Body image is more negative in patients with chronic low back pain than in patients with subacute low back pain and healthy controls

2.1 Participants

Within a multi-center study (for more detailed information refer to Heidari and colleagues [20]), 167 consecutive patients suffering from non-specific LBP were recruited from several rehabilitation and health establishments and fitness centers in North Rhine-Westphalia, Germany. Interested individuals were informed by the responsible health care provider or by one of the involved scientists, followed by a phone call. Each patient received a 10 EUR gift certificate as an incentive for study participation. Inclusion criteria for the present study were non-specific LBP [21] with pain located near the spine in the lumbar or sacral region [22], the age between 18 and 60 years and the ability to read and write in German. Exclusion criteria were assessed by a standardized questionnaire and consisted of the following parameters: specific causes of LBP (e.g. acute disc prolapse), infections, acute fractures, tumors, rheumatic diseases, other chronic pain diseases, current psychiatric disorders (i.e. psychoses). Of the 167 patients 31 declined to participate. Forty six patients were excluded for not fulfilling the inclusion criteria (6 due to missing LBP, 16 because of not fulfilling the age inclusion criterion, 24 because of exclusion criteria). Of the remaining 90 patients 13 were eliminated due to incomplete data. The final eligible population included 77 patients (Age M=36.08, SD=11.26, 53.2% female). Thirty three patients (42.9%) reported sLBP, whereas 44 patients (57.1%) suffered from cLBP.

Fifty-seven controls (Age M=35.32, SD=13.40; 42.1% female) were recruited in fitness centers, at the local university, through personal contact and among participants of workplace training. Eligibility criteria stipulated that the participants must be 18 years of age or older with an absence of LBP and having been active in sports at least for 1 h per week for at least 6 months prior to the investigation. The same exclusion criteria as stated for patients applied to the control group.

All participants received written details about the study and had to sign a consent form prior to completing several questionnaires. The study protocol was approved by the Medical Ethics Committee of the Ruhr-University Bochum.

2.2 Measures

Physical exercise

Physical exercise had to be indicated in exercise hours spent weekly (“How much time do you spend for training per week?”) and for patients before the occurrence of LPB (“How much time per week did you spend for training before the occurrence of LBP?”), as well (exercise frequency).

2.3 Data analysis

Descriptive analyses were conducted for patient samples with sLBP and cLBP and for the HC. Depending on the data distribution, data were presented as mean (M), standard derivation (SD), median or frequency (n/%). In case of a non-normal distribution of the dependent variables, log-10 transformation of the three BI-variables was conducted. If log-transformation did not lead to normalization, non-parametric analyses were performed. Depending on the distribution, group differences between patients with sLBP or cLBP and HC were calculated with ANOVAs and t-tests or Kruskal-Wallis-tests with Bonferroni-corrected post-hoc-tests. In a second step in the group of patients with sLBP and cLBP bi-variate correlations were conducted between the three respective aspects of BI (“self-acceptance”, “physical efficacy” and “health”), and pain intensity, pain disability and self-reported current exercise frequency and before onset of LBP. α-levels of significant correlations were adjusted using Bonferroni-corrections. A correlation of 0.10<r 0.30 is considered as low, 30<r<0.50 as medium, and r>0.50 as high [27]. Depending on the data distribution, Pearson or Spearman’s rank correlation was implemented. All tests were operated using SPSS, version 25 [28].

3.1 Group differences regarding the three dimensions of body image, pain parameters, exercise frequency, and gender

The analysis of variance (ANOVA) with the between-subjects factor “group” (HC, sLBP, cLBP) and BI “self-acceptance” as dependent variable showed a significant main effect (F₍₂₎=5.49, p<0.01). Patients with cLBP revealed more negative BI “self-acceptance” than HC (p<0.01) and also compared to patients with sLBP (p<0.01). No significant differences were found between patients suffering from sLBP and HC (see Table 3 and Fig. 1).

Fig. 1:

Means (SD) of body image of self-acceptance between groups. Group differences were assessed with ANOVA. **p<0.01. BI=body image; sLBP=patients with subacute pain; cLBP=patients with chronic pain.

Regarding BI “physical efficacy”, a Kruskal-Wallis-test (X²=14.55, p=0.001) and subsequent the Mann-Whitney-U-test revealed lower scores in BI “physical efficacy” in both patient groups compared to the HC (for cLBP p<0.001, for sLBP p<0.05). No differences were found between the two patient groups (see Fig. 2 and Table 3).

Fig. 2:

Boxplots and 95% confidence ranges of body image of physical efficacy between groups. Group differences were assessed with Kruskal-Wallis-tests and Mann-Whitney-U-tests. *p<0.05; **p<0.01. BI=body image; sLBP=patients with subacute pain; cLBP=patients with chronic pain.

Concerning BI “health”, the overall Kruskal-Wallis-test was significant (X²=15.12, p<0.01). Post hoc analyses indicated significant lower BI “health” scores in patients with cLBP compared to HC (z=−4.203, p<0.001). No differences were found between the subgroups of patients and between patients with sLBP and HC (see Fig. 3 and Table 3).

Fig. 3:

Medians and 95% confidence ranges of body image of health between groups; Group differences were assessed with Kruskal-Wallis-tests and Mann-Whitney-U-tests. **p<0.01. BI=body image; sLBP=patients with subacute pain; cLBP=patients with chronic pain.

No significant differences between sLBP and cLBP were found for pain intensity (t₍₇₅₎=−1.751, p>0.05) and disability (t₍₇₅₎=−1.317, p>0.1).

Concerning current exercise frequency the ANOVA with the factor “group” (HC, sLBP, cLBP) showed a significant main effect (F₍₂₎=7.73, p=0.001). Post hoc analyses indicated group differences between HC and patients with sLBP (p<0.01) and between HC and patients with cLBP (p<0.01), showing that HC were more physically active. There were no significant differences between the two pain groups (see Fig. 4).

Fig. 4:

Means of current exercise frequency between groups. Group differences were assessed with ANOVA. **p<0.01. sLBP=patients with subacute pain; cLBP=patients with chronic pain.

No significant differences were found between the two subgroups of patients in concerns of exercise frequency before the onset of LBP (t₍₇₀₎=−0.460, p>0.1). T-tests for the dependent variables indicated that over time neither sLBP-patients (t₍₃₀₎=−0.069, p>0.1) nor cLBP-patients (t₍₄₀₎=−0.091, p>0.1) differed in exercise frequency as reported before the onset of LBP and current exercise frequency.

Referring to gender, a χ² test was used to investigate gender distribution. The three pain groups showed no significant gender differences (X²₍₁₎=2.152, p>0.1). Further, comparing gender differences in the two pain groups in concerns of pain intensity and disability we calculated t-tests for sLBP (pain intensity: t₍₃₁₎=1.365, p>0.1; disability: t₍₃₁₎=0.834, p>0.1) and cLBP (pain intensity: t₍₄₁₎=1.215, p>0.1; disability: t₍₄₂₎=1.190, p>0.1) which were not significant. Finally, calculating Mann-Whitney-U-tests for possible gender effects with regard to BI, we found no differences within the three pain groups concerning “physical efficacy” (p>0.1) and “health” (p>0.1). T-tests for “self-acceptance” revealed no differences within the three pain groups equally: sLBP (t₍₃₁₎=−0.025, p>0.1), cLBP (t₍₄₂₎=−0.680, p>0.1), HC (t₍₅₅₎=−0.721, p>0.1).

3.2 Bivariate correlations between body image, pain outcomes and exercise frequency

Bivariate correlations between all three BI-dimensions and pain outcomes were calculated separately for patients with sLBP, patients with cLBP, and healthy controls. Patients with sLBP showed a moderate to high negative correlation of BI “health” with disability only (see Table 4).

In the group of patients with cLBP we found a moderate negative correlation between BI “physical efficacy” and disability. Additionally, BI “health” showed a moderate negative correlation to pain intensity and a highly negative correlation to disability. Further, cLBP-patients displayed a moderate to high positive correlation between BI “self-acceptance” and exercise frequency before LBP. BI “physical efficacy” was moderately to highly positively correlated to both current exercise frequency and exercise frequency before the onset of LBP (see Table 5). For the group of HC, correlational analyses showed a significant positive relation between BI “physical efficacy” and current exercise frequency (see Table 6).

4.1 Self-acceptance

Our results partly confirm our hypothesis that patients with cLBP will show a more negative BI “self-acceptance” than patients with sLBP and also compared to HC. Whereas the chronic patients differed significantly from both other groups, we saw no difference between the subacute and the healthy sample. These findings support an unpublished report from Deusinger [25], who found lower “self-acceptance” scores in patients with cLBP than in HC. As both studies are cross-sectional, the question of cause and consequence remains unanswered. A negative BI concerning “self-acceptance” may be responsible for initiating the process of chronification but not for the occurrence of LBP, presumably mediated by less readiness to engage in physical exercise [4]. Avoidance of physical activity has been shown to be one possible risk factor for the chronicity of LBP [14]. On the other hand, development of cLBP may lower feelings of attractiveness as one consequence of pain chronicity [29], [30]. The finding of our study that the chronic but not the subacute patients differed from the HC may indicate, that not the presence of pain itself but the chronicity of pain may lower feelings of attractiveness.

Concerning our correlation analyses with pain outcomes, in both patient groups BI “self-acceptance” was neither associated with pain intensity nor with disability. This leads us to the assumption that the occurrence of pain and disability themselves do not influence BI “self-acceptance”. However, our findings concerning exercise behavior may shed further light on the relation between LBP and BI “self-acceptance”. The positive correlation between BI “self-acceptance” and exercise frequency before onset of pain in the cLBP group indicates that feeling unattractive was associated with less exercise behavior. It can be speculated that in our sample low levels of exercise have led to the onset or chronicity of LBP as the role of physical exercise in the prevention of LBP has been shown in a number of studies [31], [32].

However, the report of current exercise frequency did not correlate with BI “self-acceptance”, possibly due to the fact that all patients were engaged in some kind of exercise treatment which may have developed to a standardized behavior. This is in line with several studies in patients with chronic musculoskeletal diseases [29], [33]. However, the low level of “self-acceptance” in the chronic sample could also be a consequence of pain and reduced exercise behavior.

4.2 Physical efficacy

Both patient groups displayed significantly more negative BI “physical efficacy” than HC, which partly confirms our hypothesis. These results suggest that both patient groups equally experienced their body as less efficient in terms of bodily strength and flexibility compared to the HC. Possibly, the presence of pain as well as the chronicity may have an effect on the BI “physical efficacy”.

Moreover, results indicate that cLBP-patients with a high BI of “physical efficacy” revealed a lower disability. Missing agility may consequently lead to the performance of movements in a very controlled manner resulting from self-protective behavior [30], [34]. Therefore, patients may feel disabled through pain and experience a more negative BI “physical efficacy”. This assumption is confirmed by research observing a connection between physical activity and physical efficacy among adolescents [35]. These results indicated that adolescents with higher self-reported efficacy show a higher level of physical activity, whereby physical efficacy is related to being satisfied with one’s efforts in training. No associations were found between the BI “physical efficacy” and the pain outcomes in the sLBP group in this sample. This might be explainable by the low pain level in our sample, not representing a typical clinical sample. Besides, patients suffering from LBP are often affected by pain only in one part of their body, whereas patients with other pain syndromes are affected in various body parts [29].

Further, for cLBP and HC we found positive correlations between BI “physical efficacy” and current exercise frequency. For cLBP we also found a positive correlation with exercise before the onset of LBP. In contrast, such a relation was not seen for sLBP. As mentioned above, physical activity has a positive impact on BI [36], [37]. Patients with cLBP who continue their physical activity seem to connote exercise positively, whereas patients with sLBP may struggle to decide whether exercise is helpful or painful. Struggling with an optimal activity level was reported for patients with persistent musculoskeletal pain by Sündermann and colleagues [30].

4.3 Health

As expected, patients with cLBP revealed a significantly poorer BI “health” than HC. However, patients with sLBP neither differ from healthy controls nor from patients with cLBP. This result suggests that patients with sLBP neither feel ill nor healthy. Maybe they are between hope of cure and fear of chronicity. On the one hand, previous research reveals that chronic pain decreases the personal well-being or health-related quality of life, evoked from disability and/or dissatisfaction [38]. In contrast, other studies indicate that individuals with a more positive BI show more health-related behavior and attitudes [4], leading to higher rates of well-being. Nevertheless, the dose-response direction still remains unclear.

Another finding of the present study was that a more negative BI “health” was associated with higher pain intensity in patients with cLBP. This result is in line with previous research having shown an impact of pain on health [4], [39], [40]. With regard to the present study, pain seems to have an impact dependent on pain duration, since patients with sLBP revealed no association. Furthermore, the higher patients with cLBP felt disabled the lower was the BI of “health”. Even patients with sLBP felt disabled, although we did not find any association between BI “health” and pain. We assume that due to the pain level in patients with sLBP not pain itself has an impact on health. Nevertheless, pain influences function, as already explored before, which may lead to disability concerning health issues.

Furthermore, we did not find any association between BI “health” and exercise frequency. Our results confirm earlier reports showing no differences in the physical activity level between cLBP and HC. Bousema and colleagues [41] suggested that patients with cLBP, who even increased their activity level, cope with their situation and are able to continue their daily routine. Moreover, previous research has shown positive effects of moderate exercise on health in LBP-patients [42], and further in BI “health” examining patients with mild depression [43] or breast cancer [44]. Further, being physically active or in contact with the aching body through physiotherapy might be the reason that our patients did not solely focus on their disabled back, but on positive aspects of their body as well, e.g. the functioning and healthy parts. Thus, they might already have a relatively high health-related BI “health” through physical activity. This is supported by our finding that both patient groups revealed equal exercise frequencies concerning hours spent on exercise and with respect to the time of measurement, current and before LBP. Importantly, most research has focused on negative BI and its improvement. However, having a negative BI does not equate having no positive BI. Among other aspects, a positive BI is defined as accepting the own body and feeling confident with it. Both, negative and positive BI are discussed as different concepts [45] and should be considered as such in prevention and therapy of LBP. Interventions improving self-esteem and lowering stress seem to be promising in the development of positive BI [46]. Another approach to modify BI might be the cognitive-behavioral therapy (CBT) [47] modified for LPB.

4.4 Gender and pain parameters and body image

In contrast to previous studies [48] we found no significant gender effects neither with regard to pain intensity and disability nor to body image. This might be explainable because all patients were physically active. As mentioned above, physical activity has a positive impact on BI. Reasons for exercising and therefore improving BI might be different, such as women exercise for attractiveness, whereas men exercise to change their look [17]. The effect seems to be the same.

4.5 Strengths and limitations

The present study has several strengths and limitations. One limitation is that all patients underwent physiotherapy or were frequently engaged in physical exercise which can influence both, BI and pain. Further, because patients are younger than a usual clinical sample, and because of reducing the patient groups due to the confounder age the present population is not representative. Moreover, by including only physically active or treated participants, respectively, we may have caused a selection effect with regard to BI. Consequently, due to the cross-sectional nature of this study, we cannot interpret cause and effect regarding pain and BI, as well as regarding physical exercise and BI. Further research should examine the BI of patients in a longitudinal way, probably yielding more information about the relationship to LBP.

However, the main strength of the current study is the differentiation between sLBP and cLBP, which may improve therapy options in LBP-treatment. Another strength is the investigation of different BI-dimensions, to the best of our knowledge, for the first time. This might help to understand the influence of both, LBP on separate BI-dimensions, and BI on LBP.