Normal aging of the lumbar spine in women

Abstract

BACKGROUND:

Lumbar lordosis is required for bipedalism.

OBJECTIVES:

To investigate age-related changes in lumbar lordosis and to clarify the relationships between lumbar lordosis and vertebral wedging and disc degeneration.

METHODS:

A total of 300 women were included in this retrospective study, 50 in each of six age groups (20-, 30-, 40-, 50-, 60- and 70-year-olds). Patients with vertebral collapse, instable fracture or disc sequestration were excluded. In each patient, lumbar lordosis angle, posterior vertebral wedging, L5-S1 intervertebral disc angle, L5/L1 vertebral height ratio and L5-S1 intervertebral disc/L1-L2 intervertebral disc height ratio were examined. Significance level was set at $p<$ 0.05 and two-sided tests were used.

RESULTS:

Significant differences were found in lumbar lordosis according to age group ( $p<$ 0.001). Lumbar lordosis correlated most strongly with L4 posterior vertebral wedging, L5 posterior vertebral wedging and L5-S1 intervertebral disc angle, in that order ( $r=$ 0.50, $r=$ 0.40, $r=$ 0.32, respectively; $p<$ 0.001).

CONCLUSION:

In this cohort, strong spinal structure was maintained during physiological aging from 20 to 40 years of age; lumbar lordosis increased by 50 years of age. Increased lordosis correlated with increased posterior vertebral wedging and loss of posterior disc height.

Keywords

Lumbar lordosis aging vertebral wedge bipedalism

1. Introduction

Lumbar lordosis is specific to human who stands for his life. Bipedalism is only possible with sagittal balance, the most important component of which is lumbar lordosis [1]. Lumbar lordosis is poorly developed or even totally absent in infants [2]. Development of lordosis increases in conjunction with bipedalism during early childhood [3]. The most frequently encountered problems in sagittal balance involve lumbar lordosis. In their study of different spine types, Roussouly et al. [4] defined harmony between spine and pelvis and described sagittal balance as that spinal alignment requiring the least amount of energy consumption during standing. Lumbar lordosis mainly results from the conformation of the L4-L5 and L5-S1 segments. In the hypolordotic spine, the load on the spine is increased and early degeneration and disc herniation may be seen. The thoracic region bends to the front in patients with lumbar flatness deformity, resulting in sagittal imbalance [4]. In the hyperlordotic spine, the load on the posterior elements is increased, increasing the risk of facet joint hypertrophy and spondylolisthesis [4]. Some previous studies have reported that lumbar lordosis decreases in old age, whereas others have reported that lordosis increases with age [5, 6].

Vertebral height increases towards the coccyx in the vertebral column; however the posterior height of the lumbar vertebrae is shorter than the anterior height, contributing to the development of lumbar lordosis [7]. The height of the anterior vertebrae is especially greater than the posterior height at L5 and L4. Therefore, posterior wedging is seen in the lumbar vertebrae. The prevalence of physiological posterior vertebral wedging increases towards the inferior lumbar region [8]. However, in osteoporotic compression fractures, anterior wedging is often seen [9].

Previous studies have investigated age-related degeneration of the facet joints, discs and end plates [10]; others have described multifidus muscle atrophy associated with aging [11]. Our primary objective in the present study was to clarify the normal aging process by evaluating age-related changes in the lumbar lordosis angle and the prevalence of posterior vertebral wedging. Our secondary purpose was to investigate the comorbidity of increased lumbar lordosis and posterior vertebral wedging and the relationship with disc degeneration.

2. Methods

Following approval of the ethics committee at our institution, power analysis was performed to determine the number of patients to include in our retrospective study. When alpha error and beta error are both considered as 0.05, a minimum of 23 cases are needed within each group in order to state that there is a statistically significant difference for a common variance of 86 and 10 units of difference between decades, according to pilot studies.

2.1 Patient population

A retrospective study of MRIs of 300 female patients, 50from each age group, was done between January 2015 and December 2015. Only patients aged 20, 30, 40, 50, 60, or 70 years were included, to minimize the overlap of parameters during decade transitions.

Exclusion criteria:

1.
Men were excluded to eliminate the effects of gender.
2.
According to the Arbeitsgemeinschaftfür Osteosynthesefragen classification [12], patientswith (A.1.1) impaction of endplates and (A.1.2) wedge impaction fracture subgroups were included. However, those with (A.1.3) vertebral body collapse and instable fractures were excluded, because these could affect the lumbar lordosis angle.
3.
Patients with transitional vertebrae, sequestrated disc herniation, rotoscoliosis, vertebral or paravertebral mass, spinal mass, and patients with previous spinal operation or infiltrative disease of the bone marrow were excluded.

2.2 MRI methods

A 1.5-T MR Unit (Signa HDxt; GE, USA) and body surface coil were used. Sagittal T1-weighted fast spin-echo (FSE), sagittal T2-weighted FSE and axial T2-weighted FSE (3680/128 repetition time/echo time, 180 $\times$ 256 matrix, 280-mm field of view and 4-mm section thickness, NEX 2) were imaged for the study.

2.3 Visual evaluation

A radiologist with 10 years of experience in evaluating the musculoskeletal system performed all MRI evaluations. All measurements were made on T1- and T2-weighted mid-sagittal plane images. The following parameters were measured:

Lumbar lordosis angle: The angle between the line passing along the superior plateau of L1 on a mid-sagittal section and the line passing along the superior plateau of S1.

Posterior vertebral wedge: L1, L4 and L5 were individually evaluated. Anterior and posterior heights of the vertebrae, including the cortex, were measured. The ratio of anterior height/posterior height was calculated (Fig. 1a).

Figure 1.

(a) L5 posterior vertebral wedging: Anterior and posterior vertebral heights were measured including the cortex and their ratio was calculated. (b) L5-S1 intervertebral disc angle (D-Cobb): Angle between the inferior plateau of L5 and superior plateau of S1 was measured. (c) Height ratio of L5/L1 vertebra: Central height of each vertebra was measured and their ratio was calculated.

L5-S1 intervertebral disc angle (D-Cobb): The angle between lines passing along the inferior plateau of L5 and the superior plateau of S1 (Fig. 1b).

Ratio of L5/L1 vertebral height: The average height of both vertebrae was measured and their ratio was calculated (Fig. 1c).

Height ratio of L5-S1 intervertebral disc/L1-L2 intervertebral disc: The average height of both discs was measured and their ratio was calculated.

2.4 Statistical analysis

Normality was determined with Shapiro-Wilk and single sample Kolmogorov-Smirnov tests, histogram, box-plot and Q-Q graphs. Variables without normal distribution were compared with one-way Kruskal-Wallis variance analysis. Post hoc comparisons were performed with Dunn’s test. Data are expressed as mean, standard deviation, median, minimum and maximum. Variability in lumbar lordosis angle according to age category was assessed with one-way variance analysis (one-way ANOVA). Variance homogeneity was evaluated with Levene’s test. Post hoc (multiple) comparisons were performed with Tukey’s HSD test as group variances were homogenous. Correlations between lumbar lordosis angle and other variables were analyzed with Pearson’s correlation for variables with normal distribution and with Spearman’s correlation test for others. Significance level was set at $p<$ 0.05 and two-sided tests were used. Analyses were performed with NCSS 10 software.

3. Results

The lumbar lordosis angle, assessed with one-way variance analysis, ranged from 42.24 ${}^{\circ}$ to 62.82 ${}^{\circ}$ (Table 1).

Table 1
Lumbar lordosis angle was analyzed with one-way variance analysis (one-way ANOVA) (SD: standard deviation)

Age	L-Cobb Mean	SD	Min	Max
20	47.53	9.6	44.80	50.25
30	49.09	7.65	46.91	51.26
40	46.28	9.7	43.52	49.04
50	53.28	7.63	51.11	55.45
60	54.76	8.78	52.26	57.26
70	60.27	8.97	57.72	62.82

Significant differences were found in lumbar lordosis angle according to age group ( $p<$ 0.001) (Tables 2 and 3). Post hoc comparisons were made with Tukey’s HSD test. The largest difference in lumbar lordosis angle was found between 40- and 70-year-olds.

Table 2

Differences in lumbar lordosis angle between age groups ( $p<$ 0.001). Post-hoc comparisons were performed with Tukey HSD test (CI: confidence interval)

Age comparisons	L-Cobb Mean difference	Standart error (95%CI)	P
20–50	5.75	0.72–10.78	$P=$ 0.015
20–60	7.23	2.2–12.2	$P=$ 0.001
20–70	12.7	7.7–17.7	$P<$ 0.001
30–60	5.67	6.1–16.2	$P=$ 0.017
30–70	11.1	6.1–16.2	$P<$ 0.001
40–50	6.99	1.96–12.02	$P=$ 0.001
40–60	8.48	3.45–13.5	$P<$ 0.001
40–70	13.9	8.96–19.02	$P<$ 0.001
50–70	6.99	1.96–12.02	$P=$ 0.001
60–70	5.51	0.48–10.5	$P=$ 0.022

Table 3

Homogeneous distribution of lumbar lordosis according to age (Tukey’s HSD test $p<$ 0.05)

	L-Cobb Mean
Age
40	46.28
20	47.53
30	49.09	49.09
50		53.28	53.28
60			54.76
70				60.27

Table 4

Lumbar lordosis angle (L-Cobb) independent from age variable, L5-S1 disc angle (D-Cobb), L5 vertebral wedging (L5 wedge), L4 vertebral wedging (L4 wedge), L5 vertebra height/L1 vertebra height (L5/L1 vertebra) and L5-S1 disc height /L1-L2 disc height (L5/L1 disc). Mean $\pm$ SD (min-max) values are shown

	L-Cobb	D-Cobb	L5 wedge	L4 Wedge	L5/L1 vertebra	L5/L1 Disk
Mean	51.87	13.36	1.19	1.06	0.99	1.28
SD	9.93	3.5	0.97	0.07	0.06	0.37
Min	24.15	4	1	0.92	0.80	0.40
Max	81	25	1.55	1.31	1.25	2.50

Correlations between lumbar lordosis and other variables were assessed with Pearson’s correlation test. Lumbar lordosis correlated most strongly with L4 posterior vertebral wedging ( $r=$ 0.50, $p<$ 0.001). The second strongest correlation was with L5 posterior vertebral wedging ( $r=$ 0.40, $p<$ 0.001) and the third was with D-Cobb ( $r=$ 0.32, $p<$ 0.001). The weakest correlation was between lumbar lordosis and L1-L2/L5-S1 disc ratio ( $r=$ $-$ 0.14, $p=$ 0.015). Although this finding was statistically significant, the low correlation was related to sample size. There was no significant correlation between lumbar lordosis and L5/L1 vertebral height ratio ( $r=$ 0.11, $p=$ 0.054).

Spearman’s correlation test revealed no association between lumbar lordosis and L1 wedging ( $r=$ $-$ 0.003 $p=$ 0.97).

Mean $\pm$ SD (min-max) was calculated for lumbar lordosis independent from age, D-Cobb, L5 vertebral wedging, L4 vertebral wedging, L5 vertebra height/L1 vertebra height and L5-S1 disc height /L1-L2 disc height (Table 4).

The distribution of L1, L4 and L5 vertebral wedging was examined according to age group (Table 5). Post hoc comparisons were performed with the Kruskal–Wallis multiple-comparison z-value test (Table 5). Z-values $>$ 1.960 were accepted as significant.

Table 5

Distribution of L5 vertebral wedging and L4 vertebral wedging according to age (Dunn’s test)

Age	Median	Min	Max	Median	Min	Max
	L5 wedge			L4 wedge
20	1.15	1	1.45	1.03	1	1.31
30	1.14	1	1.31	1.04	0.92	1.28
40	1.17	1	1.55	1	0.93	1.16
50	1.20	1.11	1.40	1.10	0.96	1.25
60	1.16	1	1.45	1.03	0.96	1.22
70	1.30	1.08	1.45	1.14	1	1.28

There were no differences in the prevalence of L5 vertebral wedging between participants 20 years of age versus those 30 or 40 years of age (z-values 0.94 and 0.43, respectively). At 50 years of age, the prevalence of L5 vertebral wedging was significantly higher than among younger patients (z-value $>$ 1.960). At 70 years of age, this value was higher than at 20, 30, 40, or 60 years of age (z-value $>$ 1.960).

There were no differences in L4 vertebral wedging prevalence between participants 20 years of age versus those 30 or 40 years of age (z-values 0.49 and 1.23, respectively). At 50 years of age, the prevalence of L4 vertebral wedging was significantly higher than at 20, 30 or 40 years of age (z-value $>$ 1.960). At 70 years of age, the prevalence was higher than at all other ages (z-value $>$ 1.960).

For L1 vertebral wedging the only significant difference was between participants 70 years of age and those 20, 50 or 60 years of age (z-value $>$ 1.960).

Figure 2.

Examples of lumbar lordosis angle measurement on T2-weighted midsagittal MRI are shown for (a) 20-year-old, (b) 30-year-old, (c) 40-year-old, (d) 50-year-old, (e) 60-year-old and (f) 70-year-old.

Figure 3.

T2-weighted midsagittal MRI: L3-L4 and L4-L5 intervertebral disc sequestrations are seen. Lumbar lordosis is flattened (lordosis angle 31.3 ${}^{\circ}$ . Sequestrated disc was an exclusion criterion of our study; thus this patient was not included in the study.

4. Discussion

During physiological aging, significant changes of lumbar lordosis were not present in women 20 to 40 years old. Similarly, no significant difference was found in the prevalence of posterior vertebral wedging at L4 or L5 among these age groups. In addition, 40-year-olds in this study had a mean lumbar lordosis angle smaller than those of 20- and 30-year-old. Why do we see changes at 50 years of age despite a young and strong lumbar spine up to 40 years of age? Paravertebral muscles provide spine stabilization; obesity and dehydration leading to loss of disc height increase the load on the lumbar vertebrae and are considered to be important factors in spine aging. In our previous study on atrophy of the multifidus muscle, which is a significant stabilizer of the lower lumbar vertebrae, muscular atrophy began at 40 years of age [11]. The partial loss of stabilization resulting from multifidus muscle atrophy, along with obesity, may increase lordosis. There is a recognized correlation between obesity and increased lordosis [13]. In their study of middle-age and elderly women, Guo et al. [14] detected increased lumbar lordosis in obese individuals. However, obesity was not assessed in the present retrospective study. Fifty years of age was a significant cut-off point for lumbar lordosis and increased posterior vertebral wedging; both of these gradually increased after 50 years of age. L1 wedging was seen at about 70 years of age and was not observed at younger ages. Accordingly, aging in the lumbar spine, which is defined as young at 20 years of age, starts at 50 years of age. Can development of hyperlordosis be prevented by maintaining paravertebral muscle strength and preventing abdominal fat increase?

Sparrey et al. [6] studied lumbar lordosis development, sagittal balance and lumbar lordosis in aging and reported that flattened lumbar deformity occurred in elderly people because of disc degeneration and vertebral osteoporotic compression fractures. This result conflicts with ours, likely for two reasons. In our study, we excluded patients with sequestrated disc herniation and vertebral collapse. Therefore our findings did not reveal loss of lumbar lordosis; on the contrary, lordosis increased during normal aging beginning at 50 years of age. Thus, vertebral collapse and severe loss of disc height may be considered the main pathologies responsible for flattened lumbar deformity (Fig. 3). Furthermore, our study included only women and we know that lumbar lordosis is 28% more frequent in women [15, 16].

The L5-S1 segment is the main site of lumbar lordosis and the most common region for degenerative changes [17]. Thus in our study the L5-S1 intervertebral disc was analyzed in detail. At this level, lumbar lordosis correlated with D-Cobb. The disc angle is formed by posterior height loss of the disc. Increased lordosis is associated with posterior height loss of the disc. Disc dehydration seen with advancing age becomes more prominent in the posterior disc, mainly due to center of gravity. Posterior vertebral wedging and increased D-Cobb may result from similar physical effects. However, no significant correlation was found between lumbar lordosis and L5-S1 or L1-L2 disc height. Because disc height was measured at mid disc, there was less effect resulting from dehydration than at the posterior disc. Lack of significant correlation between the ratio of L5/L1 vertebral height and lumbar lordosis can be explained with a similar theory to the disc height ratio. Increased lordosis was correlated with the height of the posterior vertebra and posterior disc, whereas loss of mid-level height was independent of lordosis.

One limitation of the present study is that lumbar MRI was performed with patients in the supine position; standing X-rays are more suitable for assessment of lumbar lordosis. However, when all assessed parameters are considered, MRI is a superior diagnostic tool. A second limitation is that our study included only patients with back pain.

In conclusion, strong spinal structure is maintained in women during physiological aging from 20 to 40 years of age; lumbar lordosis increases after 50 years of age. Lumbar lordosis is most strongly correlated with increased prevalence of L4 and L5 posterior vertebral wedging and loss of posterior height of the L5-S1 intervertebral disc. Multifidus muscle atrophy, which increases after 40 years of age, and obesity may play important roles in the pathophysiological process of increased lumbar lordosis. Vertebral compression fractures and severe loss of disc height may lead to hypolordosis.

Conflict of interest

The authors declare that they have no conflict of interest.

Footnotes

Acknowledgments

Thanks to Dr. Sevim Purisa for statistical evaluation.

References

Abitbol

(1987) Evolution of the lumbosacral angle. Am J Phys Anthropol 72(3): 361–372.

Choufani

Jouve

Pomero

Adalian

Chaumoitre

Panuel

(2009) Lumbosacral lordosis in fetal spine: genetic or mechanic parameter. Eur Spine J 18(9): 1342–1348.

Cil

Yazici

Uzumcugil

Kandemir

Alanay

Acaroglu

Surat

(2005) The evolution of sagittal segmental alignment of the spine during childhood. Spine 30(1): 93–100.

Roussouly

Gollogly

Berthonnaud

Labelle

Weidenbaum

(2006) Sagittal alignment of the spine and in the presence of L5-S1 isthmic lysis and low-grade spondylolisthesis. Spine 31(21): 2484-2490.

Hegazy

(2014) Midsagittal Anatomy of Lumbar Lordosis in Adult Egyptians: MRI Study Anat Res Int 2014: 370852. doi:10.1155/2014/370852.

Sparrey

Bailey

Safaee

Clark

Lafage

Schwab

Smith

Ames

(2014) Etiology of lumbar lordosis and its pathophysiology: a review of the evolution of lumbar lordosis, and the mechanics and biology of lumbar degeneration. Neurosurg Focus 36 (5): E1. doi:10.3171/2014.1/13551.

Berry

Moran

Berg

Steffee

(1987) A morphometric study of human lumbar and selected thoracic vertebrae. Spine 12(4): 362–367.

Davies

Recker

Heaney

(1989) Normal vertebral dimensions and normal variation in serial measurements of vertebrae. J Bone Miner Res 4(3): 341–349.

Yoshimura

Kinoshita

Takijiri

Oka

Muraki

Mabuchi

, et al. (2008) Association between height loss and bone loss, cumulative incidence of vertebral fractures and future quality of life: the Miyama study. Osteoporos Int 19(1): 21–28.

10.

Bernick

Cailliet

(1982) Vertebral end-plate changes with aging of human vertebrae. Spine 7: 97–102.

11.

Ekin

Kurtul Yıldız

Mutlu

(2016) Age and sex-based distribution of lumbar multifidus muscle atrophy and coexistence of disc hernia: an MRI study of 2028 patients Diagn Interv Radiol Doi 10.5152/dir.2015.15307.

12.

Gamanagatti

Rathinam

Rangarajan

Kumar

Farooque

Sharma

(2015) Imaging evaluation of traumatic thoracolumbar spine injuries: Radiological review. World J Radiol 7(9): 253–265.

13.

Vismara

Menegoni

Zaina

Galli

Negrini

Paolo Capodaglio

(2010) Effect of obesity and low back pain on spinal mobility: a cross sectional study in women. J Neuroeng Rehabil 7: 3. doi:10.1186/1743-0003-7-3.

14.

Guo

Zhang

Alimujiang (2008) Effect of BMI and WHR on lumbar lordosis and sacrum slant angle in middle and elderly women. Zhongguo Gu Shang 21(1): 30–31.

15.

Bergenudd

Nilsson

Udén

Willner

(1989) Bone mineral content, gender, body posture, and build in relation to back pain in middle age. Spine 14: 577–579.

16.

Norton

Sahrmann

Van Dillen

(2004) Differences in measurements of lumbar curvature related to gender and low back pain. J Orthop Sports Phys Ther 34: 524–534.

17.

Filler

(2007) Emergence and optimization of upright posture among hominiform hominoids and the evolutionary pathophysiology of back pain. Neurosurg Focus 23: 1–6.