nonetheless, in another mammal known to spend 70-90% of a 24-hour period in an inverted posture, the bat (Chiroptera), a predilection site for tuberculosis in the basal regions of the lungs has been documented. This finding implies that alveolar PO₂ is highest in the basal areas, the inverse of the pattern typically seen in humans, likely due to posture-induced differences in ventilation-perfusion relationships [22, 23]. While these implications may be unsettling for the bats, they thus offer intriguing insights for physiologists.

The findings also have relevance for gymnasts, a population that spends a considerable proportion of their time inverted. They may appreciate that pulmonary gas exchange appears to be largely preserved. Also, clinically postural modification by »proning« is one of the most well-documented strategies for improving pulmonary gas exchange in patients with acute respiratory distress syndrome [18]. The greater improvement observed in handstand raises the question of whether the fully inverted posture might also benefit gas exchange in such patients. However, we posit that the practical application of handstand in critical care remains entirely unfeasible.

In terms of limitations, the small sample size may have led to type II errors; however, our findings align with earlier findings and the physiological rationale. Additionally, even with handheld instruments, not all measurements met the acceptability criteria for diffusing capacity and airway resistance measurements. Acceptability criteria mostly failed due to limitations on the volume of air inspired during a handstand.

In conclusion, our findings indicate that handstand imposes constraints on respiratory mechanics, particularly by increasing small airway resistance and reducing pulmonary compliance, accompanied by a lower V_A due to compression from abdominal contents and the heart, pulmonary diffusing capacity is concurrently improved. This is conceivably attributable to increased pulmonary capillary recruitment and distension resulting from enhanced systemic venous return to the pulmonary circulation, as well as a more favourable ventilation-perfusion distribution for pulmonary gas exchange. Even though the handstand may epitomise the inversion of the upright posture in humankind’s quest for meaning, these findings likely offer very little insight into the philosophical question of what it means to be human. They do, however, point towards a truth well known to scientists: sometimes we need to flip things upside down to gain a better perspective!