Results
Interobserver agreement (weighted k) varied between 0.83 and 0.25 (n=190),
according to the variable,
disc level and pair of observers.
It was good (0.60-0.79) or moderate (0.40-0.59) in most cases.
Lowest values were found at L2/L3 level where most discs had mild degenerative changes only.
Signal intensity of the disc was lower,
size of the nuclear complex larger and intranuclear cleft thicker and more often irregular among middle aged than young men.
Irregularity and inhomogeneity of the annulus fibrosus was relatively common among both age groups,
especially at lower levels.
Inhomogeneity of nucleus pulposus was common at all levels among middle-aged men but the young men only had inhomogeneity at lower levels.
Bony endplate lesions were not common,
most found at lower levels and equally common among both age groups.
Most subchondral bone marrow changes were detected at lower levels and among the middle-aged men.
The most common morphological disc types,
or combinations of grades of the chosen degenerative MRI variables,
were different among young and middle-aged men and varied between disc levels (Figure 1,
Tables 2.
and 3.).
Young men had more often combinations with normal or grey signal intensity of nucleus and a small,
regular nuclear complex.
Bulges and inhomogeneity of annulus or nucleus were only detected at lower disc levels and most severe degeneration at L5/S1.
Middle aged men more often had a large nuclear complex and thick intranuclear cleft at any level.
Irregular INC was combined with other degenerative grades and most common at lower levels.
Irregular small sagittal complex was rare in any combination but always found in discs with other degenerative findings.
Black discs were not frequent,
only found at lowest levels,
most at L5/S1,
and about half of them had a bulge.
Most combinations of degenerative changes were rare and there was a wide variation in the appearance of degenerated discs among both age groups since MRI variables showed different grades or severity of degenerative findings,
according to disc level in particular.
Two main factors were found in the analysis (Table 4).
Variables including changes in annulus fibrosus showed strongest associations with the most common factor 1 at all disc levels and variables including focal changes and inhomogeneities in nucleus pulposus at L2/L3 and L5/S1 level,
too.
Factor 2 was less common.
Nuclear complex,
intranuclear cleft,
its focal lesion and inhomogeneity of the nucleus had a weaker association with it,
more clearly at L3/L4 and L4/L5 levels.
Bony endplate lesions and subchondral signal changes had a weak association only with factor 3.
They were excluded from the final factor analysis.
In the final analysis three factors were found (table 5).
They represent different types of degenerative processes.
Factor I is associated with changes in annulus fibrosus,
factor II with changes in nuclear complex and intranuclear cleft.
Factor 3 is associated weakly with decrease of signal intensity in nucleus pulposus (at three lower disc levels).
Variables including changes in annulus fibrosus correlated strongest with other variables and thus they could be the essential signs of the degenerative process in disc.
Sagittal complex,
inhomogeneity of the nucleus pulposus and regularity of the intranuclear cleft seem to be signs or features of a different type of degenerative process,
obviously taking place in the nucleus pulposus and inner annulus.
Signal intensity of the nucleus pulposus had a weaker association with other variables,
e.g.
variables representing internal structural changes in nuclear complex region.
Change in signal intensity of nucleus pulposus may have a different course during the degenerative process than the other variables as found earlier in a prospective study of LBP patients [7].
Bony endplate lesions and subchondral signal changes had the weakest correlation with other variables but had a weak correlation with each other and with factor 3 in the principal factor analysis.
They may be signs of a specific type of degenerative process,
manifesting as changes at the border of the disc and vertebral bone.
Bony endplate changes were as common among the young as middle aged and many were not associated with other changes.
Our results support the hypothesis that a part of them has a different etiology than the degenerated discs without such lesions.
They were rare findings.
Also many other combinations of graded variables were rare.
The most common combinations may represent slowly progressing,
age dependent degeneration.