*4.1. Global Spine Parameters*

The results for the global spine parameters derived from the current study did not differ greatly from those of previous publications using the same VRS measurement device ([19–23]; Table 6). However, there seems to be a trend towards slightly lower measurement results for the parameters in the transverse and the coronal planes. Possible explanations for the deviation of the results could be, in addition to the different cohort compositions and cohort sizes, differences in the measurement protocol and data analysis. To obtain the most accurate data quality, we used a high standardization of the measurement protocol, the use of additional markers in the course of the vertebral column (~T3 and ~T12) and the systematic removal of extreme outliers from the raw dataset. Since the comparative studies do not provide corresponding information, the question regarding the reasons for the differences cannot be answered in a well-established manner.

In the current study, significant trends towards possible differences in several global spine parameters according to sex and age cohort were revealed through explorative data analyses. While these differences were not found to be systematic for the coronal plane parameter of the "(Right Side) Apical Deviation VP-DM + max (mm)", the results for the respective sagittal plane parameters were considered highly important. The "Lumbar Lordosis (ITL-ILS) (◦ )" angle revealed a trend towards a significant difference between the female and male participants, with females showing greater lordosis angles than their male counterparts. This is in accordance with previous publications using the same VRS measuring device [19–21]. These findings also match the results of a recent systematic review and meta-analysis describing age- and sex-based effects on the lumbar lordosis angles and the range of motion based on different clinically established measurement approaches (radiological and non-radiological). The authors also found significant differences according to sex, with females having greater lumbar lordosis angles than men, but in contrast to the current findings, they also revealed indications that age possibly affected the respective spine parameters [32]. Similar correlations between VRS- and X-ray-measured results were found for the sagittal plane parameter of "Thoracic Kyphosis". The current study found a trend towards significant differences between young and old (*p* < 0.001) and between middle and old (*p* < 0.02) participants, indicating an increase in the VRSmeasured parameter with increasing age. Comparable results were published in a recent systematic review based on radiography-based Cobb angle calculations [33]. The authors described an increase in thoracic kyphosis with aging but did not find that sex affected the spine parameters.

These results confirm the importance of having VRS reference data that are not only distinguishable between subgroups according to sex, as has been the case thus far, but also according to different age cohorts. Without these data, changes in spinal posture that physiologically occur over a healthy life span could be falsely diagnosed as pathologic, resulting in troubling uncertainty for the affected patients in clinical practice.

#### *4.2. Specific Spine Parameters*

Apart from a previous sub-analysis of 100 female participants from the present study cohort, which was published by our own research group [24], this is the first paper to present systematic reference data for every vertebral body in all three dimensions for both sexes and for three different age cohorts, which were derived using the VRS "Formetric III 4D" measuring device. As already described, in contrast to the currently common clinical beliefs, the physiological spinal posture in the transverse and the coronal planes was not found to be straight and symmetric with regard to rotation and lateral flexion [24]. Instead, a systematic rotation of the mid- and lower thoracic and lumbar vertebrae towards the right side was observed and supported by explorative data analyses. This rotation was found to be more pronounced in males than females and in young compared to middle-aged and old participants. A pre-existing vertebral rotation in healthy participants was also previously described based on CT and MRI measurements [34]. In patients with a diagnosed situs inversus totalis, the side of the vertebral rotation changed, respectively [35]. These results suggest that the VRS findings of the current study are clinically comprehensible and that internal organ arrangement might be a possible physiological cause of the observed asymmetric spinal posture.

In the coronal plane, a lateral flexion to the right side was found in the case of the upper thoracic vertebrae, while the underlying vertebrae showed a systematic lateral flexion to the left. Contrary to the results for the spinal rotation, visually, the lateral flexion in the upper thoracic vertebrae seemed to be more pronounced in the female compared to the male participants, whereas no sex-specific differences were detectable in the mid- and lower thoracic or the lumbar vertebral body positions. According to age-related differences, the younger group visually seemed to demonstrate less lateral flexion than the middleaged and old participants, specifically in the thoracic spinal region. Whether or not this might be caused by posture adaptations induced by normal degenerative changes in the spine remains unclear. However, Kilshaw et al. [36], who analyzed the lumbar spine retrospectively based on abdominal radiographs and found that deformities such as lumbar scoliosis, lateral listhesis and osteoarthritis in the coronal plane started to occur after the age of 50 and steadily increased with age, previously described such an effect, albeit in a different spinal section.

No unexpected outcomes for the vertebral positions were detected in the sagittal plane parameters. The apex of the VRS-measured thoracic kyphosis, meaning the least tilted vertebrae, appeared around T8. This is in accordance with recently published findings derived from radiographic data. Here, the thoracic apex was located between T7 and T9 [37,38]. In the current study, the lumbar lordosis apex appeared between the second and the fourth lumbar vertebrae, depending on the analyzed subgroup. As already described, a statistically significant trend towards a difference between the female and male participants was found for the global parameter of "Lumbar Lordosis (ITL-ILS) (◦ )". This difference according to sex was also apparent among almost all the specific spine parameters, except for the two major curvature turning points (the most flexed (T1) and the most extended (L1) vertebral body). In the thoracic spine, males showed a greater curvature in the upper thoracic spine, and females showed greater curvature in the lower thoracic spine. The authors assume that this difference between females and males canceled each other out, which is why the sex difference did not manifest in the global variable of "Thoracic Kyphosis (ICT-ITL) (◦ )". Nevertheless, the global parameter of "Thoracic Kyphosis (ICT-ITL) (◦ )" revealed significant differences between the analyzed age cohorts caused by the significant age differences in the respective specific parameters of the upper thoracic spine. Sex differences in spinopelvic alignment and in per-level vertebral inclination have also been reported in healthy participants based on upright low-dose digital biplanar X-ray analyses [39]. Similar to the current study, more dorsally inclined vertebrae were found

in females than in males from T1 down to L2. In the current study, females showed more extended vertebral positions from T2 to L3 compared to their male counterparts. Even though the isolated raw values for vertebral inclination differed slightly between the two measurement approaches, the functional comparability of our results with those derived from X-ray analyses further supports the clinical importance of VRS as a non-invasive and simultaneously quick and easy assessment tool for spinal posture analysis in daily clinical routines.

However, despite the described functional agreement between the results based on the VRS and X-ray based measurements, the visual differences in the graphical data outputs between the analyzed subgroups could not be confirmed through statistical analyses for the specific spine parameters in the transverse or the coronal plane in the current study. One reason for this might be the large standard deviation identified in each of the analyzed variables. A normal spine anatomy also means that the results for the parameters in the transverse and the coronal planes are distribute naturally in a preferably narrow corridor reflecting an almost neutral spine position. No physiologically large differences are expected. Due to the high individuality displayed in the large standard deviation and those additional anatomical conditions, the analyzed sample of 201 healthy participants (and, partly, less than *n* = 67 in the respective subgroups) might have simply been too small to detect potential significant differences in the respective planes.

The fact that, in contrast, statistically significant trends in the possible differences between the analyzed subgroups were revealed for the parameters in the sagittal plane might be due to the presence of the two physiological major spinal curvatures, "thoracic kyphosis" and "lumbar lordosis". These anatomical conditions mean that there is a higher natural deviation from the neutral position throughout the whole vertebral column, making it easier to detect statistically significant deviations between the analyzed subgroups even in this "small" sample of 201 healthy participants.

The possibility that sagittal plane parameters will be suitable for detecting differences between subgroups and between different pathologies is in accordance with previously published research [40,41]. Artificial intelligence (AI)-driven analyses of VRS measurement results also found sagittal plane parameters to be one of the most important features with which to distinguish pathology-independent spinal posture data from healthy comparative datasets [41]. Similar results were found when dynamic VRS gait data were analyzed by AI-driven methods. Here, the parameters of the coronal and the sagittal planes were most relevant for the classifications between the sexes [40]. Whether or not sagittal spine parameters have the potential to systematically distinguish between physiological and pathological spinal postures and which parameters are specifically involved must be investigated further in the future. Nevertheless, the first results point in this direction.

#### *4.3. Limitations*

The manufacturer of the "Formetric III 4D" system recommends the use of reflective markers for spinal posture analysis only when the software is not able to identify the required visual landmarks (vertebra prominens (VP) and the two lumbar dimples (DM)) on its own. However, the use of three reflective markers for the landmarks is necessary for dynamic gait analysis. As this study is part of an overarching research project aiming to collect reference data for spinal posture in the habitual stance and when walking at four different walking speeds among the same healthy study cohort, it was necessary to mark all the participants with the three markers in order to render the stance and gait results comparable with each other. Software misinterpretations that arose in advance during the test measurements at the fast walking speeds, caused by the soft tissue and scapular motions of the participants' back surface, meant that the researchers decided to use two additional markers (~T3 and ~T12) to stabilize the systems' dynamic data analysis procedures. The researchers also decided to mark C7 and the PSIS instead of the VP and DM. This approach was chosen because marking C7 and the PSIS is recommended in cases where the VP and DM are not clearly identifiable on the surface of the participant's back. In order to standardize the measuring procedure and to render the results as comparable to each other as possible, the marking of C7 and the PSIS was determined a priori for all the participants, even though this technique was definitely more prone to palpation bias [42,43]. Furthermore, only participants with a BMI of <sup>≤</sup>30.0 kg/m<sup>2</sup> were included in the study due to data capture requirements. The procedures described enabled the collection of highly standardized data under controlled laboratory conditions. Nevertheless, the researchers are aware that this approach limits the external validity of the presented findings and, thus, their direct transferability to clinical practice.

The large number of parameters analyzed and the resulting high number of tested hypotheses also mean that the significant results have to be interpreted with caution. They are more suitable for showing trends in possible differences rather than real statistical significance. In this regard, it must also be mentioned that, retrospectively, the chosen sample size seemed to be too small to detect potential differences between the analyzed subgroups, especially for the specific spine parameters.

Finally, yet importantly, their radiation and contact-free nature mean that results derived from VRS measurements are calculated and based on mathematical algorithms. Even though their validity has been investigated in various publications, those studies mainly focused on comparisons between X-ray and VRS data captured from patients affected by different spinal pathologies (mainly scoliosis) [8,14,15,44,45]. For ethical reasons, no such comparative studies based on healthy participants are available. The results presented here, however, reveal a strong functional agreement with the results derived from clinically established measurement approaches, such as X-ray or MRI/CT scans, and VRS measurements [33,34,39]. This underlines the potential of VRS to serve as a noninvasive, quick and objective alternative for spinal posture analysis in clinical practice, especially when the therapeutic focus lies in function-orientated clinical outcomes and when pre-post measurements are required.

#### **5. Conclusions**

This study complements the existing VRS reference datasets for global spine parameters by adding normative values for different subgroups according to sex and age over an adult life span from 18 to 70 years. The closure of this gap, retrospectively, was found to be very important, because relevant changes over the life span in the isolated spine parameters became visible. Reference values for the specific spine parameters of every vertebral body from C7 to L4 in all three dimensions according to sex and age were presented and revealed visual but statistically non-significant differences between the analyzed subgroups. The sagittal plane parameters seem to have the greatest potential to detect differences between groups of participants. Whether or not those variations are possibly significant must be investigated in future studies by repeating the current project with an appropriate number of healthy participants.

The great variation in and individuality of the spinal posture displayed in the large standard deviation of the analyzed parameters, which was described previously by our research group using data derived from VRS measurements of asymptomatic female volunteers [24], were confirmed for the respective subgroups in the current study. Most importantly, and against widespread clinical expectations, the healthy human spine was found to be systematically asymmetric in the transverse and the coronal planes during upright habitual standing. There needs to be discussion in the therapeutic setting about whether approximations to an almost symmetrical spine in the respective planes are biomechanically desirable in any way [24].

**Author Contributions:** Conceptualization, J.H. and U.B.; methodology, J.H. and U.B.; software, J.K., C.W., R.W. and I.S.; validation, J.H., J.K., C.W., R.W. and I.S.; formal analysis, J.H., J.K., C.W., R.W. and I.S.; investigation, J.H.; resources, P.D. and U.B.; data curation, J.H., J.K., C.W., R.W. and I.S.; writing—original draft preparation, J.H.; writing—review and editing, J.K., C.W., R.W., I.S., P.D. and U.B.; visualization, J.H., J.K. and C.W.; supervision, J.K., P.D. and U.B.; project administration, J.K., P.D. and U.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted in accordance with the Declaration of Helsinki and was approved by the responsible ethics committee of the Rhineland-Palatinate Medical Association. The study is registered with the World Health Organization (WHO) (INT: DRKS00010834).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** Due to ethical and privacy reasons, the data presented in this study are not publicly available.

**Acknowledgments:** The authors would like to express their gratitude to all the participants for their time and interest in participating in this study. Our colleagues are also acknowledged for their professional expertise and contributions to this project and their strong support during the participant recruitment process. Special thanks also go to Amira Basic and Kjell Heitmann (DIERS Company) for their technical support and assistance.

**Conflicts of Interest:** The authors declare no conflict of interest. Technical assistance was provided by staff members of the DIERS Company in preparation for the statistical data analysis process; however, there was no external influence on the design of the study; the collection, analysis, or interpretation of the data; the writing of the manuscript; or the decision to publish the results.
