Immediate and cumulative effects of upper-body isometric exercise on the cornea and anterior segment of the human eye

Participants

An a-priori power analysis, using the GPower 3.1 software (Faul et al., 2007) and based on the smallest effect size of interest, was performed to determine the required sample size for this study. For this analysis, we assumed an effect size of 0.25, power of 0.80 and alpha of 0.05. This calculation projected a minimum sample size of 17 participants. At this point, 18 healthy young adults (9 women and 9 men: age = 25.3 ± 5.2 years, body mass = 66.3 ± 12.1 kg, height = 172.4 ± 9.1 cm) were recruited to participate in this investigation. Inclusion criteria were: (i) be free of any ocular disease, as assessed by slit lamp and direct ophthalmoscopy examination, (ii) had no history of refractive surgery or orthokeratology, and iii) not be currently taking any medication. Participants were asked to refrain from strenuous exercise 48 h preceding each visit to the laboratory, and also to avoid alcohol or caffeine consumption 12 h prior to each testing session. They were also asked to avoid the use of contact lenses for 8 h prior to the main experimental session (second visit). Participants were sport science students from the University of Granada, they were physically active through their standard academic curriculum, which included ∼6 physical activity classes per week (7.9 ± 2.6 h per week). The University of Granada Institutional Review Board (438/CEIH/2017) granted Ethical approval to carry out the study. All participants gave written informed consent prior to the study.

Anterior eye segment assessment

For the assessment of the eye anterior segment morphology, we used the Pentacam (Oculus Optikgeräte Inc., Wetzlar, Germany). This high-resolution non-contact device uses a rotating Scheimpflug camera, which rotates 360° around the optical axis, and allows to obtain multiple images of the anterior segment of the eye in 2 s, considering the corneal vertex as the reference point (Shankar et al., 2008). The acquired data are used by the Pentacam software to construct a 3-dimensional image of the anterior segment, which permits to calculate several anterior segment parameters. For this study, we considered the pupil size (PS; average diameter of the pupil over the duration of the scan), the anterior chamber angle (ACA; the smaller of the 2 angles taken in the horizontal meridian), anterior chamber volume (ACV; volume between the posterior surface of the cornea and the iris and lens over a 12 mm diameter centered on the corneal apex), and the anterior chamber depth (ACD; the distance from the corneal endothelium to the anterior surface of the crystalline). Regarding corneal morphology, we analyzed data of central corneal thickness (CCT; centered at the corneal apex), corneal volume (CV; volume over a diameter of 10 mm centered on the corneal apex), and keratometry readings (K-steep; corneal curvature in the steep central three mm zone, and K-flat; corneal curvature in the flat central three mm zone)

Upper-body isometric exercise

The isometric biceps curl exercise was always performed bilaterally, in a seated position, with the elbows flexed at 90°. Participants were instructed to maintain a constant breathing pattern (inhaling and exhaling), although an involuntary Valsalva maneuver was expected when leading to muscular failure. The load was applied through two 10-l volume capacity jugs with comfort-grip handles (Vera et al., 2019). The heaviest load that each participant could hold for 2 min (13.2 ± 4.1 kg) during the isometric biceps curl exercise was determined in the first testing session. First, participants estimated the load they perceived that they could hold an isometric contraction for a maximum of 2 min. Then, the isometric biceps curl exercise was performed following the technique described above using the 75% of the estimated load. After 30 s of effort, participants were instructed to stop the exercise when they or an experienced researcher perceived that the applied load could be hold for more or less than 2 min. Afterwards, the magnitude of the load was modified in successive sets by consensus between the participant and the researcher until the heaviest load that each participant could hold for 2 min was identified. Successive sets were separated by 3 min. The main experimental session was performed in a different day and consisted of two sets separated by 10 min. One set was performed holding the heaviest load identified in session 1 (high-intensity condition) and another set holding a load representing half of the heaviest load (low-intensity condition).

Experimental design and procedure

A repeated measures design (2 intensities ×7  points of measure) was used to test the acute effects of upper-body isometric effort performed at two intensities (high and low) on anterior eye biometrics at seven points of measure (baseline, 30, 60, 90 and 120 s during isometric effort, and 30 and 120 s after isometric effort).

Participants attended the laboratory on two occasions. In the first experimental session, participants were screened for the accomplishment of the inclusion criteria, and we individually determined the heaviest load that each participant could hold for 2 min during the biceps curl exercise (see the “upper-body isometric exercise” subsection). In the main experimental session (second visit), upon arrival to the laboratory, participants performed a standardized warm-up consisting of 5 min of jogging and upper-body dynamic stretching exercises. Then, both experimental conditions (high-intensity and low-intensity) were carried out in a randomized order and were separated by 10 min of passive rest. Participants were seated in front of the Pentacam for all the measurements. A total of seven measurements were collected for each experimental condition: one before exercise, four during the 2-minutes isometric exercise (after 30, 60, 90, and 120 s), and two after the isometric exercise (after 30 and 120 s). The isometric biceps curl exercise was always performed bilaterally, in a seated position, and with the elbows flexed at 90°. This position was maintained during the four Pentacam measurements taken during the 2-minutes of upper-body isometric exercise, while the measurements collected before and after exercise were performed in a rested condition (i.e., without performing any isometric effort). All measurements were obtained under constant environmental (∼22 °C and ∼60% humidity) and illumination conditions (∼30 lux).

Statistical analyses

Before any statistical analysis, the normal distribution of the data (Shapiro–Wilk test) and the homogeneity of variances (Levene’s test) were confirmed (p > 0.05). To assess the impact of holding weights of different magnitude on corneal morphology and anterior chamber parameters, we performed a repeated measures ANOVA for each dependent variable with the exercise intensity (high, low) and the point of measure (baseline, 30, 60, 90 and 120 s of isometric effort, and 30 and 120 s after isometric effort) as within-participant factors. The level of statistical significance was set at 0.05, and multiple comparisons were corrected with the Holm–Bonferroni procedure. The Cohen’s d effect size (d) and eta squared (η_p²) were assessed to determine the magnitude of the differences for T and F tests, respectively. The criteria for interpreting the magnitude of the Cohen’s effect size were: trivial (<0.20), small (0.20–0.50), moderate (0.60–1.19), large (1.20–2.00), and extremely large (>2.00) (Hopkins et al., 2009).

Pupil size

PS exhibited a statistically significant effect for the intensity (F_1,17 = 13.05, P = 0.002, η²_p = 0.43), point of measure (F_6,102 = 6.92, P < .001, η²_p = 0.29), and the interaction “intensity × point of measure” (F_6,102 = 4.95, P < .001, η²_p = 0.23). After it, we performed a separate one-way ANOVAs for each isometric effort intensity (high and low) with the point of measure as the within-participants factor. For the high-intensity condition, the point of measure reached statistical significance (F_6,102 = 10.47, P < .001, η²_p = 0.38) due to a lower PS at baseline compared to 30 (corrected P-value = .001, d = 1.24), 60 (corrected P-value = .006, d = 1.05), 90 (corrected P-value = .021, d = 0.86) and 120 (corrected P-value = .020, d = 0.89) seconds during isometric effort. The ANOVA for the low-intensity condition did not show statistical significance for the point of measure (F_6,102 = 1.87, P = .092) (Fig. 1).

Anterior chamber parameters

The analysis of the ACA exhibited a statistically significant effect for the point of measure (F_6,102 = 3.35, P = .005, η²_p = 0.16), but no differences were obtained for the intensity (F_1,17 = 0.70, P = .416) or the interaction “intensity × point of measure” (F_6,102 = 0.61, P = .718). Consequently, we carried out separate analyses for each exercise intensity. Statistical significance differences for the point of measure were showed for the high-intensity condition (F_6,102 = 3.11, P = .008, η²_p = 0.15), but not for the low-intensity condition (F_6,102 = 0.84, P = .543). Post-hoc analyses for the high-intensity condition evidence a narrower ACA compared to baseline after 30 (corrected P-value = .018, d = 0.95), 90 (corrected P-value = .047, d = 0.84) and 120 (corrected P-value = .034, d = 0.88) seconds of isometric exercise (Fig. 2A).

The ACV reached a statistically significant effect for the point of measure (F_6,102 = 9.95, P < .001, η²_p = 0.37), whereas the intensity (F_1,17 = 3.08, P = .098) and the interaction “intensity × point of measure” (F_6,102 = 1.03, P = .412) did not reveal statistical significance. The two separate one-way ANOVAs for each intensity revealed a statistically significant effect of the point of measure for both the high-intensity (F_6,102 = 7.83, P < .001, η²_p = 0.32) and low-intensity (F_6,102 = 4.98, P < .001, η²_p = 0.23) conditions. Post-hoc comparisons revealed for the high-intensity condition a reduced ACV compared to baseline after 120 s of isometric effort (corrected P-value = .002, d = 1.20), as well as after 30 s of passive recovery (corrected P-value = .021, d = 0.92). No post-hoc differences were observed in the low-intensity condition (Fig. 2B). Lastly, the analysis of ACD did not show differences for the intensity (F_1,17 = 1.86, P = .191), point of measure (F_6,102 = 1.56, P = .165), or the interaction “intensity × point of measure” (F_6,102 = 0.60, P = .727).

Pachymetry

CV did not show statistically significant differences for the intensity (F_1,17 = 0.05, P = .819), point of measure (F_6,102 = 0.38, P = .889), or the interaction “intensity × point of measure” (F_6,102 = 0.46, P = .996). Similarly, the main factors of intensity (F_1,17 = 0.73, P = 0.406) and point of measure (F_6,102 = 0.90, P = .497), as well as the interaction “intensity × point of measure” (F_6,102 = 0.46, P = .837) were far from reaching statistical significance for CCT.

Keratometry readings

For K-flat, a significant effect was found for the point of measure (F_6,102 = 5.29, P < .001, η²_p = 0.24), whereas no differences were obtained for the intensity (F_1,17 = 0.41, P = .528) or the interaction “intensity × point of measure” (F_6,102 = 1.44, P = .208). Separate unifactorial ANOVAs revealed a statistically significant effect of the point of measure for the high-intensity condition (F_6,102 = 5.09, P < .001, η²_p = 0.23), but not for the low-intensity condition (F_6,102 = 1.88, P = .092). Post-hoc analyses evidenced in the high-intensity condition a lower K-flat compared to baseline measurements at 90 (corrected P-value = .006, d = 0.88) and 120 (corrected P-value = .001, d = 1.10) seconds of isometric effort. Also, the K-flat measured after 120 s of upper-body isometric exercise was lower than the obtained after 30 (corrected P-value = .017, d = 0.81) and 120 (corrected P-value = .029, d = 0.78) seconds of passive recovery (Fig. 3A).

The analysis of K-steep did not reach statistically significant differences for the point of measure (F_6,102 = 2.06, P = .069), the intensity (F_1,17 = 0.01, P = .938) and the interaction “intensity × point of measure” (F_6,102 = 0.48, P = .825) (Fig. 3B).