Scoliosis in elderly patients may cause severe low back pain, leg pain, neurological deficits, and trunk imbalances in coronal and sagittal planes. It is defined as a spinal deformity with a coronal Cobb angle of >10 degrees in skeletally mature patients.1 A study in the United States found a prevalence of lumbar scoliosis (Cobb angle ≥11 degrees) of 8.9% in 2973 individuals aged 40 years or older, and a close positive association between age and scoliosis prevalence.2 There is an almost linear increase in scoliosis prevalence from the fifth to eighth decade of life, with a sharp increase in patients over 90 years old. Therefore, the number of patients with adult spinal deformity can be expected to increase rapidly in aging populations.
Spinal curvature not only causes musculoskeletal disorders, but can also cause disorders of visceral organs close to the spine. Scoliosis patients with a severe thoracic curve can develop pulmonary problems, such as decreased total lung capacity,3 and reductions in forced vital capacity have been observed in adult idiopathic scoliosis patients with scoliosis >40 degrees or kyphosis >50 degrees.4 In adults with a degenerative spine, sagittal deformities such as lumbar kyphosis or vertebral fractures of the thoracic or lumbar spine have been associated with gastroesophageal reflux disease (GERD), hiatal hernia, and other disorders of the digestive organs.5–8 Imagama et al9 suggested that the lumbar lordosis angle, sagittal balance, and strength of the back muscles significantly affect the presence of GERD symptoms.
GERD is a chronic disorder involving a retrograde flow of the gastroduodenal contents into the esophagus, causing heartburn and acid regurgitation. In GERD’s multifactorial pathophysiology, the function of natural antireflux mechanisms, such as esophageal peristalsis, the thoracic-abdominal pressure gradient, the lower esophageal sphincter, and the diaphragm are impaired.10 A hiatal hernia, which can weaken the lower esophageal sphincter and allow stomach contents to move back into the esophagus, is another important predisposing factor for GERD. The prevalence of GERD with weekly or more frequent reflux symptoms is reported to be 10% to 20% in Europe and North America,11 and 6.5% to 9.5% in Japan.12 However, the prevalence of GERD is increasing in these ageing societies. Questionnaires such as the Frequency Scale for the Symptoms of Gastroesophageal Reflux Disease (FSSG)13 and the Questionnaire for the Diagnosis of Reflux Esophagitis (Quest)14 indicate a prevalence of GERD as high as 37.6% in Japan.15
The association between scoliosis and GERD in adult patients has not been precisely evaluated. We hypothesized that spinal deformity in the coronal plane, as in the sagittal plane, at the thoracolumbar/lumbar spine in elderly patients may contribute to the development of gastric and esophageal disorders. This study was conducted to elucidate the influence of trunk deformities on GERD.
MATERIALS AND METHODS
Patients with 40 years of age or older were considered for this study, regardless of whether there was any spinal deformity, if they were seen at our hospital for spinal disorders, and had undergone whole-standing frontal and lateral spine x-rays. Patients were asked to fill out the Quest questionnaire. Whole-spine x-ray and Quest were taken prospectively from 236 patients from 2010 to 2012. Forty-six patients were excluded from the study, 40 patients due to incomplete answer to Quest, 5 patients due to previous spinal surgery, and 1 patient due to oral steroid intake. There were no patients who had undergone surgical treatment for esophageal, gastric, or duodenal intestinal diseases.
Finally, 190 patients (mean age 70.6±8.6 y; range, 44–90 y old; 51 males and 139 females) were included in the study. The diagnoses for these patients included degenerative lumbar kyphoscoliosis, lumbar canal stenosis, lumbar disk hernia, degenerative lumbar spondylosis, cervical myopathy, and degenerative cervical spondylosis. Data such as body weight, height, and drug regimens were collected for each patient.
Quest was developed by Carlsson et al14 to screen GERD patients. It consists of 7 items that evaluate the sensation experienced by patients and the relationship between the occurrence of symptoms and factors known to provoke, exacerbate, or relieve those symptoms. Each item is assigned a positive, neutral, or negative score, and the scores are added to obtain a total ranging from −7 to +18. A Japanese language translation of Quest was found to have a sensitivity of 65%, a specificity of 74%, and an accuracy of 73% in a Japanese population when the cutoff score was set at 6 points.16 In the present study, patients with a Quest score of ≥6 points were considered positive for GERD.
The following radiographic parameters were measured. In the lateral view, the T5–T12, T10–L2, and T12–S1 (lumbar lordosis) angles; the pelvic incidence, pelvic tilt, sacral slope (SS), and sagittal balance (the distance between the C7 plumb line and the posterosuperior edge of S1) were measured. Kyphosis was expressed as a positive value, and lordosis as a negative value. In the frontal view, the coronal balance was measured as the distance between the C7 plumb line and the central sacral vertical line, and was expressed as positive if the C7 plumb line was shifted to the right and negative if shifted to the left. The Cobb angle in the thoracolumbar/lumbar and thoracic spine was also measured. To discriminate the direction of the coronal curve, the right convex curve was defined as negative and the left convex curve as positive. Patients with a curve larger than absolute value of 10 degrees were considered to have degenerative scoliosis.
Radiographic measurements were obtained by one of the authors based on a single measurement using standardized techniques. This author has 15 years of experience in spine surgery and was blinded to the Quest scores.
All data are presented as the mean±SD. Statistical analyses were performed using SPSS version 21. Relationships between Quest score and radiologic parameters, including Cobb angle, sagittal alignment, and trunk balance, were evaluated using Pearson or Spearman correlation coefficient analysis. Odds ratios (OR) with a 95% confidence interval were analyzed for potential risk factors of GERD using univariate and multivariate logistic regression analyses; a P-value <0.05 was considered statistically significant.
Table 1 lists the patient characteristics and mean values for measured variables. The mean absolute value of the Cobb angle in all patients was 14.4±15.4 degrees in the thoracic spine, and 23.6±20.6 degrees in the lumbar spine. Of the 126 patients with lumbar degenerative scoliosis (a curve larger than 10 degrees), 42 had a right convex curve (Cobb angle <−10 degrees) and 84 had a left convex curve (Cobb angle >10 degrees). There was no significant difference in the magnitude of the lumbar curve between the 2 groups with different curve directions. The mean thoracic Cobb angle was 14.3±10.8 degrees in patients with a right convex lumbar curve and −22.8±17.5 degrees in those with a left convex lumbar curve. The remaining 64 patients had an almost straight spine, for which the absolute value of the lumbar curve was <10 degrees. The mean Quest score of all the patients was 3.7±4.3, and 31.0% (59 patients) were defined as having GERD symptoms using 6 points as the cutoff.
Table 2 shows correlations between the Quest score and tested variables, by Pearson correlation coefficient analysis. The Quest score correlated positively with the lumbar Cobb angle (correlation coefficient=0.260) and negatively with the thoracic Cobb angle (correlation coefficient=−0.148) (P<0.05). No other variables, including sagittal parameters such as thoracolumbar kyphosis, lumbar lordosis, or sagittal balance, correlated significantly with the Quest score. Correlations within variables, tested with Spearman correlation analysis, revealed a strong association between the thoracic Cobb angle and lumbar Cobb angle (correlation coefficient=−0.842), and between the SS and lumbar lordosis (correlation coefficient=−0.761); these were the only variables with a correlation coefficient larger than 0.7 or smaller than −0.7.
The ORs for GERD risk factors were tested by univariate logistic regression analysis (Table 3). The presence of GERD symptoms (Quest ≥6) was used as a dependent variable. Of the estimated independent variables, the coronal lumbar curve had an OR of 1.021 and the thoracic curve an OR of 0.983 (P<0.05). Finally, multivariate logistic regression analysis was conducted with all variables excluding the thoracic Cobb angle and SS, which had high correlation coefficient rates with the lumbar Cobb angle and lumbar lordosis, respectively (Table 4). The lumbar curve was significantly associated with the presence of GERD, with an OR of 1.021 (P<0.05). No other variables, including sagittal parameters, showed significant correlations. As shown in Table 5, when the lumbar curve was categorized according to the Cobb angle (21 patients with Cobb <−30 degrees, large right convex curve; 126 patients with Cobb −30 to +30 degrees, relatively small curve; or 43 patients with Cobb >+30 degrees, large left convex curve), a Cobb angle >+30 degrees was strongly associated with the presence of GERD, with an OR of 10.925 (P<0.05).
Aging, being overweight, hiatal hernia, smoking, regular consumption of alcohol, and Helicobacter pylori negativity are risk factors for GERD.17,18 Kyphosis induced by a vertebral fracture or sagittal spinal malalignments is also related to the incidence or size of hiatal hernias, and to GERD.6,8,19 Miyakoshi et al5 evaluated GERD in osteoporosis patients using FSSG and assessed correlations with radiologic parameters from total spine lateral x-rays, and found that the angle of lumbar kyphosis and the number of lumbar vertebral fractures were significantly associated with GERD. Imagama and colleagues used FSSG to evaluate GERD symptoms in 245 subjects seen for health checkups and evaluated sagittal parameters with the SpinalMouse. They reported that the lumbar lordosis angle, sagittal balance, and back muscle strength are significant factors related with the presence of GERD.9 However, these studies did not investigate the relationship between coronal deformities and GERD.
Several case reports have suggested that hiatal hernia is associated with scoliosis/kyphoscoliosis.20–23 Schuchert et al24 used chest x-rays to screen for scoliosis (Cobb angle ≥10 degrees) in 320 patients who underwent repair of a giant paraesophageal hernia. Although only the coronal curve of the thoracic spine was evaluated, they found significant scoliosis in 29.1% of the patients, which is a much higher incidence than seen in the general population. Not only in adult patients, but also in adolescent scoliosis patients have been reported to be associated with hiatal hernia. Dickson and Harrington25 noted a 6.2% incidence of hiatal hernia in patients with adolescent idiopathic scoliosis before surgery; this decreased to 2.1% after scoliosis-correction surgery. This study suggests that coronal rather than sagittal deformity may be related to GERD, because kyphotic deformity is uncommon in adolescent idiopathic scoliosis patients.
Although these reports support our hypothesis that scoliosis is a risk factor for GERD, they lack the detailed skeletal measurements available from standing whole-spine x-rays. Furthermore, no previous study has taken the direction of the curve into account when evaluating scoliosis as a risk factor for GERD. Our study used measurements from whole-spine standing x-rays, both frontal and lateral views, and these data enabled us to assess spinal deformities more accurately than with the SpinalMouse or from the partial spinal views from chest x-rays. Unlike sagittal spinal parameters, which are usually expressed as a positive value in kyphosis and a negative value in lordosis, curve direction is generally not indicated in the measurements of curve magnitude in scoliosis. However, the distinction between right and left curves could be useful in evaluating the relationship between scoliosis and GERD, as the gastric-esophageal junction is located on the left side. Therefore, we defined a left convex thoracolumbar/lumbar curve as a positive value and a right convex curve as negative value to discriminate the curve direction when analyzing GERD risk factors.
We have successfully shown for the first time that scoliosis in the thoracolumbar/lumbar region is a significant risk factor for GERD symptoms. The OR was 1.021 with a 95% confidence interval of 1.010–1.033, suggesting that a left convex curve increases the risk of GERD (as we defined a left curve as a positive value). When the curve was categorized by Cobb angle, a left convex curve with substantial magnitude (Cobb >+30 degrees) was associated more strongly with GERD (OR=10.925).
The possible mechanisms by which left convex scoliosis induces GERD include anatomic changes in the diaphragm. As the diaphragm is attached to the spine at the thoracolumbar junction, a curve with an apex at this level may distort the esophageal hiatus, leading to hiatal hernia and finally to GERD. When evaluating spinal abnormalities in patients with hiatal hernia, Gavala and Zarabini26 found that 24% of the patients had scoliosis and that the apex of the scoliotic curve was at the diaphragm in 12 of 14 patients. A left thoracolumbar/lumbar curve could also increase intra-abdominal pressure by narrowing the abdominal space and pushing abdominal organs such as the stomach and gastroesophageal junction toward the abdominal wall, leading to gastric acid reflux or to hiatal hernia. This mechanism was suggested by a case reported by Yang et al,27 in which a CT scan showed that the axis between the esophagus and the stomach was deviated to the left side in a patient with severe left convex thoracolumbar neuromuscular kyphoscoliosis. They also described improvements in both GERD symptoms and axis deviation after posterior correction surgery for scoliosis for this patient. Our results, which revealed an association between GERD and a left convex curve, but not a right convex curve of equivalent magnitude, also support this explanation.
The FSSG and Quest questionnaires are widely used to evaluate GERD. For this study, we used Quest rather than FSSG because of its superiority in distinguishing GERD from other conditions.28
In contrast to previous studies, we did not find sagittal alignments such as thoracic kyphosis, thoracolumbar kyphosis, or lumbar lordosis to be significant risk factors for GERD symptoms. This may be because the sagittal parameters for patients included in this study were almost normal (mean angle of thoracic kyphosis, 23.6±15.6 degrees; thoracolumbar kyphosis, 16.2±16.4 degrees; lumbar lordosis, −26.2±20.1 degrees), as the subjects were collected with a focus on coronal curvature, not sagittal alignments. This discrepancy might also be due to differences in the questionnaires used to diagnose GERD (Quest instead of FSSG).
Our study has several limitations. First, GERD symptoms were assessed only by Quest score, without diagnostic endoscopy or other additional tests. As the sensitivity, specificity, and accuracy of Quest are reported to be 65%, 74%, and 73% with a cutoff score of 6 points,16 our study may include 25%–35% false positives or negatives. Second, our subjects were patients seen at our hospital for spinal disorders, so those patients with a Cobb angle <10 degrees were not healthy control subjects. Finally, we did not analyze the efficacy of corrective surgery or other interventions for GERD. Further study is necessary to determine whether surgical correction of scoliosis improves the Quest score, which would support the idea that scoliosis is a major cause of GERD in adult scoliosis patients.
In conclusion, our study demonstrates that scoliosis is associated with the prevalence of GERD. We should be aware that GERD may be present in patients with adult scoliosis, especially those with a left thoracolumbar/lumbar curve.
1. Aebi M. The adult scoliosis
. Eur Spine J. 2005;14:925–948.
2. Kebaish KM, Neubauer PR, Voros GD, et al. Scoliosis
in adults aged forty years and older: prevalence and relationship to age, race, and gender. Spine (Phila Pa 1976). 2011;36:731–736.
3. Tsiligiannis T, Grivas T. Pulmonary function in children with idiopathic scoliosis
4. Jackson RP, Simmons EH, Stripinis D. Coronal and sagittal plane spinal deformities correlating with back pain and pulmonary function in adult idiopathic scoliosis
. Spine (Phila Pa 1976). 1989;14:1391–1397.
5. Miyakoshi N, Kasukawa Y, Sasaki H, et al. Impact of spinal kyphosis on gastroesophageal reflux disease
symptoms in patients with osteoporosis. Osteoporos Int. 2009;20:1193–1198.
6. Kusano M, Hashizume K, Ehara Y, et al. Size of hiatus hernia correlates with severity of kyphosis, not with obesity, in elderly Japanese women. J Clin Gastroenterol. 2008;42:345–350.
7. Yamaguchi T, Sugimoto T, Yamauchi M, et al. Multiple vertebral fractures are associated with refractory reflux esophagitis in postmenopausal women. J Bone Miner Metab. 2005;23:36–40.
8. Yamaguchi T, Sugimoto T, Yamada H, et al. The presence and severity of vertebral fractures is associated with the presence of esophageal hiatal hernia in postmenopausal women. Osteoporos Int. 2002;13:331–336.
9. Imagama S, Hasegawa Y, Wakao N, et al. Influence of lumbar kyphosis and back muscle strength on the symptoms of gastroesophageal reflux disease
in middle-aged and elderly people. Eur Spine J. 2012;21:2149–2157.
10. Herbella FA, Patti MG. Gastroesophageal reflux disease
: from pathophysiology to treatment. World J Gastroenterol. 2010;16:3745–3749.
11. Dent J, El-Serag HB, Wallander MA, et al. Epidemiology of gastro-oesophageal reflux disease: a systematic review. Gut. 2005;54:710–717.
12. Kinoshita Y, Adachi K, Hongo M, et al. Systematic review of the epidemiology of gastroesophageal reflux disease
in Japan. J Gastroenterol. 2011;46:1092–1103.
13. Kusano M, Shimoyama Y, Sugimoto S, et al. Development and evaluation of FSSG: frequency scale for the symptoms of GERD. J Gastroenterol. 2004;39:888–891.
14. Carlsson R, Dent J, Bolling-Sternevald E, et al. The usefulness of a structured questionnaire in the assessment of symptomatic gastroesophageal reflux disease
. Scand J Gastroenterol. 1998;33:1023–1029.
15. Watanabe T, Urita Y, Sugimoto M, et al. Gastroesophageal reflux disease
symptoms are more common in general practice in Japan. World J Gastroenterol. 2007;13:4219–4223.
16. Nagano K, Kubo M, Goto M, et al. The diagnosis of GERD: a study of a questionnaire (QUEST) for patients complaining of upper gastrointestinal symptoms. Shinyaku to Rinsho. 1998;47:841–859.
17. Miyamoto M, Haruma K, Kuwabara M, et al. High incidence of newly-developed gastroesophageal reflux disease
in the Japanese community: a 6-year follow-up study. J Gastroenterol Hepatol. 2008;23:393–397.
18. Labenz J, Jaspersen D, Kulig M, et al. Risk factors for erosive esophagitis: a multivariate analysis based on the ProGERD study initiative. Am J Gastroenterol. 2004;99:1652–1656.
19. Yoshimura M, Nagahara A, Ohtaka K, et al. Presence of vertebral fractures is highly associated with hiatal hernia and reflux esophagitis in Japanese elderly people. Intern Med. 2008;47:1451–1455.
20. Bianchi M, Briguglio A, Fontana PF, et al. Staticdynamic modifications of the vertebral axis and the pathogenesis of diaphragmatic hernia. Chir Organi Mov. 1960;48:48–57.
21. Comte H. Esophageal hiatal hernia in the kyphoscoliotic aged. Maroc Med. 1953;32:872–874.
22. Hoeffel JC, Lascombes P, Schmitt M, et al. Peptic esophagitis and scoliosis
in children. Ann Pediatr (Paris). 1992;39:561–565.
23. Tatekawa Y, Kanehiro H, Nakajima Y. Laparoscopic modified Thal fundoplication for gastroesophageal reflux in a patient with severe scoliosis
and sliding esophageal hiatal hernia. J Pediatr Surg. 2006;41:E15–E18.
24. Schuchert MJ, Adusumilli PS, Cook CC, et al. The impact of scoliosis
among patients with giant paraesophageal hernia. J Gastrointest Surg. 2011;15:23–28.
25. Dickson JH, Harrington PR. Pre- and postoperative evaluation of scoliotic patients for hiatal hernia. South Med J. 1973;66:489–493.
26. Gavala S, Zarabini GE. The pathogenesis of diaphragmatic hernia of the hiatus in its possible relations to change of the spine. Ann Radiol Diagn (Bologna). 1961;34:481–494.
27. Yang JH, Kasat NS, Suh SW, et al. Improvement in reflux gastroesophagitis in a patient with spinal muscular atrophy after surgical correction of kyphoscoliosis: a case report. Clin Orthop Relat Res. 2011;469:3501–3505.
28. Danjo A, Yamaguchi K, Fujimoto K, et al. Comparison of endoscopic findings with symptom assessment systems (FSSG and QUEST) for gastroesophageal reflux disease
in Japanese centres. J Gastroenterol Hepatol. 2009;24:633–638.