Introduction

Rotator cuff (RC) lesions are the most common reason for shoulder pain, dysfunction and discomfort [1, 2]. It is commonly  acknowledged that the prevalence of rotator cuff tears increases with age [3,4,5]. Various reports indicate that rotator cuff lesions occur in up to 22% of all patients over 65 years of age [6]. Partial thickness tears remain difficult to identify both in clinical tests and with magnetic resonance imaging [7]. However, the lesion size increases without surgical treatment [8]. Initially repairable lesions develop to irreparable lesions, while the acromio-humeral distance decreases with an increase of gleno-humeral arthritis after 4–5 years. However, the ideal therapy of RC lesions is still controversial including conservative and operative treatment, while in case of an operation the arthroscopic approach has advantages over an open and mini-open technique [9, 10]. This includes less soft tissue damage, reduced pain, faster rehabilitation, lower postoperative infection rates avoiding accordingly partial deltoid muscle insufficiencies and shoulder stiffness. At later stages the clinical results do not show significant differences in comparison of open or arthroscopic surgery [11].

The results of operatively reconstructed RC lesions are in general good and have been improved during the last 15 years using improved anchor systems with lower re-tear rates using newer self-reinforcing knotless systems [12]. There is still an ongoing intense research effort to understand biomechanics, treatment results and prognostic factors making RC lesions one of the most highly investigated orthopedic pathology. This fact is reflected obviously in the review article by McElvany et al. including 2.383 studies with 108 inclusion criteria from 1980 till 2012 [13].

A major concern after RC repair is still the occurrence of re-tears or biomechanical healing failure with a rate of 27 up to 46% depending on rupture patterns, RC morphology and surgical technique, such as single and double-row reconstruction [14,15,16]. This is of specific interest as RC re-tears seem to occur between 12 and 24 months postoperatively as recently shown in a MRI based follow-up study [17].

Additionally, many biomechanical studies have been performed to optimize stability and, therefore, reduce healing failure. Moreover, animal studies examining tendon-to-bone healing showed that the repair tissue differs dramatically from the native bone–tendon interface. This might also correlate with healing failures in human shoulders [18,19,20].

Identified negative prognostic factors for RC re-tears after reconstruction are fatty degeneration of the RC muscle according to Goutailler et al. [21] and according to Thomazeau [22]. Relevant further factors are the tear size according to Bateman et al. and the tendon retraction according to Patte et al. leading to impaired clinical outcome and higher re-tear rates for greater lesion sizes [21, 23, 24]. Another independent negative prognostic factor was identified by Chung et al. showing that osteoporosis is accompanied with higher failure rates [25].

The aim of this study was to correlate, if obesity was a risk factor for RC repair as obesity was proven to be a significant negative prognostic factor for postoperative complications in orthopedic trauma surgery [26]. The hypothesis of this study was, therefore, that obesity causes poorer Constant and DASH scores and also provokes higher re-tear rates after mini-open and arthroscopic double-row RC reconstruction.

To the best to our knowledge there are no comparable studies available, yet.

Materials and methods

Study population

Inclusion criteria were a complete rupture of one of the RC tendons, a reconstruction with anchors and a non-operated contra-lateral shoulder as well as surgery between 2006 and 2010. Two hundred and fifty-three patients were identified of which 146 patients with RC tears were included retrospectively in this case–control study of operatively treated RC tears. Exclusion criteria for the study were death (3), severe illness (7), residence abroad (21) or travel distance to our clinic of more than 2 h (76).

Seventy-five patients were treated using a mini-open technique and 71 patients arthroscopically by two different surgeons in a single-center for sports injuries and arthroscopic surgery. The operative procedure was standardized and identical for the mini-open and arthroscopic group using for the medial row resorbable Bio-Corkscrew FT Anchors (Arthrex) and for the lateral strutting row the knotless resorbable SwiveLock© Anchor (Arthrex). In all cases simultaneous subacromial decompression and a tenotomy or tenodesis of the long biceps tendon was performed. The ages of the subgroups were similar with 59.0 ± 9.1 years in the arthroscopy group and 56.7 ± 10.1 in the mini-open group. The analysis and classification of the body-mass index (BMI) was performed according to the WHO definition with three categories [27]: BMI normal (n = 38 patients, BMI between 18.5 and 25), BMI pre-obese (n = 73 patients, BMI between 25 and 30) and BMI obese (n = 30 patients, BMI over 30). Table 1 shows more detailed information of the patients' weight pattern.

Table 1 Patient demographics according to BMI
Full size table

The tear size (TS) was categorized in four types: TS group 1 (n = 36): small rupture of a single tendon up to Bateman I/Patte (1) TS-group 2 (n = 56): rupture of a single tendon up to Bateman II/Patte (2) TS-group 3 (n = 44): rupture of several tendons up to Bateman III/Patte (3) TS-group 4 (n = 10): rupture of several tendons up to Bateman IV/Patte 3.

Follow-up examinations took place at a mean of 43 ± 16 months after initial surgery (minimum of 24 months).

None of the patients had relevant body-weight change overtime from surgery to time of examination.

The study was performed with the consent of the local ethical committee.

Outcome measures

RC integrity was examined by means of a high-resolution ultrasound of all patients using a Siemens Sono site with 13.5 MHz linear-array probe. In all patients both shoulders were examined and the integrity was judged in a standardized multi-planar manner as previously recommended [28, 29].

The follow-up examination included the DASH score and Constant–Murley shoulder score [30]. The range of motion was noted and shoulder pain was assessed by using the VAS (visual analogue scale).

Statistical analysis

The statistical analysis was performed with Stata software v.10.0. Nonparametric tests were used depending on the sample size. Quantitative variables were tested by the Wilcoxon test for paired groups. A “p” value of less than 0.05 was considered to be statistically significant.

Results

Figure 1 shows a homogeneous distribution of the TS in each BMI group (Fig. 1). This analysis was necessary to objectify an independent influence of the BMI on the clinical outcome and re-tear rate.

Fig. 1
figure 1

Distribution of RC tears according to BMI

Full size image

In the clinical examinations obese patients had significantly worse results in the DASH and Constant–Murley score at the follow-up examination (Fig. 2). Interestingly there were no differences between normal-weight patients and pre-obese patients (Constant–Murley p = 0.22; DASH p = 0.47).

Fig. 2
figure 2

Comparison DASH/Constant score according to body weight

Full size image

In comparison, there are similar clinical results with no significant differences for both the arthroscopic and the open treatment: Constant–Murley score 78.3 ± 18.2 arthroscopic vs. 77.0 ± 21.8 for open surgery; DASH 12.7 ± 18.2 arthroscopic vs. 15.6 ± 21.6 for open surgery (p = 0.81). The subjective results showed a median of 27.6 ± 4.4 points after arthroscopic reconstruction and 27.4 ± 4.4 after mini-open procedure. The objective results revealed a median score of 39.2 ± 3.7 (arthroscopic) vs. 38.5 ± 3.6 (mini-open) without significant differences. A detailed overview is provided in Table 2.

Table 2 Comparison of the Constant–Murley scores (*Wilcoxon test, **Fishers exact test)
Full size table

The length of surgery was not longer in obese patients (57.2 min ± 10.2) compared to the other groups (normal 67.9 min ± 22.3/pre-obese 74.3 min ± 25.3).

The ultrasound results between mini-open and arthroscopic surgery were equivalent without relevant differences. An overview of the ultrasound follow-up is given in Table 3 [28, 29].

Table 3 Comparison of the ultrasound examination at follow-up
Full size table

The re-tear rate revealed 15.8% in BMI-group normal; 8.2% in BMI-group pre-obese and 28.6% in the BMI-group obese (p = 0.0086), respectively. However, there was no significant difference between the BMI-group normal and the BMI-group pre-obese (p = 0.12). In addition, the range of motion (ROM) was significantly limited in the obese group in comparison to the pre-obese group for abduction (p = 0.0401), flexion (p = 0.0007), outward rotation (p = 0.0155), inward rotation 90° (p = 0.008) and outward rotation 90° (p = 0.0151)—(Fig. 3).

Fig. 3
figure 3

Range of motion at follow-up by subgroups

Full size image

The analysis of postoperative quality of life due to the SF-36 and EQ-5D showed also significantly impaired results for obese patients (Fig. 4).

Fig. 4
figure 4

Quality of life due to the SF-36 and EQ-5D

Full size image

The independent analysis of the RC repair failure according to TS showed for group TS 1 a re-tear rate of 11.1%, and for the group TS 2 a re-tear rate of 12.5% (TS 1 vs. TS2 p = 1.00). Bigger lesions as for TS 3 had a re-tear rate of 18.2% (TS 1 vs. TS3 p = 0.53), and for the group TS 4 a re-tear rate of 30.0% (TS 1 vs. TS 4 p = 0.16).

Discussion

The key finding of this study was that obesity causes less favorable results in the Constant and DASH scores and showed higher re-tear rates after rotator cuff repair. The same negative influence was found for a BMI > 30 for the mini-open as well as the arthroscopic RC repair.

Risk factors for rotator cuff tears can be grouped in anatomical, therefore, mechanical, sports and lifestyle factors, metabolic and epidemic factors. Anatomic risk factors are accepted to derive from outlet and inlet shoulder impingement [31]. Inlet impingement meaning shoulder dysfunction due to dysbalanced rotator cuff tears, tendinitis calcarea or shoulder instability. As reported by Bigliani osteoarthritis of the AC joint, an overhanging acromion and a narrow subacromial space are accepted mechanical outlet risk factors [32].

Sportsman affected by rotator cuff lesions are mainly reported to be found in sports where a rotation in the shoulder is an essential element such as in handball, baseball, javelin or in gymnastics [33, 34].

Nowadays the focus is also on metabolic factors for the prevalence of RC tears. There is an interesting published case–control study of 381 patients with RC tears treated by arthroscopy and 220 control patients. This study proofed that obesity, measured through BMI and body fat is a significant risk factor for the occurrence and severity of a rotator cuff tear [35]. A possible explanation might be that higher mechanical strain due to arm weight on the shoulder joint and consecutively on the RC tendons may have negative effects.

In a recently published study with 206 patients Djerbi et al. reported on cardiovascular risk factors for the appearance of RC tears. This study identified smoking, high blood pressure, diabetes (DM), alcoholism, dyslipidemia, obesity and cardiovascular history as such factors [14].

Other recent studies report that also hypertension [36] and body fat [35] are proven risk factors for RC tears.

Kim et al. reported on postoperative outcomes after RC repair of patients suffering from metabolic syndrome with a high BMI, DM and dyslipidemia [37]. Significantly higher re-tear rates were seen in this study group of 180 patients for such patients. However, as obese patients tend to have DM and dyslipidemia it is difficult to separate BMI, DM and dyslipidemia as independent factors.

Warrender et al. already described obesity as having a negative impact on the functional outcomes and recovery in patients with a BMI > 30 [38]. Furthermore, it is generally known that DM and hypercholesterolemia have a negative influence on tendon healing [39]. However, although DM and dyslipidemia can negatively affect clinical outcomes, Dhar et al. found no difference in re-tear rates [40]. This is in contrast to our results as this analysis revealed that evident obesity causes in general inferior clinical results and reduced multi directional range of motion as well as higher re-tear rates.

The presented study has certain limitations. The retrospective design provides less evidence than prospective studies. As aforementioned healing after RC repair is multi-factorial and does also depend on other measurable factors such as tear size, tendon quality and muscle atrophy or as in this study BMI. Simple mechanical deficits such as poor bone quality or inadequate suture fixation as well as intense or early active mobilization may be further reasons for healing failure. The preoperative fatty degeneration of the RC was not measured and ultrasound was used as reference to control cuff integrity. Therefore, a logically consistent study should prospectively evaluate RC integrity after double-row repair by MRI performing a multi-factorial analysis (RC size, muscle atrophy, anatomic patterns etc.) to objectivize a hierarchy of these aforementioned risk and influencing factors. However, in this study, obesity could be identified as an independent negative factor with higher re-tear rates after cuff repair.

Conclusion

Obesity (BMI > 30) showed higher re-tear rates for both mini-open and arthroscopic RC repair. In addition, obesity causes less favorable results in ROM  as well as in the Constant and DASH scores.

Bearing these results in mind, one must reflect indications for cuff repair in obese patients and also motivate these patients for a healthier lifestyle.