2025 年 133 巻 1 号 p. 39-43
Several reports have raised concerns about poor posture in children worldwide. As Japan began to experience rapid economic growth toward the end of the 20th century, poor posture (hyperkyphosis) in children became a particular problem. However, little research has been conducted on hyperkyphosis among Japanese elementary school students, and sample sizes have been inadequate. This study aimed to determine the prevalence and characteristics of hyperkyphosis among Japanese elementary school students to help understand and prevent hyperkyphosis. A total of 403 Japanese elementary school students (208 boys and 195 girls) were enrolled in this study. The SpinalMouse system was used to measure thoracic kyphosis and upper thoracic, lower thoracic, lumbar lordosis, and sacral anteversion angles in the standing position. Hyperkyphosis was defined as a thoracic kyphosis angle of >40°. The participants were assigned to two groups: hyperkyphosis and non-hyperkyphosis. The prevalence of hyperkyphosis among Japanese elementary school students was approximately 25.0%. The upper and lower thoracic kyphosis angles were compared separately. The hyperkyphosis group had a significantly higher degree of kyphosis (upper and lower) than the non-hyperkyphosis group. The lumbar lordosis angle was significantly greater in the hyperkyphosis group than in the non-hyperkyphosis group. Additionally, the upper thoracic spine angle was significantly greater than the lower thoracic spine angle in the hyperkyphosis group. A strong negative correlation was found between the lumbar lordosis angle and sacral anteversion angles in both groups. The results of this study indicated a hyperkyphosis prevalence of approximately 25% and suggested that the upper thoracic spine may be characterized by a greater degree of kyphosis in Japanese elementary school students.
A series of recent reports have raised concerns about poor posture not only among adults but also among children, with lifestyle changes such as the increased use of smartphones and home video game consoles thought to be playing a role (Sainz de Baranda et al., 2020a; 2020b).
Thoracic hyperkyphosis is one of the most common forms of poor posture. Feng et al. reported an association between hyperkyphosis and non-specific lower back pain (LBP) in young adults (Feng et al., 2017). In their extensive study of Portuguese children and adolescents, Azevedo et al. found that hyperkyphosis increased the risk of developing LBP (Azevedo et al., 2022). In a prospective study of young baseball players, hyperkyphosis was found to be a risk factor for elbow injuries (Sakata et al., 2017). Therefore, it is of paramount clinical importance to mitigate and prevent hyperkyphosis, which has various effects on the body, including musculoskeletal disorders.
The epidemiology of hyperkyphosis in elementary school students was studies by Sainz de Baranda et al., who reported that among 731 Spanish elementary school students, the prevalence of hyperkyphosis was 27.4% with no significant difference between sexes (Sainz de Baranda et al., 2020a). They also reported that students with thoracic hyperkyphosis presented with lumbar hyperlordosis (Sainz de Baranda et al., 2020a). Koyama et al. found that the prevalence of hyperkyphosis was 30.1% (25/83) in elementary school students in Japan. In addition, the upper thoracic spine angle was significantly greater than the lower thoracic spine angle in the hyperkyphosis group (Koyama et al., 2022). This is important for understanding the (clinical) cases of thoracic hyperkyphosis in Japan, and is helpful for future efforts. However, the sample size in that study was small compared to that of other reports. To enhance the reliability of the results, it is necessary to increase the sample size.
This study aimed to investigate the prevalence and characteristics of hyperkyphosis in Japanese elementary school students and obtain basic knowledge to help understand and prevent hyperkyphosis in the future.
The participants were 403 Japanese elementary school students (208 boys and 195 girls) living in Tokyo. Table 1 presents the data for the grade distribution. Kruusamäe et al. reported that sagittal spinal alignment is affected by competitive sports (Kruusamäe et al., 2015). Children who were not regularly involved in sports were included. The participants were free of musculoskeletal disorders at the time of measurement. The study protocol was approved by the Ethics Committee of Tokyo Ariake University of Medical and Health Sciences (approval number 320). Written informed consent was obtained from all participating elementary school students and their parents before participation. Information regarding the purpose of the study, potential risks, and the protection of participants’ rights was provided. Exercise and applicable measurements were duly conducted from November 2021 to February 2023.
Number of participants at each grade level
| Grade | Male (n = 208) | Female (n = 195) | Total (n = 403) |
|---|---|---|---|
| 1st | 35 | 32 | 67 |
| 2nd | 36 | 32 | 68 |
| 3rd | 34 | 32 | 66 |
| 4th | 31 | 31 | 62 |
| 5th | 34 | 28 | 62 |
| 6th | 38 | 40 | 78 |
The participants’ physical characteristics (height and weight) were determined. Information on grade, age, and current musculoskeletal disorders was obtained using a questionnaire. The Rohrer index was calculated as (weight (kg)/height (cm)3 × 107).
Measurement of sagittal spinal alignmentA SpinalMouse (Idiag AG, Volketswil, Switzerland) was used to measure sagittal spinal alignment. This device is non-invasive and can easily assess sagittal spinal alignment. It is particularly useful for assessing non-structural postural misalignment in the sagittal plane. The participants were asked to stand in a comfortable position with their feet shoulder-width apart and their eyes at the horizon level. Boys wore no clothes from the waist up, whereas girls wore a patient gown open at the back to allow direct skin contact with the SpinalMouse during measurement. SpinalMouse is highly reliable in assessing sagittal spinal alignment, with good intra- and inter-rater validity and reliability compared to radiography (Kellis et al., 2008; Mannion et al., 2014). It has also been reported to be highly reliable in assessing spinal alignment in children (Livanelioglu et al., 2016). The thoracic kyphosis angle was calculated as the angle between adjacent thoracic vertebrae; these were measured and added together from the first to the twelfth thoracic vertebrae, represented by (+). The lumbar lordosis angle was calculated as the angles between adjacent lumbar vertebrae; these were measured and added together from the first to the fifth lumbar vertebrae, represented by (–). The angle between the line joining the first and third sacral vertebrae and the vertical line was defined as the sacral anteversion angle with an anterior tilt (+) and a posterior tilt (–). The upper thoracic angle was calculated as the angle between the adjacent thoracic vertebrae; this was measured and added together from the first to the seventh thoracic vertebrae. The lower thoracic angle was calculated as the angle between the adjacent thoracic vertebrae; this were measured and added together from the seventh thoracic vertebra to the first lumbar vertebra (Figure 1).
Sagittal spine alignments (thoracic kyphosis angle, lumbar lordosis angle, and sacral anteversion angle) in standing positions, as acquired using the SpinalMouse® device
The criterion for hyperkyphosis in this study was a thoracic kyphosis angle of 40° or more, as reported in previous studies by Feng et al. (2018) and Jung et al. (2022) using a SpinalMouse. Participants were divided into two groups based on their thoracic kyphosis angle: those with a thoracic kyphosis angle of 40° or more (referred to as the ‘hyperkyphosis group’) and those with a thoracic kyphosis angle of less than 40° (referred to as the ‘non-hyperkyphosis group’), to investigate the characteristics of hyperkyphosis.
Statistical analysisThe chi-square test was used to assess the differences in the prevalence of hyperkyphosis, sex, and grade levels. The hyperkyphosis and non-hyperkyphosis groups were compared by performing a normality test (Shapiro–Wilk test) for each parameter (physical characteristics and sagittal spinal alignment of the participants), followed by an unpaired t-test. If normality was not found, the Mann–Whitney U-test was used. Pearson’s product–moment correlation coefficient was used to examine the association between each sagittal spinal alignment, and Spearman’s rank correlation coefficient was used when normality was not observed. A P value of less than 0.05 was considered statistically significant. All statistical analyses were performed using IBM SPSS Statistics software (IBM Corp., Armonk, NY, USA).
The prevalence of hyperkyphosis (>40°) among the Japanese elementary school students was 25.3% (104/403). The prevalence of hyperkyphosis (>40°) in the lower (grades 1 and 2), middle (grades 3 and 4), and upper grades was 26.7% (36/135), 28.9% (37/128), and 22.1% (31/140), respectively (P = 0.433). The prevalence of hyperkyphosis (>40°) was not associated with the grade level (Table 2).
Comparison of the prevalence of hyperkyphosis (lower grade vs middle grade vs upper grade)
| HK (n = 104) | Non-HK (n = 299) | |
|---|---|---|
| Lower grade (1st and 2nd) | 36 (26.7) | 99 (73.3) |
| Middle grade (3rd and 4th) | 37 (28.9) | 91 (71.1) |
| Upper grade (5th and 6th) | 31 (22.1) | 109 (77.9) |
Values are presented as n (%). HK, hyperkyphosis.
This study compared 104 patients in the hyperkyphosis group (with a thoracic kyphosis angle of 40° or greater) and 299 patients in the non-hyperkyphosis group (with a thoracic kyphosis angle of less than 40°) with respect to physical characteristics, sex, and sagittal spinal alignment. No statistically significant differences were observed between the two groups in any of the parameters (Table 3). In addition, when the upper and lower thoracic kyphosis angles were compared separately, the hyperkyphosis group had a significantly higher degree of kyphosis than the non-hyperkyphosis group (P < 0.001). The lumbar lordosis angle was also significantly greater in the hyperkyphosis group than in the non-hyperkyphosis group (P < 0.01). However, there was no significant difference in sacral anteversion angle between the two groups (Table 3). In addition, the upper thoracic spine angle was significantly greater than the lower thoracic spine angle in the hyperkyphosis group (P < 0.001). In contrast, no significant differences were observed between the upper and lower thoracic kyphosis angles in the non-hyperkyphosis group (Table 4).
Comparison of the physical characteristics and sagittal spinal alignment of the participants
| HK (n = 104) | Non-HK (n = 299) | P-value | |
|---|---|---|---|
| Age (years) | 9.1 (1.7) | 9.3 (1.9) | 0.492 |
| Height (cm) | 134.6 (11.2) | 135.9 (12.2) | 0.327 |
| Body weight (kg) | 32.2 (8.3) | 32.4 (8.6) | 0.886 |
| Rohrer index | 130.7 (19.3) | 127.5 (17.4) | 0.158 |
| Sex (male:female) | 62:42 | 146:153 | 0.058 |
| Thoracic kyphosis angle (degrees) | 45.2 (4.0) | 27.5 (9.3) | <0.001 |
| Upper thoracic angle (degrees) | 25.9 (5.9) | 12.4 (12.0) | <0.001 |
| Lower thoracic angle (degrees) | 19.5 (5.6) | 15.1 (7.8) | <0.001 |
| Lumbar lordosis angle (degrees) | –24.0 (10.2) | –20.3 (11.5) | <0.01 |
| Sacral anteversion angle (degrees) | 11.1 (9.0) | 12.7 (9.5) | 0.134 |
Values are presented as n or mean (standard deviation). HK, hyperkyphosis
Comparison of the upper thoracic angle and lower thoracic angle
| Group | Upper thoracic angle (degrees) | Lower thoracic angle (degrees) | P-value |
|---|---|---|---|
| HK | 25.9 (5.9) | 19.5 (5.6) | <0.001 |
| Non-HK | 12.4 (12.0) | 15.1 (7.8) | 0.063 |
Values are mean (standard deviation). HK, hyperkyphosis
In both groups, there was no significant correlation between thoracic kyphosis and lumbar lordosis angles (Table 5). A strong negative correlation was found between the lumbar lordosis angle and sacral anteversion angle in both groups (Table 5) (P < 0.001 for both).
Correlation analysis between each sagittal spinal alignment
| Group | Sagittal spinal alignment | Correlation coefficient | P value |
|---|---|---|---|
| HK (n = 104) | Thoracic kyphosis angle vs lumbar lordosis angle | –0.12 | 0.217 |
| Thoracic kyphosis angle vs sacral anteversion angle | –0.06 | 0.504 | |
| Lumbar lordosis angle vs sacral anteversion angle | –0.88 | <0.001 | |
| Non-HK (n = 299) | Thoracic kyphosis angle vs lumbar lordosis angle | –0.14 | <0.05 |
| Thoracic kyphosis angle vs sacral anteversion angle | –0.08 | 0.140 | |
| Lumbar lordosis angle vs sacral anteversion angle | –0.85 | <0.001 |
HK, hyperkyphosis
In this study, the prevalence of hyperkyphosis (>40°) among the 403 Japanese elementary school students was 25.3% (104/403). A comparison of the hyperkyphosis and non-hyperkyphosis groups revealed that children with hyperkyphosis had significantly larger upper thoracic, lower thoracic, and lumbar lordosis angles., The upper thoracic spine angle was significantly greater than the lower thoracic spine angle in the hyperkyphosis group.
A previous study of 731 elementary school students in Spain reported a hyperkyphosis prevalence of 27.4%, which was similar to our results, although the sample size and measurement methods differed from those used in this study (Sainz de Baranda et al., 2020a). The definition of hyperkyphosis and the methods for measuring it vary widely among previous studies, making it difficult to compare the results. Although a previous study reported higher thoracic kyphosis angles in 12-year-old students than in 9-year-old students, this study did not find significant differences by grade level (Sainz de Baranda et al., 2020b). While sagittal spinal alignment may vary with age, the prevalence of hyperkyphosis in elementary school students may not be influenced by the grade level. Given that 25% of Japanese elementary school students present with hyperkyphosis, which can have numerous effects on the body, an ongoing study on sagittal spinal alignment in children is being considered.
Our results also supported the association between hyperkyphosis and hyperlordosis of the lumbar spine in elementary school students, as reported by Sainz de Baranda et al. (2020a). Czaprowski et al. (2018) classified poor posture in children and found different types of postures associated with hyperkyphosis of the thoracic spine. Among them, kyphotic-lordotic and swayback postures were associated with hyperkyphosis of thoracic and lumbar hyperkylordosis. It is possible that many elementary school students with hyperkyphosis in the present study also had kyphotic-lordotic and swayback postures. Moreover, the characteristics of elementary school students with thoracic hyperkyphosis revealed that the curvature was particularly pronounced in the upper thoracic vertebrae in this study. This result is consistent with that of a previous study by Koyama et al. (2022). Because the sample size in this study was larger than that in Koyama et al. (2022), we conclude that this study provides evidence of the characteristics of elementary school students with thoracic hyperkyphosis.
Concerning each sagittal spinal alignment in the standing position, a strong negative correlation was found between the lumbar lordosis angle and sacral anteversion angles in both groups. Although previous studies have been conducted in adults, many have reported a correlation between the lumbar lordosis angle and sacral anteversion angle (Vialle et al., 2005; Xu et al., 2022). In the present study, no significant correlation was found between the thoracic kyphosis and lumbar lordosis angles in either group. Endo et al. (2016) reported a weak correlation between the thoracic kyphosis and lumbar angles in a study of sagittal spinal alignment using radiographs in healthy adults. Although there were differences among the patients, this report contrasts with the findings of our study. The patients may have hyperkyphosis even if the lumbar lordosis angle is normal. We believe that elementary school students with thoracic hyperkyphosis may exhibit poor posture due to other factors, irrespective of the lumbar lordosis angle.
We speculate that these results may differ between adults and children.
The results of this study indicate a hyperkyphosis prevalence of approximately 25%, and suggest that the upper thoracic spine may be characterized by a greater degree of kyphosis in Japanese elementary school students. The results of this study may help improve and prevent poor posture in elementary school students. Therefore, educational programs are necessary for the treatment of hyperkyphosis in Japanese children.
Limitations of the study and scope for future researchThis study had several limitations. First, it assessed only sagittal spinal alignment in the standing position. A previous study on postural changes during seated smartphone use reported that the thoracic kyphosis angle increases over time. Ishii et al. (2017) conducted a survey of children’s sitting habits in Japan and reported that approximately 60% of the participants sat for two hours or more per day, mostly during screen time, such as watching television. In addition, elementary school students are often seated at school, so this position accounts for a high percentage of their daily activities. Therefore, it is necessary to evaluate children’s sitting postures and examine the results together with those of the present study.
Second, this study did not examine the lifestyle of elementary school students, especially the frequency of smartphone use, which may have affected hyperkyphosis. Thus, the relationship between current posture and frequency of smartphone use could not be determined. We plan to conduct further research based on these findings.
Finally, this study did not measure head position or cervical spine alignment, and further research is needed to increase the reliability of the results.
All the authors conceived the study design and protocols. K.K., T.I., K.F., Y.S., A.N., Y.K., S.N., and M.U. were responsible for data collection and analysis, and K.A. supervised the project. K.K. drafted the manuscript.
The authors declare no competing interests.
