The Incidence of Pars Interarticularis Defects in Athletes

Data Sources

The authors performed a comprehensive search of the published medical literature, using the following electronic databases from their dates of inception to September 2017: PubMed, Ovid Medline, EMBASE, Google Scholar, Cochrane Database of Systematic Reviews, and Cochrane Database of Controlled Trials. Searches were performed with the following terms as MeSH headings and keywords, with Boolean operations, including “AND” and “OR”: “spondylolysis,” “sports,” “low back pain,” and “pars defects.” Synonyms were used to identify the remainder of relevant studies. Citations and abstracts were retrieved. A hand search of the bibliographies was also performed to identify relevant articles missed by the electronic search. Two independent researchers performed the literature search.

Study Selection Criteria

Articles were included in the current review if their primary subject matter was the classification, epidemiology, risk factors, diagnosis, or management of pars interarticularis defects, or if they discussed pars interarticularis defects in athletic populations. Articles were excluded if the athletic population was not a significant focus of the article, or if they discussed the total incidence of spinal disorders without providing statistics specifically on pars interarticularis defects. Articles were also excluded if they were not in English, unless the statistics from such articles were recounted in other English articles, in which case they were included in the current review. The articles were reviewed for inclusion or exclusion independently by 2 of the authors, and disputes were resolved by group consensus.

Epidemiology

The incidence of spondylolysis in neonates is zero, ^17,24 4.4% in 6-year-olds, and 6% in adults. ¹⁷ The incidence in athletic populations is considerably higher (Table 2). Approximately 75% of spondylolysis will develop into spondylolisthesis. ¹⁷

OPEN IN VIEWER

Table 2.

Incidence of Pars Interarticularis Defects by Sport.

Sport	Sample Size	Athletes With at Least One Pars Interarticularis Defect
		n	%
Archery
Rossi et al 26	26	0	0
Soler et al 27	44	1	2.27
Total	70	1	1.43
Badminton
Kobayashi et al 31	3	2	66.66
Soler et al 27	38	2	5.26
Total	41	4	9.76
Baseball
Rossi et al 26	21	1	4.76
Kobayashi et al 31	60	37	61.67
Kobayashi et al 31,a	3	0	0
Kono et al 36	91	14	15.38
Akimoto et al 37	143	18	12.59
Hasegawa et al 59	95	22	23.16
Matsumoto et al 60	40	11	27.5
Wakitani et al 61	71	38	53.52
Total	524	141	26.91
Basketball
Kono et al 36	66	9	13.64
Rossi et al 26	174	17	9.77
Kobayashi et al 31	30	15	50
Soler et al 27	288	19	6.6
Akimoto et al 37	96	11	11.46
Total	654	71	10.86
Baton twirling
Kobayashi et al 31	1	0	0
Bobsledding
Rossi et al 26	36	5	13.88
Rossi et al 26,b	25	2	8
Soler et al 27	15	3	20
Total	76	10	13.16
Boxing
Rossi et al 26	27	3	11.11
Soler et al 27	21	3	14.29
Total	48	6	12.5
Canoeing
Rossi et al 26	69	8	11.59
Soler et al 27	162	10	6.17
Total	231	18	7.79
Cricket (fast bowling)
Crewe et al 41	46	15	32.61
Ranson et al 34	28	12	42.86
Hardcastle et al 42	22	12	54.54
Engstrom et al 43	51	18	35.29
Foster et al 44	82	9	10.98
Annear et al 45	20	9	45
Elliott et al 46	20	11	55
Total	269	86	31.97
Cycling
Rossi et al 26	95	13	13.68
Soler et al 27	175	11	6.29
Total	270	24	8.89
Diving
Rossi et al 26	57	23	40.35
Soler et al 27	8	0	0
Total	65	23	35.38
Equestrian
Rossi et al 26	83	5	6.02
Soler et al 27	16	0	0
Total	99	5	5.05
Fencing
Rossi et al 26	143	19	13.28
Soler et al 27	56	6	10.71
Total	199	25	12.56
Football (American)
Jones et al 51	104	5	4.81
Rossi et al 26	400	65	16.25
McCarroll et al 52	145	22	15.2
Soler et al 27	13	0	0
Semon et al 53	58	12	20.69
Iwamoto et al 54	742	77	10.38
Ferguson et al 55	12	6	50
Abe et al 56	210	42	20
Total	1720	229	13.32
Golf
Rossi et al 26	38	2	5.26
Soler et al 27	52	1	1.92
Total	90	3	3.33
Gymnastics
Rossi et al 26	673	112	16.64
Mohriak et al 62	18	1	5.56
Kobayashi et al 31	5	2	40
Toueg et al 11	92	6	6.52
Sward et al 32	52	8	15.38
Toueg et al 63	93	6	6.45
Soler et al 27	235	33	14.04
Bennett et al 64	13	4	30.77
Kono et al 36	49	4	8.16
Akimoto et al 37	61	5	8.2
Jackson et al 65	100	11	11
Total	1391	192	13.80
Handball
Rossi et al 26	42	3	7.14
Kobayashi et al 31	1	0	0
Soler et al 27	67	5	7.46
Total	110	8	7.27
Hockey (ice and field)
Rossi et al 26	170	13	7.64
Soler et al 27	106	3	2.83
Donaldson et al 25	25	11	44
Suzuki et al 28	63	10	15.87
Total	364	37	10.16
Javelin
Schmitt et al 66	21	14	66.67
Martial arts
Rossi et al 26	64	10	15.62
Kobayashi et al 31	3	1	33.33
Kobayashi et al 31	2	1	50
Soler et al 27	43	4	9.3
Kono et al 36	38	4	10.53
Akimoto et al 37,c	40	5	12.5
Akimoto et al 37,d	49	5	10.2
Kuroki et al 67	21	7	33.33
Total	260	37	14.23
Motorcycling
Rossi et al 26	8	0	0
Mountaineering
Soler et al 27	63	1	1.59
Paddleball
Soler et al 27	20	2	10
Pole vaulting
Rebella et al 68	135	4	2.96
Rowing
Maurer et al 69	22	6	27.27
Rossi et al 26	246	19	7.72
Soler et al 27	77	13	16.88
Total	345	38	11.01
Rugby
Rossi et al 26	65	7	10.76
Kobayashi et al 31	5	1	20
Soler et al 27	40	2	5
Abe et al 56	169	52	30.77
Total	279	62	22.22
Sailing
Rossi et al 26	128	22	17.18
Shooting
Rossi et al 26	76	8	10.52
Soler et al 27	81	4	4.94
Total	157	12	7.64
Skating (ice)
Rossi et al 26	42	3	7.14
Skating (roller)
Soler et al 27	7	0	0
Skiing
Rossi et al 26	154	25	16.23
Rossi et al 26,e	18	2	11.11
Soler et al 27	77	6	7.79
Total	249	33	13.25
Soccer
Kobayashi et al 31	47	25	53.19
Sward et al 32	31	2	6.45
Soler et al 27	55	1	1.82
Murase et al 70	160	14	8.75
Kono et al 36	264	23	8.71
Akimoto et al 37	320	28	8.75
Kyo et al 71	37	18	48.65
Matsumoto et al 60	60	19	31.67
Total	974	130	13.35
Sumo
Nakagawa 72	37	5	13.51
Swimming
Rossi et al 26,f	307	34	11.07
Kobayashi et al 31	1	1	100
Engstrom et al. 43	20	4	20
Soler et al 27	176	18	10.23
Soler et al 27,g	11	1	9.09
Kono et al 36	55	10	18.18
Akimoto et al 37	117	12	10.26
Total	687	80	11.64
Table tennis
Rossi et al 26	1	0	0
Kobayashi et al 31	4	3	75
Kono et al 36	25	3	12
Akimoto et al 37	34	4	11.76
Total	64	10	15.63
Tennis
Rossi et al 26	306	36	11.76
Kobayashi et al 31	10	4	40
Sward et al 32	30	3	10
Soler et al 27	91	1	1.1
Maquirriain et al 33	139	3	2.16
Alyas et al 73	33	9	27.27
Rajeswaran et al 35	98	29	29.59
Kono et al 36	53	8	15.09
Akimoto et al 37	67	9	13.43
Total	827	102	12.33
Track and field
Rossi et al 26	353	61	17.28
Rossi et al 26,h	54	11	20.37
Kobayashi et al 31	13	4	30.77
Soler et al 27	685	61	8.91
Kono et al 36	144	19	13.19
Akimoto et al 37	206	23	11.17
Soler et al 27,i	9	0	0
Soler et al 27,j	90	7	7.78
Total	1554	186	11.97
Volleyball
Rossi et al 26	150	16	10.66
Kobayashi et al 31	12	1	8.33
Soler et al 27	70	7	10
Külling et al 12	29	6	20.69
Akimoto et al 37	53	2	3.77
Total	314	32	10.19
Weight lifting
Rossi et al 26	112	25	22.32
Kotani et al 74	26	8	30.77
Soler et al 27	85	11	12.94
Granhed et al 75	13	2	15.38
Total	236	46	19.49
Wrestling
Rossi et al 26	80	20	25
Soler et al 27	143	16	11.19
Sward et al 32	30	2	6.67
Granhed et al 75	32	4	12.5
Total	285	42	14.74

^a Softball.

^b Luge, not bobsledding.

^c Judo.

^d Kendo.

^e Water skiing.

^f Synchronized swimming and water polo.

^g Synchronized swimming.

^h Pentathlon/triathlon.

ⁱ Pentathlon.

^j Triathlon.

Classification

The currently accepted classification system for spondylolysis (Table 3) (type IIa spondylolisthesis) is based on the progression of pathological changes that occur in pars interarticularis defects. Radiological magnetic resonance imaging (MRI) changes are grouped in 5 grades (grades 0-4). ¹ Grade 0 is a normal pars interarticularis, with no evidence of stress reaction. Grade 1 refers to T2 signal abnormalities of the pars interarticularis but not of the adjacent pedicle or articular process, representing a bone marrow stress reaction (edema) without cortical disruption. Grade 2 refers to T2 signal abnormalities and thinning, fragmentation or irregularity of the pars interarticularis on T1 or T2 image, representing incomplete pars interarticularis fractures that have not yet caused cortical disruption. Grade 3 refers to complete unilateral or bilateral cortical disruption (spondylolysis) with T2 signal abnormalities, representing acute complete fractures of the pars interarticularis. Grade 4 involves cortical disruption without abnormal T2 signals, representing old pars interarticularis fractures that have not united. ⁷⁶

OPEN IN VIEWER

Table 3.

Classification of Pars Interarticularis Defects (Type IIa Spondylolisthesis) by Radiological and Pathological Features.^a

Grade	Radiological Change	Pathological Change
0	Nil	Nil
1: Stress reaction	T2 signal abnormalities, with no pars defect on T1/T2 images	Bone marrow edema
2: Incomplete fracture	T2 signal abnormalities, with thinning, fragmentation, or irregularity of the pars interarticularis on T1 or T2 images	Bone marrow oedema with pars defect (incomplete fracture), but spondylolysis (cortical disruption) has not yet occurred
3: Complete fracture	Cortical disruption on T1/T2 images, with T2 signal abnormalities present	Cortical disruption (spondylolysis) has occurred and bone marrow edema is present, but reunion is still possible
4: Chronic complete fracture	Cortical disruption on T1/T2 images, with no T2 signal abnormalities present	Complete pars interarticularis fracture that has never reunited, with no associated bone marrow edema

^a Spondylolysis is characterized by cortical disruption (grade 3 and 4 defects). Grade 1 and 2 defects are prespondylolytic.

The 5 grades of radiological changes can be used in various combinations to produce classifications of varying categories. Currently, most classifications use a 5-category system (grade 0, grade 1, grade 2, grade 3, grade 4), but the most reliable is likely to be a 3-category system (grade 0, grades 1-3, grade 4). ¹ In any case, the classification used must achieve adequate accuracy while still being able to differentiate between the various stages of pars interarticularis damage.

The currently accepted classification of spondylolisthesis (Table 4) is based on etiology: type I (dysplastic), type II (isthmic), type III (degenerative), type IV (traumatic), type V (pathologic), and type VI (postsurgical). ² Dysplastic spondylolisthesis occurs due to a congenital defect in the neural arch. Isthmic spondylolisthesis occurs due to a defect in the pars interarticularis, and has 3 subtypes: type IIa (spondylolytic, occurring due to stress fractures of the pars, such as by repetitive hyperextension and twisting), type IIb (repeated microtrauma occurs to the pars; as it heals, nonlinear forces cause the pars to elongate and thin, making it susceptible to future fracture), and type IIc (acute traumatic fracture of the pars). Degenerative spondylolisthesis occurs due to progressive degeneration of the facet joint complex (eg, capsule, ligaments). Traumatic spondylolisthesis is caused by acute spinal trauma, which leads to a posterior column fracture, but not pars fracture. Pathologic spondylolisthesis occurs due to a pathological process such as infection, malignancy, or endocrine disorder. Postsurgical spondylolisthesis occurs due to postsurgical lumbar instability. The focus of this review is pars interarticularis defects in athletes (type IIa spondylolisthesis).

OPEN IN VIEWER

Table 4.

Classification of Spondylolisthesis by Etiology.^a

Type	Etiology	Pathogenesis
I	Dysplastic	Congenital defect in the neural arch
II	Isthmic	Pars interarticularis defect
IIa		Stress fracture of the pars (spondylolysis)
IIb		Repeated microtrauma and nonlinear forces cause elongation of the pars
IIc		Acute traumatic fracture of the pars
III	Degenerative	Degeneration of the facet joint complex (capsule, ligaments)
IV	Traumatic	Acute traumatic fracture of posterior column, but not the pars
V	Pathologic	Infection, neoplasm, endocrine disorder, or other pathology causes vertebral instability
VI	Postsurgical	Postsurgical lumbar instability

^a Pars interarticularis defects in athletes correspond to type IIa spondylolisthesis.

Spondylolisthesis may also be classified into developmental and acquired causes. ^77,78 Developmental spondylolisthesis corresponds to the dysplastic and isthmic etiologies. Acquired spondylolisthesis corresponds to the traumatic, pathologic and postsurgical etiologies.

Risk Factors

Known risk factors for pars interarticularis defects include childhood/adolescence, male sex, race, particular sports, and other lifestyle choices that lead to repetitive hyperextension and rotation of the lumbar spine. ^{24,79 -81} Possible risk factors include family history and spina bifida occulta. ^4,5

Spondylolysis is more common in children since their bones are still in the growth stages, with weaker osteochondral junctions and thinner cortices. Children also participate more frequently in physical activity than adults, not allowing sufficient time for recovery from microtrauma. ⁷⁹ Another reason that children develop pars defects more commonly than adults is that children can develop injuries due to unaccustomed loads. In other words, a child who is an expert at soccer may be injured playing a sport they are not accustomed to (eg, tennis), even if tennis presents similar loading forces on the body. Contrastingly, adults do not often develop injuries from unaccustomed loads. ⁸² Males develop spondylolysis more commonly than females with a ratio of 2:1, ^24,80 but females are more likely to develop severe spondylolisthesis. ⁸³ Incidence in white males is 6.4%, black males is 2.8%, white females is 2.3%, and black females is 1.1%. ²⁴ Eskimos have the highest incidence, ⁸¹ supporting a genetic predisposition to spondylolysis. ⁵ Spondylolysis occurs more commonly in athletes ⁸¹ due to repetitive hyperextension and twisting movements of the lumbar spine, as previously mentioned.

Family history and spina bifida occulta (SBO) are possible risk factors for pars defects. Sixty-three percent of L5 spondylolysis exhibits terminal-stage defects despite young skeletal age, while most L3/L4 spondylolysis exhibits early-stage defects. ⁴ This suggests that certain risk factors (possibly genetic) predispose to L5 spondylolysis, but not to spondylolysis at other lumbar levels. Another possibility could be the fact that L5 experiences more stress than other lumbar levels. Of the patients studied by Sakai et al, ⁴ 93% of L5 spondylolysis patients suffered from SBO, while 0% of the L3/L4 spondylolysis patients suffered from SBO. This indicates a strong correlation between L5 spondylolysis and SBO, either due to a common genetic predisposition, with autosomal dominant inheritance ^84,85 or because SBO directly predisposes to spondylolysis. This is supported by Yamada et al, ⁵ who studied 3 brothers—2 of whom were twins—who all developed lumbar spondylolysis and concomitant SBO. It is postulated that defective osteogenesis during the growth period leads to both SBO and spondylolysis. ⁴

Athletes have been shown to have larger sacrohorizontal angles (angle between the lumbar vertebra and upper end-plate of the sacrum) than nonathletes, ³² leading to increased lumbar lordosis. Increased lumbar lordosis increases the risk of pars interarticularis defects due to greater shear and compressive forces on the lumbar spine. ^86,87 Lumbar lordosis may also be increased during adolescence, when rapid bone growth causes tightness of the iliopsoas and thoracolumbar fascia. ^19,88 Additionally, specific sporting positions may be associated with a higher body mass index, such as defensive players in rugby, and higher body mass index is associated with increased lumbar lordosis. ^89,90

Investigations

Imaging modalities are the mainstay of diagnosis in pars defects since clinical assessment is unreliable, ⁸ but there is no universally agreed algorithm for the diagnostic workup. ⁹¹ A trial of conservative management may be attempted before undertaking any imaging investigations, if the lower back pain is of recent onset and there is insufficient clinical suspicion for spondylolysis. ⁹² Oblique and lateral X-rays of the lumbar spine may be performed as an initial investigation, with lucency of the pars interarticularis indicating a pars defect. The pathognomonic sign on oblique lumbar X-ray is the “Scotty dog collar” sign (the “Scotty dog” is the appearance of the normal spine, and the “collar” indicates the nondisplaced pars fracture). ⁹³ Computed tomography (CT), MRI, and single photon emission computed tomography (SPECT) are all more sensitive than X-ray ⁸ and allow greater appreciation of the spinal anatomy, but their exact role in the diagnostic workup remains debated. Generally, early-stage disease is best detected by SPECT, but MRI has an increasingly positive role in this area. ⁹³ CT is excellent for assessing more progressive disease, determining fracture size and extent, and providing a baseline on which to assess adequacy of healing. ⁹³ In all cases, a high index of suspicion is required to detect less common types of pars defects, such as unilateral spondylolysis, spondylolysis of the upper lumbar vertebra (L1-L3), multilevel spondylolysis, and early stage lesions in which cortical disruption has not yet occurred. ^94,95

Management

There is no gold standard protocol for the management of pars interarticularis defects. Further studies are required to compare conservative therapy to surgical therapy, and to compare Buck’s repair with Scott’s wiring technique, Morscher technique and other novel surgical techniques involved in these defects. The literature summatively suggests a mean return-to-play time of 3.7 months for conservative therapy, and 7.9 months for operative therapy. ⁹⁶

Only 5 studies in this review reported the management instituted for their participants, ^{25,42,43,52,65} and these were all case series (Table 5). Of 74 players with spondylolysis in these studies, 70 (94.59%) underwent conservative treatment and 4 (5.41%) underwent surgical treatment. Sixty-one (82.43%) returned to their previous level of play, 6 (8.11%) retired, ^25,42 and the disposition of the final 7 was not reported. ⁴²

OPEN IN VIEWER

Table 5.

Studies in Which Treatment Modality and Return-to-Play Time Were Reported.^a

	No. of Athletes With Pars Defects	Conservative Therapy	Surgical Therapy	Returned to Previous Level of Play	Retired
Donaldson et al 25	11	11	0	10	1
Hardcastle et al 42	12	9	3	Not reported	5
Engstrom et al 43	18	18	0	18	0
McCarroll et al 52	22	21	1b	22	0
Jackson et al 65	11	11	0	11	0
Total	74	70	4	61	6

^a Five studies reported treatment modality for participants found to have spondylolytic defects.

^b Intervertebral disc surgery, not pars interarticularis surgery.

Surgical Therapy

Surgical intervention is indicated if there is failure of conservative treatment after 6 months, persistent back pain after 9 to 12 months, or pars pseudoarthrosis (nonunion after 9-12 months). ^97 104,105 Spondylolisthesis more than 50% in those who have not reached skeletal maturity, neurological deficit and radiculopathy are relative indications for surgical management. ¹⁰⁴ There are several methods of surgical management. A Buck’s repair involves screw fixation of the pars interarticularis. ¹⁰⁶ Variations include the Scott wiring technique ¹⁰⁷ and the Morscher technique ¹⁰⁸ (Table 6). Minimally invasive techniques have also recently been described, with quicker postoperative recovery but greater technical difficulty. ^{97,109 -112} Currently, there are no randomized controlled trials comparing different techniques of pars repair.

OPEN IN VIEWER

Table 6.

Buck’s Repair, Scott’s Wiring Technique, and Morscher Technique Are Used for Surgical Fixation of Pars Defects.

Buck’s repair	Screw fixation of the pars interarticularis to compress the defect
Scott’s wiring technique	A wire is passed through the transverse processes of the vertebrae, then wrapped around the spinous processes. This stabilizes the defect and allows bony repair
Morscher technique	Bone graft fills the defect then screws are inserted into the superior articular processes. A hook hangs over the lamina and is secured by a lock

Comparison of surgical outcomes is difficult as surgeons use various intraoperative and postoperative regimens in treating pars defects. Most surgeons immobilize patients for 3 months postoperatively in a rigid lumbosacral orthosis, then allow graduated return to activity. ^{104, 113} Most surgeons allow resumption of sport after 6 months postoperatively in noncontact sports, and 12 months postoperatively in contact sports, although some surgeons advise patients to never recommence contact sports. ^{104, 113} Approximately 50% of surgeons advise patients never to recommence collision. ¹¹³ In all cases, patients must be fully rehabilitated before returning to athletic activity. This involves no pain with sport-specific activity, as well as full strength and range of motion. Naturally, some athletic activities can never be recommenced using these criteria since they require extreme ranges of motion that cannot be recommenced after surgery, such as dancers who require lumbar hyperextension for their vocation. ^{112, 113} The main factors ultimately influencing return to athletic activity are symptomaticity and time from surgery; it is unclear whether radiographic appearance is one of the least important ^104,112,113 or most important factors ^112,114 influencing return to athletic activity.

Prevention

Prevention of pars defects by targeting modifiable risk factors may be an important adjunct against this disorder. For example, notifying cricketers of the increased risk using a mixed front-and-side action or tennis players of the increased risk using a topspin serve may lead to decreased incidence of pars defects. Coaches should also be trained in the provision of safe training routines, and high-risk maneuvers should be reserved for competitive play and not employed during training sessions.