Abstract
BACKGROUND CONTEXT
Several spinal abnormalities have been studied using magnetic resonance imaging (MRI). However, in children these studies were sparsely performed in general populations. Examining young children's spines is important since the shape of the bone is largely determined during the growth spurt. Furthermore, it is so far unknown if associations between weight status and spinal abnormalities, which are known for adolescents and adults, are already present in young children.
PURPOSE
We aimed to present the prevalence of structural abnormalities in the prepubertal pediatric spine on MRI and their association with measures of the children's body weight and body composition.
STUDY DESIGN
Cross-sectional study embedded in a prospective population-based birth cohort study.
PATIENT SAMPLE
For this study, participants from the Generation R Study were selected based on the availability of MRI data of the lumbar spine and accelerometry data at the age of 9 years.
OUTCOME MEASURES
The presence of structural abnormalities of intervertebral discs and vertebrae was scored on MRI. The body mass index-standard deviation [BMI-SD] score was calculated from objectively measured weight and height, and body composition measurements were obtained by a dual-energy X-ray absorptiometry scan.
METHODS
A semiquantitative scoring tool to assess the intervertebral discs and vertebrae of the lumbar spine on conventional MRI was designed for this purpose. Proportions of children with spinal abnormalities on at least one lumbar vertebral level were presented. Logistic regression was used to analyze associations between abnormalities and weight and body composition. We declare not to have any financial conflicts of interests.
RESULTS
We included 559 children (median age of 9.88 years (interquartile range 6.74–10.02), 48.5% boys). Most frequently observed abnormalities of the intervertebral discs were abnormal signal intensity (24.9%), decreased or collapsed disc height (37.6%), disc bulging (73.3%), and abnormal nuclear shape (29.1%). Vertebral endplate irregularities and lumbosacral transitional vertebrae were seen in respectively 40% and 9.3% of the participants. Except for disc bulging, all abnormalities were predominantly present at the L5 level. Only the presence of endplate irregularities was associated with a higher body weight (BMI SD score (odds ratio [OR] 1.50 [95% confidence interval [CI] 1.21–1.86]) and BMI SD change (OR 1.48 [95% CI 1.07–2.03])) and increased body mass values in body composition measurements (% body fat (OR 1.05 [95% CI 1.02–1.09), fat mass index (OR 1.23 [95% CI 1.09–1.39]), and fat-free mass index (OR 1.30 [95% CI 1.06–1.59])) in adjusted analyses.
CONCLUSIONS
Structural spinal abnormalities, especially disc bulging, endplate irregularities, and an abnormal disc height, are already present in children aged 9 years from a Dutch population-based cohort. Of those abnormalities, endplate irregularities are associated with various weight and body composition measurements.
Keywords
Introduction
Several spinal abnormalities have been studied using magnetic resonance imaging (MRI). In a recent systematic review, we showed that disc degeneration, herniated discs, spondylolysis, and endplate changes are the most reported abnormalities among children [
[1]
]. The prevalence of disc degeneration was already 22%–44% in children and adolescents, indicating that these features are already present at a very young age [[1]
]. In adults and adolescents, some studies found associations between spinal abnormalities and low back pain, which causes a high global burden [2
, 3
, 4
]. Yet, abnormalities such as disc degeneration are also found in asymptomatic populations [[5]
,[6]
]. In children, structural spinal abnormalities have been studied sparsely in open populations. Most studies so far were performed in specific groups of athletes, in a hospital setting in specific patient populations, or in adolescents and, therefore, give little insight in the abnormalities occurring in the general pediatric population [[1]
]. The latter is remarkable, as back complaints are already present in younger children and seem to start during child's growth [[7]
,[8]
]. Examining the prepubertal spine, prior to the growth spurt, is especially important since the shape of the bone is largely determined during the growth spurt.Several factors have been suggested to be associated with structural abnormalities of the spine. In adolescents and adults, there seems to be an association between disc degeneration and the presence of obesity, that may consequently lead to pain and diminished physical and social functioning [
[9]
]. Literature indicates that higher fat mass and higher body weight can interfere with bone mass accumulation and may therefore negatively impact spinal development [[10]
]. However, it is so far unknown if associations between weight status and spinal abnormalities on MRI are already present in young children.Therefore, the aim of this study was to investigate the association between measures of body weight and composition, and structural spinal abnormalities identified on MRIs of prepubertal children.
Methods
Study population
For the present study, we selected our study population from the Generation R Study, which is a population-based prospective birth cohort study. From April 2002 to January 2006 pregnant mothers in Rotterdam, The Netherlands, were invited to participate in the study. Several questionnaires and measurements on both mother and child were performed during pregnancy and after birth. The study was approved by the Medical Ethical Committee of the Erasmus Medical Center, Rotterdam, and written informed consent was obtained from the parents of all participants [
[11]
]. For the present study we took a sample from the participating children of the mid-childhood visits at 9 years of age, of which the data collection took place from 2012 up to 2015. At this age all children were invited for two visits at the research center; the first for general measurements and the second for an MRI scan of the brain, heart, lungs, abdomen, hip and lumbar spine. In addition to the regular visits and questionnaires, a randomly selected sample of children was invited to participate in a study on sleep behaviors for which accelerometry data was collected. All children of whom MRI data of the lumbar spine of sufficient quality and accelerometry data was available, were included for the analyses of the current study purpose.Body composition measurements
At the first visit the children's weight was measured without heavy clothing and shoes using a mechanical personal scale, and height was measured in standing position using a Harpenden stadiometer (Holtain Limited, Crymych, UK). Body mass index-standard deviation (BMI SD) scores were calculated according to the Dutch reference growth curves [
[12]
]. The BMI-SD score at the age of 6 years, which was measured and calculated similarly, was used to define the change in BMI-SD score over the past 3 years. The weight status was defined using the cut-off values by Cole [[13]
], whereby the three categories for underweight were combined, resulting in four categories: underweight, normal weight, overweight, and obesity. Body fat and fat-free mass were measured using a dual-energy X-ray absorptiometry (DXA) scanner (iDXA densitometer (General Electric, formerly Lunar Corp., Madison, WI, USA)). These data were used to calculate the percentage body fat (fat mass (kg)/total body mass (kg)*100%), fat mass index (FMI, fat mass (kg)/height (m)2), and fat-free mass index (FFMI, fat-free mass (kg)/height (m)2).MRI
Structural spinal abnormalities were scored on MRI scans of the hips and pelvic region which also included the entire or major part of the lumbar spine, acquired during the second visit of the participating children at the research center. A dedicated 3.0 Tesla MRI scanner (Discovery MR750w, GE Healthcare, Milwaukee, WI, USA), operated by trained research technicians, was used. All children were scanned in supine position, according to standard imaging and positioning protocols. The MRI protocol consisted of two matched three-dimensional (3D) acquisitions; a T2-weighted fat-suppressed 3D scan and a mildly T1-weighted 2-point DIXON 3D gradient echo scan (LAVAFlex) providing two separate volumes rendering reconstructed water-only and a fat-only images. Geometrical distortions from the large FOV selected were accounted for by using a 3D gradient de-warping algorithm. A semiquantitative scoring tool to assess the lumbar spine was designed based on existing literature on spinal abnormalities [
[1]
], discussion with an experienced MSK radiologist (EO) and orthopedic surgeon (JR), and pilot testing in a test set of available MRI's. The items on the intervertebral discs and vertebrae that were included in the scoring tool, e.g. disc degeneration, endplate irregularities, and spondylolisthesis, are presented in Appendix A, and images with some examples of abnormalities are shown in Fig. B1, Fig. B2 in (Appendix B). All items were evaluated on the T2 weighted images, except from the presence of transitional vertebrae which was assessed using the water-only reconstruction from the 2-point Dixon acquisition. Two observers, MH and a research student, were trained by an experienced MSK radiologist (EO) and scored the MRI's independently. A randomly selected subsample of 20 MRI's was scored by both observers to assess the inter-observer reliability. The percentage agreement ranged between 55% and 100% for the different items (Supplementary Table 1).Covariates
Data on age and sex were available for all children. The child's ethnicity was based on the birth country of both parents and categorized in three categories: Dutch, other Western, and non-Western. Skeletal age was estimated from DXA-scans of the left hand, according to the standardized references in the Greulich and Pyle atlas [
[14]
]. Details of the used methods can be found elsewhere [[15]
].Data analysis
We used descriptive statistics to describe the baseline characteristics of the study sample. The prevalence of the abnormalities was given both as the number of participants with presence of that particular abnormality at one or more levels, and subdivided by the five lumbar vertebral levels. Cut-off points to define abnormal scores were defined based on consensus among the authors. The seven most prevalent abnormalities were further analyzed with respect to potential associations with the weight and body composition measurements. Logistic regression was used to analyze the associations between spinal abnormalities and weight and body composition, both unadjusted and adjusted for sex, ethnicity, skeletal age, and time between the two research visits. The BMI-SD change was additionally adjusted for the BMI-SD score at the age of 6 years. Results were presented as odds ratios (OR) and their corresponding 95% confidence intervals (95% CI). All analyses were complete case analyses, performed using SPSS software (IBM Corp. Released 2016. IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp.), with the level of statistical significance set at p<.05.
Results
Out of the 9,749 live born children at the beginning of the Generation R Study, 7,392 participated in the measurements at the age of 9 years, and 3,231 of them received an MRI of the spine. For 559 of those participants, accelerometry data were additionally available, and these subjects were therefore included in the present study (Fig. 1). Their median age was 9.88 years (interquartile range 9.74–10.02), 48.5% were boys, and 83.5% had a Dutch ethnicity (Table 1). The mean BMI-SD score was 0.069 (SD 0.92) and the median percentage fat mass was 24.18% (interquartile range 20.52–28.71). The included participants in the present study had more often a Dutch nationality, a lower skeletal age, and a lower BMI-SD score, lower percentage body fat and lower fat mass index compared to the excluded participants within the follow-up phase at the age of 9 years (Supplementary Table 2).
Table 1.Participant characteristics (N=559)
| Sex, boys | 271 (48.5) |
| Age, years | 9.9 (9.7–10.0) |
| Ethnicity | |
| Dutch | 467 (83.5) |
| Other Western | 34 (6.1) |
| Non-Western | 58 (10.4) |
| Skeletal age, years | 9.2 (1.2) |
| BMI, SD score | 0.07 (0.9) |
| Weight status | |
| Underweight | 40 (7.2) |
| Normal weight | 464 (83.0) |
| Overweight | 48 (8.6) |
| Obese | 7 (1.3) |
| % body fat | 24.2 (20.5-28.7) |
| FMI, kg/m2 | 3.9 (3.2-5.0) |
| FFMI, kg/m2 | 12.5 (1.0) |
Values presented as number (%) for categorical factors, or median (interquartile range) or mean (SD) for continuous factors.
This table is based on nonimputed data; missings were 2 (0.4%) for skeletal age.
BMI, body mass index; FMI, fat mass index; FFMI, fat free mass index.
Prevalence of structural spinal abnormalities
Most of the abnormalities were present at the L5–S1 intervertebral level, except for disc bulging, which was predominantly present at the upper lumbar levels (Table 2). Concerning the intervertebral discs, one or more discs with a Pfirrmann grade of 3 or higher were present in 18 participants (3.4%), while a decreased disc height as a separate item was present in 200 participants (37.6%). Disc herniation was seen in only 13 participants (2.5%), disc bulging in 385 participants (73.3%), and an abnormal nuclear shape in 154 participants (29.1%). Irregularities of the vertebral body endplates were seen in 208 (40.0%) of the participants. Only one participant had a low-degree spondylolisthesis (26%–50% anterior displacement), and lumbosacral transitional vertebrae were seen in 51 participants (9.3%), most often of type 1. Figs. 2a-f present the distribution of the most prevalent abnormalities over the lumbar vertebral levels.
Table 2.Structural abnormalities graded on MRI in 9-year-old children (N=559)
| L1–L2 | L2–L3 | L3–L4 | L4–L5 | L5–S1 | Total at any level | ||
|---|---|---|---|---|---|---|---|
| DISC DEGENERATION | |||||||
| Pfirrmann classification | |||||||
| Disc structure | Homogeneous, bright white | 231 (71.7) | 360 (68.6) | 345 (63.7) | 243 (44.8) | 137 (25.4) | 448 (84.1) |
| Inhomogeneous with or without horizontal bands | 91 (28.3) | 165 (31.4) | 197 (36.3) | 299 (55.2) | 385 (71.4) | 477 (88.2) | |
| Inhomogeneous, gray | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 16 (3.0) | 16 (3.1) | |
| Inhomogeneous, gray to black | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | |
| Inhomogeneous, black | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 1 (0.2) | 1 (0.2) | |
| Distinction nucleus/annulus | Clear | 304 (96.8) | 495 (96.7) | 515 (95.5) | 421 (78.0) | 325 (60.6) | 540 (100) |
| Unclear | 10 (3.2) | 17 (3.3) | 24 (4.5) | 119 (22.0) | 210 (39.2) | 259 (49.4) | |
| Lost | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 1 (0.2) | 1 (0.2) | |
| Signal intensity | Hyperintense, isointense to cerebrospinal fluid | 341 (98.8) | 528 (99.6) | 538 (99.1) | 523 (96.3) | 419 (77.3) | 543 (100) |
| Intermediate | 4 (1.2) | 2 (0.4) | 5 (0.9) | 19 (3.5) | 111 (20.5) | 125 (23.5) | |
| Intermediate to hypointense | 0 (0) | 0 (0) | 0 (0) | 1 (0.2) | 11 (2.0) | 12 (2.3) | |
| Hypointense | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 1 (0.2) | 1 (0.2) | |
| Disc height | Normal | 332 (96.0) | 526 (99.1) | 541 (99.6) | 487 (89.7) | 376 (69.6) | 543 (100) |
| Normal to slightly decreased | 14 (4.0) | 5 (0.9) | 2 (0.4) | 56 (10.3) | 162 (30.0) | 198 (37.2) | |
| Collapsed disc space | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 2 (0.4) | 2 (0.4) | |
| Pfirrmann grade | Grade 1 | 228 (71.0) | 358 (68.2) | 344 (63.5) | 242 (44.6) | 137 (25.4) | 447 (83.9) |
| Grade 2 | 93 (29.0) | 167 (31.8) | 197 (36.3) | 300 (55.4) | 385 (71.4) | 476 (88.0) | |
| Grade 3 | 0 (0) | 0 (0) | 1 (0.2) | 0 (0) | 10 (1.9) | 11 (2.1) | |
| Grade 4 | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 5 (0.9) | 5 (1.0) | |
| Grade 5 | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 2 (0.4) | 2 (0.4) | |
| Disc displacement | |||||||
| Disc herniation | No herniation | 309 (99.4) | 505 (98.6) | 533 (99.8) | 535 (99.8) | 530 (99.3) | 499 (97.5) |
| Protrusion | 2 (0.6) | 7 (1.4) | 1 (0.2) | 1 (0.2) | 4 (0.7) | 13 (2.5) | |
| Extrusion | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | |
| Sequestration | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | 0 (0) | |
| Disc bulging | No | 146 (49.2) | 178 (34.4) | 230 (42.8) | 455 (84.4) | 531 (98.7) | 140 (26.7) |
| Yes | 151 (50.8) | 339 (65.6) | 307 (57.2) | 84 (15.6) | 7 (1.3) | 385 (73.3) | |
| Annulus | |||||||
| High intensity zones | No | 341 (99.4) | 525 (99.4) | 538 (99.4) | 538 (99.4) | 534 (99.1) | 523 (99.1) |
| Yes | 2 (0.6) | 3 (0.6) | 3 (0.6) | 3 (0.6) | 5 (0.9) | 5 (0.9) | |
| Nucleus | |||||||
| Nuclear shape | Normal | 324 (95.9) | 513 (97.3) | 525 (97.2) | 495 (91.7) | 416 (77.2) | 376 (70.9) |
| Abnormal | 14 (4.1) | 14 (2.7) | 15 (2.8) | 45 (8.3) | 123 (22.8) | 154 (29.1) | |
| VERTEBRAL BODY | |||||||
| Endplate irregularities | No | 294 (97.7) | 469 (94.4) | 508 (95.3) | 485 (90.8) | 357 (67.5) | 312 (60.0) |
| Yes | 7 (2.3) | 28 (5.6) | 25 (4.7) | 49 (9.2) | 172 (32.5) | 208 (40.0) | |
| Schmorl nodes | No | 305 (100) | 495 (99.2) | 532 (99.4) | 536 (100) | 529 (99.4) | 491 (98.4) |
| Yes | 0 (0) | 4 (0.8) | 3 (0.6) | 0 (0) | 3 (0.6) | 8 (1.6) | |
| OTHER | |||||||
| Spondylolisthesis | 0%–25% | 555 (99.8) | |||||
| 26%–50% | 1 (0.2) | ||||||
| 51%–75% | 0 (0) | ||||||
| 76%–100% | 0 (0) | ||||||
| >100% | 0 (0) | ||||||
| Transitional vertebrae | None | 498 (90.7) | |||||
| Type I | 28 (5.1) | ||||||
| Type II | 15 (2.7) | ||||||
| Type III | 8 (1.5) | ||||||
| Type IV | 0 (0) |
Values presented as number (%). Underlined values are considered normal.
Associations between abnormalities and weight/body composition measurements
The associations between MRI abnormalities of the spine and weight and body composition measures are presented in Table 3. The presence of endplate irregularities was associated with almost all weight and body composition measures: the highest adjusted ORs were seen for the BMI-SD score (OR 1.50, 95% CI 1.21–1.86) and BMI-SD change (OR 1.48 95% CI 1.07–2.03). Furthermore, endplate irregularities were associated to both FMI (adjusted OR 1.23, 95% CI 1.09–1.39) and FFMI (adjusted OR 1.30, 95% CI 1.06–1.59). Some associations were present for the Pfirrmann grade, disc bulging, and transitional vertebrae in the crude analyses, but did not hold in the adjusted analyses. For signal intensity, disc height, and nuclear shape no associations were seen (Table 3).
Table 3.Associations between spinal MRI abnormalities and weight and body composition measurements
| Signal intensity | Disc height | Pfirrmann Grade | Disc bulging | Nuclear shape | Endplate irregularities | Transitional vertebrae | ||
|---|---|---|---|---|---|---|---|---|
| Abnormal (n=132) | Abnormal (n=200) | Abnormal (n=18) | Yes (n=385) | Abnormal (n=154) | Yes (n=208) | Yes (n=51) | ||
| BMI, SD score | Crude OR | 1.14 (0.91; 1.41) | 1.07 (0.88; 1.29) | 1.74 (1.05; 2.87)* | 0.90 (0.72; 1.11) | 1.03 (0.84; 1.27) | 1.50 (1.23; 1.83)** | 1.43 (1.05; 1.96)* |
| Adjusted OR | 1.09 (0.87; 1.38) | 1.04 (0.85; 1.28) | 1.59 (0.91; 2.76) | 0.93 (0.74 1.17) | 1.04 (0.84; 1.30) | 1.50 (1.21; 1.86)** | 1.28 (0.92; 1.78) | |
| Weight status | ||||||||
| Underweight | Crude OR | 0.47 (0.18; 1.23) | 1.40 (0.70; 2.77) | N/A # OR could not be calculated because none of the children with abnormal Pfirrmann Grade had underweight. Bold values represent statistically significant ORs (*p<.05; **p<.001).). This table is based on nonimputed data; missings were 2 (0.4%) for skeletal age, and 20 (3.6%) for BMI-SD score at age 6 years BMI, body mass index; FMI, fat mass index; FFMI, fat-free mass index. | 0.75 (0.36; 1.57) | 1.32 (0.65; 2.65) | 0.57 (0.27; 1.21) | 0.87 (0.26; 2.95) |
| Adjusted OR | 0.48 (0.18; 1.27) | 1.46 (0.73; 2.94) | N/A # OR could not be calculated because none of the children with abnormal Pfirrmann Grade had underweight. Bold values represent statistically significant ORs (*p<.05; **p<.001).). This table is based on nonimputed data; missings were 2 (0.4%) for skeletal age, and 20 (3.6%) for BMI-SD score at age 6 years BMI, body mass index; FMI, fat mass index; FFMI, fat-free mass index. | 0.71 (0.33; 1.53) | 1.30 (0.64; 2.65) | 0.60 (0.28; 1.29) | 0.94 (0.27; 3.27) | |
| Normal weight | Crude OR | Reference | Reference | Reference | Reference | Reference | Reference | Reference |
| Adjusted OR | Reference | Reference | Reference | Reference | Reference | Reference | Reference | |
| Overweight/obese | Crude OR | 0.99 (0.51; 1.93) | 1.33 (0.74; 2.38) | 1.75 (0.49; 6.28) | 0.50 (0.28; 0.91)* | 1.16 (0.62; 2.16) | 1.79 (1.00; 3.19)* | 1.51 (0.64; 3.56) |
| Adjusted OR | 0.86 (0.43; 1.70) | 1.39 (0.75; 2.56) | 1.18 (0.31; 4.50) | 0.59 (0.32; 1.10) | 1.19 (0.62; 2.27) | 1.59 (0.87; 2.92) | 1.25 (0.51; 3.06) | |
| % body fat | Crude OR | 1.02 (0.98; 1.05) | 1.01 (0.98; 1.04) | 1.08 (1.01; 1.16)* | 0.97 (0.94; 1.00) | 1.02 (0.99; 1.05) | 1.05 (1.02; 1.08)* | 1.02 (0.98; 1.07) |
| Adjusted OR | 1.00 (0.97; 1.04) | 1.02 (0.99; 1.06) | 1.07 (0.98; 1.15) | 1.00 (0.96; 1.03) | 1.02 (0.99; 1.06) | 1.05 (1.02; 1.09)* | 1.03 (0.98; 1.08) | |
| FMI, kg/m | Crude OR | 1.07 (0.95; 1.20) | 1.05 (0.94; 1.17) | 1.30 (1.05; 1.62)* | 0.89 (0.79; 0.99)* | 1.06 (0.94; 1.19) | 1.24 (1.11; 1.38)** | 1.16 (1.00-1.35) |
| Adjusted OR | 1.02 (0.89; 1.16) | 1.07 (0.95; 1.21) | 1.25 (0.96; 1.63) | 0.95 (0.84; 1.08) | 1.08 (0.95; 1.22) | 1.23 (1.09; 1.39)* | 1.14 (0.96–1.34) | |
| FFMI, kg/m | Crude OR | 1.09 (0.89; 1.32) | 1.02 (0.85; 1.22) | 1.23 (0.78; 1.94) | 0.93 (0.77; 1.13) | 0.94 (0.77; 1.14) | 1.24 (1.03; 1.48)* | 1.47 (1.11; 1.96)* |
| Adjusted OR | 1.11 (0.90; 1.38) | 0.94 (0.77; 1.15) | 1.24 (0.75; 20.4) | 0.84 (0.68; 1.04) | 0.92 (0.74; 1.14) | 1.30 (1.06; 1.59)* | 1.33 (0.97; 1.81) | |
| BMI-SD change, SD score | Crude OR | 1.17 (0.84; 1.63) | 1.12 (0.83; 1.51) | 1.61 (0.75; 3.46) | 1.08 (0.79; 1.49) | 0.94 (0.69; 1.27) | 1.33 (0.98; 1.80) | 1.19 (0.73; 1.95) |
| Adjusted OR | 1.12 (0.79; 1.59) | 1.07 (0.78; 1.45) | 1.62 (0.70; 3.76) | 1.12 (0.80; 1.57) | 0.94 (0.69; 1.30) | 1.48 (1.07; 2.03)* | 1.11 (0.67; 1.84) |
Odds Ratios (OR) and 95% confidence intervals.
1 Adjusted for sex, ethnicity, skeletal age (0.4% missing), and time between visits.
2 Additionally adjusted for BMI-SD score at age 6 years (3.6% missing).
# OR could not be calculated because none of the children with abnormal Pfirrmann Grade had underweight. Bold values represent statistically significant ORs (*p<.05; **p<.001).). This table is based on nonimputed data; missings were 2 (0.4%) for skeletal age, and 20 (3.6%) for BMI-SD score at age 6 yearsBMI, body mass index; FMI, fat mass index; FFMI, fat-free mass index.
Discussion
In this study several structural abnormalities were found to be present in large numbers in the prepubertal spine on MRI. The most observed findings were disc bulging (73.3%), endplate irregularities (40.0%), and an abnormal disc height (37.6%). Assessment of associations between spinal abnormalities and weight and body composition measurements showed that the presence of endplate irregularities was associated with a higher body weight (BMI SD score and BMI SD change) and increased body mass values in body composition measurements (% body fat, FMI, and FFMI).
The prevalence rates of spinal abnormalities found in the present study are comparable with existing literature in a general population for some specific spinal abnormalities. Kjaer et al. found comparable prevalences of abnormal signal intensity, abnormal disc height, and disc herniation in 13-year-old adolescents [
[2]
]. On the other hand, for other abnormalities our findings were different compared to the existing literature. The prevalence of 3.5% for an abnormal Pfirrmann grade for disc degeneration is lower than previous reported prevalence of 9%–57% in asymptomatic and nonathletic children aged 8–17 years [[1]
], whereas bulging and endplate irregularities had a remarkably higher prevalence than reported so far in literature (73.3% vs. 5%–46%, and 40.0% vs. 0%–21% respectively) [[1]
]. These differences might be explained by the different definitions or criteria used in the studies. Furthermore, there is large heterogeneity in study populations, with the children in the present study in general being younger than in other studied populations. Moreover, this is the first study investigating these features in a large general population sample of such a young age and indicates that especially disc bulging, endplate irregularities, and an abnormal disc height are already present at 9 years of age. The value of these findings, both on current complaints and in the future, is not clear yet and has to be further investigated.In adults, many musculoskeletal complaints are associated with weight status. For example, overweight is a well-known risk factor for osteoarthritis [
[16]
,[17]
] and low back pain [[18]
]. Literature also shows that musculoskeletal complaints are more often seen in overweight children [[19]
,[20]
]. We therefore investigated the association between structural spinal abnormalities and weight status. While an association between disc degeneration and overweight or a higher BMI has been described by several studies in adults and adolescents [[9]
,21
, 22
, 23
], we did not find this association in our pediatric population. The separate items of disc degeneration did not show any association. Crude analyses on Pfirrmann grade did show associations between BMI, percentage body fat and fat mass index and an abnormal Pfirrmann grade. However, these associations did not hold in adjusted analyses and seemed to depend on the skeletal age (abnormal Pfirrmann grade by skeletal age: OR 1.53, 95% CI 1.03–2.29). Yet, we did find endplate irregularities to be associated with higher body weight and increased body mass. This is in contrast with the findings of Li et al., who did not find an association between endplate changes and BMI in participants aged 15–55 years from a general population [[24]
]. Moreover, they suggested that endplate changes might be developmental rather than degenerative [[24]
]. Our finding of endplate irregularities, and not disc degeneration, being associated with children's body weight, might be explained by the mechanism of endplate defects being the precursor of developing disc degeneration, which did not yet occur in our very young study population [[23]
].Several underlying mechanisms may explain the association between endplate irregularities and weight status. In the present study we found that endplate irregularities were associated with both a higher fat mass index and a higher fat-free mass index. This suggests that the mass does matter and not the “content” of the mass. This supports a mechanism of mechanical loading causing damage on the vertebral endplates and is in line with Kuisma et al., who found that BMI and waist circumference, and not the percentage fat mass, contributed to the likelihood of Modic changes of the endplates in middle-aged males [
[25]
]. More research seems mandatory to better understand this mechanism and to understand future consequences of these changes at such a young age and the potential impact of weight status on this. Unfortunately, there is no data available on back complaints in the studied cohort at the age of 10 years, but ongoing measurements will give us the opportunity to investigate these complaints in adolescence. However, it might be even more interesting to understand the consequences into adulthood.Strengths and limitations
A major strength of this study is that it was part of a population based birth cohort study, in which it was possible to assess MRI data of 559 prepubertal children, in relation to weight and body composition. To our knowledge, this is the first study investigating this variety of spinal abnormalities in such a large number of children at this young age. We applied a very detailed scoring tool that also included all separate items of the Pfirrmann grade and consequently gives more insight in the specific abnormal features at this age. However, there are some limitations that need to be addressed. First, we had to take the limitations in the quality of the research MRI scans into account. Because of the relatively low quality of MRIs resulting from short scan time, it was not possible to assess spondylolysis, Modic changes, and neurocentral synchondrosis, which are potentially interesting features to examine [
[1]
,[26]
,[27]
]. As a consequence, some categories, for example, nuclear shape, were simplified into dichotomous scores. Also, the results on the distinction between the nucleus and annulus were not further analyzed, as these were not considered reliable. Furthermore, the inter-rater reliability varied largely between the scored items but this is likely explained by the low prevalence of abnormalities, resulting in a strong effect of a single difference between the raters on the percentage agreement. Finally, selection bias toward a more healthy and Western population might have occurred, as indicated by the analyses on the representativeness of the study sample for the total study cohort, which shows that more children with a Dutch ethnicity and more children with a lower weight status were included in the study sample (Supplementary Table 2). This is also known for the general follow-up of the Generation R Study [[28]
], and might have resulted in an underestimation of the prevalence of certain abnormalities, which seem to be associated with a higher weight status. Since our sample is not fully representative of the Dutch, and worldwide, population of children, caution is advised in generalizing the results.Conclusions
Structural spinal abnormalities, especially disc bulging, endplate irregularities, and an abnormal disc height, are already present in children aged 9 years from a Dutch population-based cohort. Of those abnormalities, endplate irregularities are associated with various weight and body composition measurements. Further research with longitudinal designs are needed to investigate the association between present abnormalities and low back complaints later in life, and the role of weight and body composition on the long term.
Acknowledgments
The Generation R Study is conducted by the Erasmus Medical Center in close collaboration with Faculty of Social Sciences of the Erasmus University Rotterdam, the Municipal Health Service Rotterdam area, Rotterdam, and the Stichting Trombosedienst & Artsenlaboratorium Rijnmond (STAR-MDC), Rotterdam. We gratefully acknowledge the contribution of children and parents, general practitioners, hospitals, midwives and pharmacies in Rotterdam. The general design of the Generation R Study is made possible by financial support from the Erasmus Medical Center, Rotterdam, the Erasmus University Rotterdam, ZonMw, the Netherlands Organisation for Scientific Research (NWO), and the Ministry of Health, Welfare and Sport. The work of MH was financially supported by a EUR Fellowship grant from the Erasmus University Rotterdam.
Supplementary materials
Appendix A. Scoring items with references
Tabled
1
| Categories | Remarks | |
|---|---|---|
| Disc Degeneration | ||
| Pfirrmann classification | ||
| Disc structure | Homogeneous, bright white | Examined midsagittal on a T2-weighted MR image. Categories according to Pfirrmann et al. (2001)[1] |
| Inhomogeneous with or without horizontal bands | ||
| Inhomogeneous, gray | ||
| Inhomogeneous, gray to black | ||
| Inhomogeneous, black | ||
| Distinction nucleus/annulus | Clear | Examined midsagittal on a T2-weighted MR image. Categories according to Pfirrmann et al. (2001)[1] |
| Unclear | ||
| Lost | ||
| Signal intensity | Hyperintense, isointense to cerebrospinal fluid | Examined midsagittal on a T2-weighted MR image. Categories according to Pfirrmann et al. (2001)[1] |
| Intermediate | ||
| Intermediate to hypointense | ||
| Hypointense | ||
| Disc height | Normal | Examined midsagittal on a T2-weighted MR image. Categories according to Pfirrmann et al. (2001)[1] |
| Normal to slightly decreased | ||
| Collapsed disc space | ||
| Pfirrmann grade | Grade 1 | Examined midsagittal on a T2-weighted MR image. Categories according to Pfirrmann et al. (2001)[1] |
| Grade 2 | ||
| Grade 3 | ||
| Grade 4 | ||
| Grade 5 | ||
| Disc displacement | ||
| Disc herniation | No herniation | First evaluated on midsagittal T2 images, if there is no herniation/bulging into the spinal canal: 0. If there is herniation/bulging: further examination on the axial images. Definitions according to Fardon et al. (2014)[2] |
| Protrusion | ||
| Extrusion | ||
| Sequestration | ||
| Disc bulging | No | Bulging into the spinal canal evaluated on midsagittal T2 images. Definition according to Fardon et al. (2014)[2] |
| Yes | ||
| Annulus | ||
| High intensity zones | No | Examined on midsagittal T2-weighted images, according to Aprill et al. (1992)[3] |
| Yes | ||
| Nucleus | ||
| Nuclear shape | Normal | Graded on transverse T2-weighted images. Grading from Raininko et al. (1995)[4] , simplified into two categories: “normal” for a symmetric and regular shaped nucleus, or “abnormal” for an asymmetric or irregular shaped nucleus. |
| Abnormal | ||
| VERTEBRAL BODY | ||
| Endplate irregularities | No | Any irregularities in the endplates around the graded disc (e.g., L4–L5 involves the caudal endplate of the L4 vertebra and the cranial endplate of the L5 vertebra), assessed on sagittal T2-weighted images. |
| Yes | ||
| Schmorl nodes | No | Schmorl nodes according to the definition of Fardon et al. (2014)[2] in the endplates around the graded disc (e.g., L4–L5 involves the caudal endplate of the L4 vertebra and the cranial endplate of the L5 vertebra), assessed on sagittal T2-weighted images. |
| Yes | ||
| OTHER | ||
| Spondylolisthesis | 0%–25% | Graded according to the Meyerding classification[5] by use of sagittal T2-weighted images, and if needed additionally 2-point Dixon water-only images. |
| 26%–50% | ||
| 51%–75% | ||
| 76%–100% | ||
| >100% | ||
| Transitional vertebrae | None | Presence, type, and bilaterality of lumbosacral transitional vertebrae graded according to the Castellvi classification[6] by use of coronal 2-point Dixon water-only images. |
| Type I | ||
| Type II | ||
| Type III | ||
| Type IV |
References
- 1.Pfirrmann CW, Metzdorf A, Zanetti M, Hodler J, Boos N. Magnetic resonance classification of lumbar intervertebral disc degeneration. Spine (Phila Pa 1976). 2001;26(17):1873-8.
- 2.Fardon DF, Williams AL, Dohring EJ, Murtagh FR, Gabriel Rothman SL, Sze GK. Lumbar disc nomenclature: version 2.0: recommendations of the combined task forces of the North American Spine Society, the American Society of Spine Radiology, and the American Society of Neuroradiology. Spine (Phila Pa 1976). 2014;39(24):E1448-65.
- 3.Aprill C, Bogduk N. High-intensity zone: a diagnostic sign of painful lumbar disc on magnetic resonance imaging. Br J Radiol. 1992;65(773):361-9.
- 4.Raininko R, Manninen H, Battie MC, Gibbons LE, Gill K, Fisher LD. Observer variability in the assessment of disc degeneration on magnetic resonance images of the lumbar and thoracic spine. Spine (Phila Pa 1976). 1995;20(9):1029-35.
- 5.El-Feky M, Gaillard F, al. e. Spondylolisthesis grading system [Available from: https://radiopaedia.org/articles/spondylolisthesis-grading-system.
- 6.Castellvi AE, Goldstein LA, Chan DP. Lumbosacral transitional vertebrae and their relationship with lumbar extradural defects. Spine (Phila Pa 1976). 1984;9(5):493-5.
Appendix B – Examples of MRIs with common abnormalities
Fig. 1Flow diagram for the study sample selection.
Fig. 2(A) The distribution of abnormal signal intensity over the lumbar vertebral levels. (B) The distribution of abnormal disc height over the lumbar vertebral levels. (C) The distribution of abnormal Pfirrmann grade over the lumbar vertebral levels. (D) The distribution of disc bulging over the lumbar vertebral levels. (E) The distribution of abnormal nuclear shape over the lumbar vertebral levels. (F) The distribution of endplate irregularities over the lumbar vertebral levels.
Fig. B1T2-weighted image of a participant with a decreased disc height at L4–L5 (A), disc bulging at L2–L3 and L3–L4 (B), and endplate irregularities at L5–S1 (C).
Fig. B2T2-weighted image of a participant with abnormal signal intensity at L4–L5 (A), disc bulging at L2–L3 and L3–L4 (B), and endplate irregularities at L5–S1 (C).
References
- The prevalence of abnormalities in the pediatric spine on MRI: a systematic review and meta-analysis.Spine. 2020; 45: E1185-E1196
- An epidemiologic study of MRI and low back pain in 13-year-old children.Spine. 2005; 30: 798-806
- MRI findings of disc degeneration are more prevalent in adults with low back pain than in asymptomatic controls: a systematic review and meta-analysis.AJNR Am J Neuroradiol. 2015; 36: 2394-2399
- The global burden of low back pain: estimates from the Global Burden of Disease 2010 study.Ann Rheum Dis. 2014; 73: 968-974
- Systematic literature review of imaging features of spinal degeneration in asymptomatic populations.AJNR Am J Neuroradiol. 2015; 36: 811-816
- Back pain and MRI changes in the thoraco-lumbar spine of young elite Mogul skiers.Scand J Med Sci Sports. 2017; 27: 983-989
- Viz hub.Institute for Health Metrics and Evaluation, Seattle, USA2019 ([internet][cited 2019 September 9]Available from:)
- Prevalentie van nek- en rugklachten naar leeftijd en geslacht in de huisartsenpraktijk.Rijksinstituut voor Volksgezondheid en Milieu (RIVM). 2018; ([cited 2019 September 9]Available from:)
- A population-based study of juvenile disc degeneration and its association with overweight and obesity, low back pain, and diminished functional status.J Bone Joint Surg Am. 2011; 93: 662-670
- Obesity is a risk factor for fracture in children but is protective against fracture in adults: a paradox.Bone. 2012; 50: 457-466
- The Generation R Study: design and cohort update 2017.Eur J Epidemiol. 2016; 31: 1243-1264
- Continuing positive secular growth change in The Netherlands 1955-1997.Pediatr Res. 2000; 47: 316-323
- Extended international (IOTF) body mass index cut-offs for thinness, overweight and obesity.Pediatr Obes. 2012; 7: 284-294
- Radiographic atlas of skeletal development of the hand and wrist.2nd ed. Stanford University Press, Palo Alto, United States1959
- Skeletal maturation in relation to ethnic background in children of school age: the generation R study.Bone. 2020; 132115180
- Osteoarthritis, obesity and weight loss: evidence, hypotheses and horizons - a scoping review.Obes Rev. 2014; 15: 578-586
- Weight and osteoarthritis.Am J Clin Nutr. 1996; 63: 430S-432S
- The association between obesity and low back pain: a meta-analysis.Am J Epidemiol. 2010; 171: 135-154
- Overweight and obesity are associated with musculoskeletal complaints as early as childhood: a systematic review.Obes Rev. 2014; 15: 52-67
- Orthopedic complications of overweight in children and adolescents.Pediatrics. 2006; 117: 2167-2174
- Association of abdominal obesity with lumbar disc degeneration–a magnetic resonance imaging study.PLoS One. 2013; 8: e56244
- The association of lumbar intervertebral disc degeneration on magnetic resonance imaging with body mass index in overweight and obese adults: a population-based study.Arthritis Rheum. 2012; 64: 1488-1496
- Vertebral endplate defect as initiating factor in intervertebral disc degeneration: strong association between endplate defect and disc degeneration in the general population.Spine (Phila Pa 1976). 2018; 43: 412-419
- Two subtypes of intervertebral disc degeneration distinguished by large-scale population-based study.Spine J. 2016; 16: 1079-1089
- Are the determinants of vertebral endplate changes and severe disc degeneration in the lumbar spine the same? A magnetic resonance imaging study in middle-aged male workers.BMC Musculoskelet Disord. 2008; 9: 51
- Degenerative disk disease: assessment of changes in vertebral body marrow with MR imaging.Radiology. 1988; 166: 193-199
- Morphometric analysis of neurocentral synchondrosis using magnetic resonance imaging in the normal skeletally immature spine.Spine (Phila Pa 1976). 2010; 35: 76-82
- The generation R study: design and cohort update 2012.Eur J Epidemiol. 2012; 27: 739-756
Article info
Publication history
Published online: October 09, 2020
Accepted:
October 2,
2020
Received in revised form:
October 1,
2020
Received:
July 9,
2020
Footnotes
FDA device/drug status: Not applicable.
Author disclosures: MMvdH: Nothing to disclose. EHGO: Nothing to disclose. JJMR: Nothing to disclose. SMAB-Z: Consulting: Pfizer (A); Trips/Travel Reimbursement: OARSI congress (A per year); Board of Director: Associate editor Osteoarthritis & Cartilage (B per year); Grants: Several grants from The Netherlands Organization for Health Research and Development, Dutch Arthritis Association (I over last 10 years). MvM: Grant: EUR Fellowship.
Identification
Copyright
© 2020 The Author(s). Published by Elsevier Inc.
User license
Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0) | How you can reuse 
Elsevier's open access license policy
Creative Commons Attribution – NonCommercial – NoDerivs (CC BY-NC-ND 4.0)
Permitted
For non-commercial purposes:
- Read, print & download
- Redistribute or republish the final article
- Text & data mine
- Translate the article (private use only, not for distribution)
- Reuse portions or extracts from the article in other works
Not Permitted
- Sell or re-use for commercial purposes
- Distribute translations or adaptations of the article
Elsevier's open access license policy
31140-2&id=gr1.jpg){kind=link}
31140-2&id=gr2.jpg){kind=link}
31140-2&id=gr3.jpg){kind=link}
31140-2&id=gr4.jpg){kind=link}