Osteoporosis International

, Volume 24, Issue 8, pp 2269–2273 | Cite as

Relationship of thoracic kyphosis and lumbar lordosis to bone mineral density in women

  • A. Pavlovic
  • D. L. Nichols
  • C. F. Sanborn
  • N. M. DiMarco
Original Article



The relationship between spinal curvature and bone mineral density (BMD) in women was examined. Significant relationships were observed between spinal curvature and BMD in both pre- and postmenopausal women. Excessive spinal curvature may be associated with low bone mass in premenopausal women.


The purpose of this study was to examine the associations between spinal measurements of thoracic and lumbar curvatures and bone mineral density in pre- and postmenopausal women.


The data for this study were obtained from the Texas Woman’s University Pioneer Project. Female participants (n = 242; premenopausal n = 104, postmenopausal n = 138) between the ages of 18 and 60 years were evaluated on multiple health measures. Thoracic and lumbar curvatures were measured with a 24-in. (60 cm) flexicurve. Bone mineral density was assessed via dual-energy X-ray absorptiometry (Lunar DPX IQ, version 4.6e). Pearson correlations and logistic regression analysis were used to examine the associations between the obtained spinal curvature measurements and bone mineral density. Significance was set at p < .05.


Significant correlations were observed for the femoral neck and lumbar spine bone mineral density with thoracic and lumbar curve in premenopausal women (r = −.344 to − .525; p < .001). Slightly weaker, but significant, correlations were observed for femoral neck and lumbar spine in relation to thoracic and lumbar curve in postmenopausal women (r = −.288 to −.397; p < .01). Premenopausal women with thoracic curvature greater than 4 cm had a greater risk of having low bone mass compared to premenopausal women with less than 4 cm of curvature (odds ratio = 3.982, 95 % CI = 1.206, 13.144).


The observed negative relationship suggests that as either thoracic or lumbar curvature increases, the regional bone mineral density decreases in both pre- and postmenopausal women.


BMD Low bone mass Spinal curvature 


Osteoporosis is a disease of the skeletal system that is characterized by low bone mass. Decreased bone mass is associated with increased risk of bone fractures, especially in the wrist, hip, and spine. Although osteoporosis is quite common, certain populations are at greater risk for developing the disease. Individuals at increased risk for developing osteoporosis are females, older adults, Caucasians, and postmenopausal women. Other risk factors include small body size, consuming diets that are deficient in calcium, and being physically inactive [1].

Slight spinal curvature is a normal component of the spine’s structure. However, in excess, spinal curvature can be considered a deformity. Two types of spinal deformities include kyphosis and lordosis. Kyphosis, commonly referred to as “hunchback” or “dowager’s hump,” occurs when there is excessive outward curvature of the thoracic spine. Lordosis, also known as “swayback,” is characterized by excessive inward curvature of the lumbar spine. In individuals with excessive spinal curvature, direct vertebral loading of the skeleton may be compromised which may ultimately result in decreased bone mineral density (BMD). In elderly men and women, a significant negative relationship has been found between thoracic curvature and thoracic BMD (r = −0.56; p < .05) and lumbar BMD (r = −0.57; p < .05; [2]). Similarly, Ettinger et al. [3] found that kyphosis was inversely related to BMD in postmenopausal women (ages: 65–91 year). Significant negative correlations were found in all four BMD sites (distal radius, calcaneus, lumbar spine, and femoral neck). Furthermore, the 10 % of women with the most severe kyphosis had approximately 7–17 % lower BMD at the four measured sites as compared to the rest of the sample.

However, consistent relationships between spinal curvature and BMD are not always found. Sinaki and associates [4] examined the correlation between thoracic kyphosis and lumbar lordosis and BMD (L2–L4) in 65 estrogen-deficient women who were between the ages of 48 and 65 years. Bone mineral density was not significantly related to either thoracic kyphosis or lumbar lordosis. Although findings vary, spinal curvature and BMD seem to be moderately, indirectly related in postmenopausal women. To date, the relationship has not been examined in premenopausal women. Therefore, the purpose of this study was to examine the relationships between spinal measurements of thoracic kyphosis and lumbar lordosis and BMD in pre- and postmenopausal women. The hypothesis was that greater curvature in the spine would be associated with lower bone mass in both pre- and postmenopausal women.


Participants for this study were women who volunteered to be a part of the Pioneer Project, a longitudinal research study devoted to examining women’s health across the lifespan. Generally healthy women who ranged in age from 18 to 60 years with no preexisting conditions or diseases were enrolled. Exclusion criteria included resting blood pressure >200 mmHg systolic or 115 mmHg diastolic; weight over 275 lb; pregnant or attempting to become pregnant; within 6 months postpartum; inability to stand freely; or any indication of cardiovascular disease. This study was approved by the Texas Woman’s University Institutional Review Board, and all women gave informed consent before participating in the study.

Baseline data from 104 premenopausal and 138 postmenopausal women consisted of the following: height, weight, vital signs, waist and hip circumference, pulmonary function, blood and urine sampling, bone density measurements, cardiovascular stress test, assessments of spinal curvature, muscle strength and coordination, TestWell® Wellness Assessment, the Harvard Food Frequency Questionnaire, and a medical and reproductive history questionnaire. Details on data collected that were used for the current study are presented below. Menopausal status was determined based on a series of questions related to absence of menstrual cycles and time since the last menstrual cycle. Women older than 40 years who had not had a menstrual cycle for more than 12 months were classified as postmenopausal.

Bone mineral density

Measurements for determination of bone mineral density of lumbar spine (L2–L4), total body, and right proximal femur (femoral neck, Ward’s area, and trochanter) were conducted using a Lunar DPX IQ dual-energy X-ray absorptiometer (DXA, software version 4.6b, Lunar Radiation Corporation, Madison, WI). Lean tissue mass and fat mass were determined from total body measurements performed with DXA. To avoid possible problems with changes in size or position of regions of interest, the same certified technician performed all analyses using the same technique for all measurements. Scan speed was kept constant throughout the study for each participant and was determined based on the height and weight of the participant.

Short-term precision error of DXA measurements in the Texas Woman’s University laboratory was determined by performing three repeated measurements on 14 adults, according to the method outlined by Glüer et al. [5]. Precision error was 1.2 % or less for all bone density measurements. Quality assurance and control measurements were done as recommended by the International Society of Clinical Densitometry [6].

Spinal curvature

A flexicurve was used to measure spinal curvature [7, 8]. The flexicurve is a strip of malleable metal 24 in. (60 cm) in length. With the participant standing with normal posture, the flexicurve was manipulated to match the curvature of the spine. The proximal end of the flexicurve was placed on the seventh cervical vertebrae, and the level of the lumbosacral joint was marked on the flexicurve. The flexicurve was then placed on a large piece of graph paper and the spinal curve was transferred to the paper by running a pencil along the flexicurve. The ends of the curve were joined with a straight line, and the amount of deviation (in centimeter) from the straight line was used as the measure of both lumbar and thoracic curvature.

Statistical analysis

Descriptive statistics (means and standard deviations) were computed for all relevant variables. To examine the relationship between spinal curvature and BMD, zero-order Pearson correlations were determined for both thoracic and lumbar curvature and lumbar and femoral neck BMD. Because of the potential influence of age on BMD, partial correlations controlling for age were also computed. The magnitude of the correlation coefficients were compared between pre- and postmenopausal women using Fisher Z-transformations.

Quartiles of thoracic curvature were determined based on determining the 25th, 50th, and 75th percentile of the sample. Approximate levels of curvature (0–2, 2–3, 3–4, and >4 cm) for each of the four quartiles was then determined. Differences in lumbar BMD across the four levels of curvature (0–2, 2–3, 3–4, and >4 cm) in pre- and postmenopausal women were compared with Factorial ANOVA (menopausal status × 4 levels of curvature). Sidak post hoc tests were used when appropriate.

Women were also classified as having low bone mass, based on a T-score at the femoral neck of −1.0 or less. Logistic regression was used to calculate the risk of having low bone mass based on two levels of thoracic curvature (less than or greater than 4 cm, the 75th percentile); age, height, and weight were used as covariates in the analysis. Level of significance was set at .05 for all statistical analyses.


Descriptive characteristics of participants are presented in Table 1. The age range for the premenopausal women was 19–53 years, while for postmenopausal women, the range was 41–60 years. Mean values of both thoracic and lumbar curvature were similar between pre- and postmenopausal women. Ranges for lumbar curvature were 0–15 and 0–19 cm for pre- and postmenopausal women, respectively. Thoracic curvature ranges were 0–13 and 0–11 cm, respectively. Data for both spinal curvatures were positively skewed and thus there were a few outliers (Z-score > 3.0; n = 8), which were removed prior to further statistical analysis.
Table 1

Descriptive characteristics of participants


Premenopausal (n = 104)

Postmenopausal (n = 138)

Age (year)

39.5 ± 6.7

52.0 ± 4.6

Weight (kg)

64.4 ± 11.2

67.6 ± 13.6

Height (cm)

164.6 ± 6.8

164.4 ± 5.6

Femoral Neck BMD (g/cm2)

0.958 ± .119

0.909 ± .138

Lumbar Spine BMD (g/cm2)

1.157 ± .124

1.149 ± .168

Kyphosis (cm)

3.2 ± 2.2

3.4 ± 2.2

Lordosis (cm)

2.7 ± 3.4

3.2 ± 4.1

Low BMDa (%)



aDefined as a T-score at the femoral neck of less than −1.0

Using Pearson correlations, significant negative relationships (r = −0.29 to −0.53) were found between both spinal measurements (thoracic and lumbar curvatures) and BMD (femoral neck and lumbar spine) in both pre- and postmenopausal women (Table 2, Fig. 1). These relationships remained significant after controlling for age (Table 2). Although the relationships appear to be stronger in the premenopausal group, no significant differences in the magnitude of the relationship between pre- and postmenopausal women were found based on Fisher Z-transformations (z < 1.96).
Table 2

Zero-order and first-order correlations for BMD and spinal curvature

BMD Site

Premenopausal (n = 100)

Postmenopausal (n = 134)

Thoracic curvature

Lumbar curvature

Thoracic curvature

Lumbar curvature


Zero-order correlations

Femoral Neck





Lumbar Spine (L1–L4)






First-order correlations—controlling for age & weight

Femoral Neck





Lumbar Spine (L1–L4)





*p < 0.01 (statistically significant)

Fig. 1

Relationship between lumbar spine BMD and thoracic curvature in premenopausal and postmenopausal women

Regardless of menopausal status, a significantly lower lumbar BMD was found in women who had a spinal curvature >4 cm compared to those in the three lower quartiles (Fig. 2). No significant differences were observed between the first three quartiles. There was also no significant interaction of menopausal status and spinal curvature.
Fig. 2

Lumbar spine BMD in premenopausal and postmenopausal women across the four quartiles of curvature. *Lumbar BMD was significantly lower in women with more than 4 cm of curvature when compared to the other three quartiles

In premenopausal women with greater than 4 cm of thoracic curvature, there was a 4-fold greater risk of having low bone mass compared to premenopausal women with less than 4 cm of curvature (odds ratio = 3.982, 95 % CI = 1.206, 13.144). Using the covariates of age, height, and weight lowered the odds ratio slightly to 3.605, but it was still significant (95 % CI = 1.020, 12.739). The association between low bone mass and thoracic curvature was also significant in postmenopausal women (odds ratio = 2.652, 95 % CI = 1.108, 6.348) but not after adjustment for age, height, and weight (odds ratio = 1.964, 95 % CI = 0.752, 5.132)


The present study found significant negative relationships (r = −0.26 to −0.53) between both measures of spinal curvature (thoracic and lumbar) and BMD (femoral neck and lumbar spine) in both pre- and postmenopausal women. Our findings are similar to that of Edmondston and colleagues [2], who also found thoracic curvature significantly correlated with lumbar BMD (r = −0.57). One possible explanation for the observed negative relationship is the reduction in direct skeletal loading of the spine in individuals with excessive spinal curvature. Another explanation may be low back extensor strength which may result in reduced muscular pull on the bone. However, Sinaki and associates [4] failed to find a significant relationship between BMD (L2–L4) and spinal curvature in women between the ages of 48 and 65 years. They found that back extensor strength was a strong predictor of both kyphosis and lordosis, and, therefore, a high degree of curvature was most likely attributed to low back extensor strength.

When women in the present study were placed into quartiles based on spinal curvature, participants with the most amount of curvature (>4 cm) had significantly lower lumbar spine BMD as compared to all other quartiles (Fig. 2). This was true for both pre- and postmenopausal women. Similarly, Ettinger et al. [3], found that women with the most severe kyphosis had approximately 7–17 % lower BMD at various sites (spine, calcaneus, proximal radius, and distal radius). Based on these findings, individuals with excessive spinal curvature should be encouraged to monitor their bone health regardless of age.

The other unique finding of the present study was that for premenopausal women, thoracic curvature greater than 4 cm was associated with a 4-fold greater risk of having low bone mass (T-score < −1.0) when compared to premenopausal women with curvature less than 4 cm. This finding is similar to other studies in postmenopausal women in which kyphosis has been associated with a greater risk for both low bone mass and fracture [9, 10]. To our knowledge, the present study is the first to observe such a relationship in premenopausal women. However, some caution should be observed when interpreting the association in the present study as only 20 of the premenopausal women had low bone mass.

The incidence of thoracic kyphosis and lumbar lordosis may be reduced through exercise, specifically strengthening of the back extensors. A study by Mika, Unnithan, and Mika [11] examined differences in thoracic kyphosis and back extensor strength in women with osteoporosis. They found that the women with the weakest back extensor strength (<35 N) had significantly higher mean values of thoracic kyphosis as compared to women with back extensor strength >60 N. These findings suggest the importance of maintaining adequate amounts of muscular strength, which may ultimately prevent spinal deformities that lead to bone loss.

Although the present study only focuses on thoracic and lumbar curvature, scoliosis, abnormal lateral curvature of the spine, has also been associated with low BMD. Adolescent girls diagnosed with scoliosis were found to have significantly lower areal and volumetric BMD compared to their age-matched controls. However, a relationship was not observed between severity of scoliosis and BMD [12]. A similar relationship was found in postmenopausal women with scoliosis [13] who had significantly lower total hip and femoral neck BMD compared to their counterparts without scoliosis.

Three potential limitations are noted for this study. First is assessment of spinal curvature utilizing a flexicurve. Although studies have validated the flexicurve for measuring spinal curvature [7, 8], other measurement techniques are considered more accurate. A second limitation was that menstrual status was classified only as pre- or postmenopausal, and we therefore did not have a control for perimenopausal status. Perimenopausal women typically have declining estrogen levels, and bone loss may already be occurring in these women. Future studies should examine pre-, peri-, and postmenopausal relationships to spinal curvature and BMD.

The final limitation is that the design of this study was cross-sectional, and therefore the progression or timing of spinal curvature is unknown. Thus it is possible that the low bone mass predates the spinal curvature. Longitudinal studies should follow premenopausal women with spinal curvature overtime (especially if the curvature exceeds 4–6 cm) to determine how much bone loss, if any, may be occurring, and if that bone loss can be attributed to the spinal curvature.

In conclusion, we found that spinal curvature is negatively correlated with BMD in both pre- and postmenopausal women. Regardless of age or menopausal status, women with increased amounts of thoracic or lumbar curvature had significantly lower lumbar spine BMD, and those with thoracic curvature greater than 4 cm had a significantly greater risk of having low bone mass at the femoral neck. Although, previous studies have examined the relationship between spinal curvature and BMD, to our knowledge, this is the first study that examined the relationship in premenopausal women. We have shown that the negative relationship exists in all women, and that premenopausal women should not be overlooked simply on the basis of their menstrual status or age.


Conflicts of interest



  1. 1.
    U.S. Preventive Services Task Force (2011) Screening for osteoporosis: U.S. preventive services task force recommendation statement. Ann Intern Med 154:356–364. doi:10.1059/0003-4819-154-5-201103010-00307 CrossRefGoogle Scholar
  2. 2.
    Edmondston SJ, Singer KP, Price RI, Day RE, Breidahl PD (1994) The relationship between bone mineral density, vertebral body shape and spinal curvature in the elderly thoracolumbar spine: an in vitro study. Br J Radiol 67:969–975PubMedCrossRefGoogle Scholar
  3. 3.
    Ettinger B, Black DM, Palermo L, Nevitt MC, Melnikoff S, Cummings SR (1994) Kyphosis in older women and its relation to back pain, disability and osteopenia: the study of osteoporotic fractures. Osteoporos Int 4:55–60PubMedCrossRefGoogle Scholar
  4. 4.
    Sinaki M, Itoi E, Rogers JW, Bergstralh EJ, Wahner HW (1996) Correlation of back extensor strength with thoracic kyphosis and lumbar lordosis in estrogen-deficient women. Am J Phys Med Rehabil 75:370–374PubMedCrossRefGoogle Scholar
  5. 5.
    Gluer CC, Blake G, Lu Y, Blunt BA, Jergas M, Genant HK (1995) Accurate assessment of precision errors: how to measure the reproducibility of bone densitometry techniques. Osteoporos Int 5:262–270PubMedCrossRefGoogle Scholar
  6. 6.
    Baim S, Binkley N, Bilezikian JP, Kendler DL, Hans DB, Lewiecki EM, Silverman S (2008) Official Positions of the International Society for Clinical Densitometry and executive summary of the 2007 ISCD Position Development Conference. J Clin Densitom 11:75–91. doi:10.1016/j.jocd.2007.12.007 PubMedCrossRefGoogle Scholar
  7. 7.
    de Oliveira TS, Candotti CT, La Torre M, Pelinson PP, Furlanetto TS, Kutchak FM, Loss JF (2012) Validity and reproducibility of the measurements obtained using the flexicurve instrument to evaluate the angles of thoracic and lumbar curvatures of the spine in the sagittal plane. Rehabil Res Pract 2012:186156. doi:10.1155/2012/186156 PubMedGoogle Scholar
  8. 8.
    Greendale GA, Nili NS, Huang MH, Seeger L, Karlamangla AS (2011) The reliability and validity of three non-radiological measures of thoracic kyphosis and their relations to the standing radiological Cobb angle. Osteoporos Int 22:1897–1905. doi:10.1007/s00198-010-1422-z PubMedCrossRefGoogle Scholar
  9. 9.
    Cunha-Henriques S, Costa-Paiva L, Pinto-Neto AM, Fonsechi-Carvesan G, Nanni L, Morais SS (2011) Postmenopausal women with osteoporosis and musculoskeletal status: a comparative cross-sectional study. J Clin Med Res 3:168–176. doi:10.4021/jocmr537w PubMedGoogle Scholar
  10. 10.
    Huang MH, Barrett-Connor E, Greendale GA, Kado DM (2006) Hyperkyphotic posture and risk of future osteoporotic fractures: the Rancho Bernardo study. J Bone Miner Res 21:419–423. doi:10.1359/JBMR.051201 PubMedCrossRefGoogle Scholar
  11. 11.
    Mika A, Unnithan VB, Mika P (2005) Differences in thoracic kyphosis and in back muscle strength in women with bone loss due to osteoporosis. Spine (Phila Pa 1976) 30:241–246CrossRefGoogle Scholar
  12. 12.
    Cheng JC, Qin L, Cheung CS, Sher AH, Lee KM, Ng SW, Guo X (2000) Generalized low areal and volumetric bone mineral density in adolescent idiopathic scoliosis. J Bone Miner Res 15:1587–1595. doi:10.1359/jbmr.2000.15.8.1587 PubMedCrossRefGoogle Scholar
  13. 13.
    Jaovisidha S, Kim JK, Sartoris DJ, Bosch E, Edelstein S, Barrett-Connor E, Rojanaplakorn P (1998) Scoliosis in elderly and age-related bone loss: a population-based study. J Clin Densitom 1:227–233PubMedCrossRefGoogle Scholar

Copyright information

© International Osteoporosis Foundation and National Osteoporosis Foundation 2013

Authors and Affiliations

  • A. Pavlovic
    • 1
  • D. L. Nichols
    • 1
  • C. F. Sanborn
    • 1
  • N. M. DiMarco
    • 1
  1. 1.Department of Kinesiology and Institute for Women’s HealthTexas Woman’s UniversityDentonUSA

Personalised recommendations