Relationship Between Quality of Life and Self-Efficacy in Persons With Spinal Cord Injuries

Methods 

Results 

Table 1 shows the descriptive statistics for the entire group of 106 as well as for able-bodied Australians. The SCI group was found to have significantly lower QOL scores on all 8 SF-36 domains when compared with Australian norms standardized for age and sex. Table 1 also shows the mean MSES level for the 106 persons with SCI. Pearson correlation analyses failed to find any significant associations between self-efficacy (MSES scores) and sex, age, level or completeness of lesion, and time since injury. Only level of lesion was found to be significantly associated with the QOL physical functioning domain (r=.34, P<.05; tetraplegia associated with lower QOL scores). No other significant relationships were found between QOL and age, sex, time since injury, and completeness of the lesion.

Table 1. Descriptive Statistics for the 106 Persons With SCI for the SF-36 Domains and the MSES Scores

	Variable	Mean ± SD	95% CI	SE	Mean Australian Norms ± SE	t	Effect Size
	Physical functioning	37.5±25	32.7–42.3	2.43	82.5±0.2	18.5⁎	1.80
	Role–physical	42.2±39	34.7–49.7	3.79	79.8±0.4	9.9⁎	0.96
	Bodily pain	51.9±29	46.2–57.6	2.87	76.8±0.3	8.6⁎	0.85
	General health	55.6±26	50.6–60.6	2.53	71.6±0.2	6.4⁎	0.61
	Vitality	52.7±22	48.4–57.1	2.17	64.5±0.2	5.4⁎	0.54
	Social functioning	65.7±29	59.9–71.4	2.88	84.9±0.2	6.6⁎	0.66
	Role–emotional	70.4±32	62.9–78.0	3.80	82.8±0.3	3.2⁎	0.39
	Mental health	69.3±20	65.5–73.2	1.94	75.9±0.2	3.4⁎	0.33
	MSES score	84.5±21	80.4–88.6	2.06	NA	NA	NA

NOTE. Australian able-bodied SF-36 norms standardized for age and sex. Abbreviations: CI, confidence interval; NA, not applicable; SE, standard error.

⁎ P<.01.

The MSES correlated significantly with all the SF-36 domains (r=.38 for bodily pain; r=.41 for mental health and role−physical; r=.43 for role−emotional; r=.54 for physical functioning; r=.60 for vitality; r=.64 for general health and social functioning, variances ranging from 14% to 41%). This consistent mild to moderate positive relationship between QOL and MSES (high self-efficacy related to higher QOL) was supported by the finding of substantial differences in QOL as a function of self-efficacy. Figure 1 shows the influence on the 8 SF-36 QOL domains as a function of self-efficacy level, that is, low self-efficacy versus high self-efficacy. MANOVA confirmed that significant differences existed between the 2 groups over the 8 measures (Wilks λ=.53, F8,97=10.7, P<.001). Scheffé post hoc tests showed that significant differences existed (df=104, P<.001) in all SF-36 domains between the 2 groups. Table 2 shows that effect sizes for QOL differences between population norms and persons with SCI possessing low self-efficacy ranged from moderate to very large, but effect sizes for the difference between the high self-efficacy group and the general population on QOL were moderate to large for only 3 domains (physical functioning, role−physical, bodily pain). No substantial differences existed between the general population and those persons with SCI having high self-efficacy for the remaining 5 domains (general health, vitality, social functioning, role−emotional, mental health).

View Large Image Download to PowerPoint

Fig 1. The influence of self-efficacy on the 8 SF-36 QOL domains. SCI persons with low self-efficacy (low SE) are compared with those with higher self-efficacy (high SE). NOTE. 95% confidence intervals (CIs) are shown. Abbreviation: SE, standard error.

Table 2. Differences Between the General Population and the SCI Group With Low Versus High Self-Efficacy as a Function of the 8 SF-36 Domains

	Variable	Mean Low Self-Efficacy ± SD	Difference to Norms (effect size)	Mean High Self-Efficacy ± SD	Difference to Norms (effect size)	Mean Australian Norms ± SE
	Physical functioning	27.7±21	2.61	48.0±25	1.38	82.5±0.2
	Role–physical	29.7±37	1.35	55.7±37	0.65	79.8±0.4
	Bodily pain	42.2±29	1.19	62.4±27	0.53	76.8±0.3
	General health	42.3±24	1.22	70.1±20	0.07	71.6±0.2
	Vitality	41.6±20	1.14	64.8±18	−0.02	64.5±0.2
	Social functioning	49.8±29	1.21	82.8±19	0.11	84.9±0.2
	Role–emotional	58.2±41	0.60	83.6±32	−0.02	82.8±0.3
	Mental health	63.1±21	0.61	76.0±17	−0.00	75.9±0.2

NOTE. Australian able bodied SF-36 norms standardized for age and sex.

No significant differences were found in the QOL domains as a function of sex, completeness of lesion (complete vs incomplete), time since injury, or age at injury. However, neurologic level (tetraplegia vs paraplegia) was found to influence QOL significantly in some domains (fig 2). MANOVA detected significant differences between the 2 groups (Wilks λ=.79, F8,95=3.2, P<.01) and Scheffé post hoc tests found significant differences between the tetraplegic and paraplegic groups (df=104, P<.05) only in the physical functioning and pain SF-36 domains. Figure 3 shows the QOL domains that differed as a function of pain intensity (low pain vs high pain group, median split of 42 on the SF-36 pain intensity Q.7). MANOVA analysis detected significant differences between the high and low pain intensity groups in QOL (Wilks λ=.69, F7,98=6.4, P<.001). Scheffé post hoc tests (df=104, P<.01) confirmed that differences existed between the 2 groups in all SF-36 domain functions except physical functioning (see fig 3).

View Large Image Download to PowerPoint

Fig 2. The influence of level of lesion as a function of those with tetraplegia versus paraplegia on the 8 SF-36 QOL domains. NOTE. 95% CIs are shown.

View Large Image Download to PowerPoint

Fig 3. The influence of those SCI subjects with high levels of pain intensity versus those with low levels of pain intensity on the 8 SF-36 QOL domains. NOTE. 95% CIs are shown.

Table 3 displays a 2×2 contingency table showing the number of participants with SCI who reported they had high versus low levels of pain intensity as a function of low versus high levels of self-efficacy. There was a significant tendency (χ12 test=6.25, P<.05) for people with high levels of pain to have low levels of self-efficacy, and vice versa (low levels of pain and high levels of self-efficacy). The positive predictive value or likelihood of pain being associated with low self-efficacy was found to be high (P=.67), with an odds ratio of 2.7 for someone with high pain intensity having low self-efficacy. A question may arise concerning the independence of the self-efficacy construct, that is, that self-efficacy is just another measure of pain. This is unlikely given that the strength of the relationship between self-efficacy and pain intensity was not strong (ϕ=.24, P<.05; ϕ2=.06 [or variance explained 6%]). Furthermore, reference to table 3 shows that 40 of the sample (38%) had self-efficacy and pain scores inconsistent with the notion that self-efficacy is another measure for pain intensity. The combination effect of a person with SCI having 2 QOL risk factors, that is, low self-efficacy and high pain intensity, was determined. The QOL differences (effect sizes [ES]) with the general population were increased even more in comparison with decreases associated with low self-efficacy and high pain intensity individually except for physical functioning (mean ± SD, 29.9±21; ES=2.5; mean role−physical, 21.9±36; ES=1.6; mean general health, 36.9±15; ES=2.3; mean vitality, 35.4±23; ES=1.3; mean social functioning, 37.9±18; ES=2.6; mean role−emotional, 48.7±41; ES=.83; mean mental health, 59.1±22; ES=.76).

Table 3. A 2×2 Contingency Table With the Number of Participants With SCI Who Reported They Had High Versus Low Levels of Pain Intensity as a Function of Low Versus High Levels of Self-Efficacy

	Self-Efficacy	High Pain (≤42)	Low Pain (>42)	Percentage
	Low self-efficacy (≤89)	37(35)	18(17)	52
	High self-efficacy (>89)	22(21)	29(27)	48
	Percentage	56	44	100

NOTE. Values are n (percent) or percent. χ12 test=6.25, P<.05, ϕ=.24, ϕ2=.06 [6% of the variance]; positive predictive value, .67; odds ratio, 2.7.

Discussion 

The findings of this research confirmed the results of a number of prior studies that have investigated the QOL of people with SCI.4, 7, 8, 25 The SCI sample that participated in the present study had consistently lower QOL scores than those in the Australian general population across all the 8 SF-36 domains. Not surprisingly, SCI had a large impact on physical functioning, physical role limitations, and pain. These 3 areas were found to have large effect sizes ranging from 1.8, .96, to .85, respectively, but general health difference between the SCI and general population was moderate (ES=.61). This comes as no surprise because prior studies have established that persons with SCI judge areas such as physical loss and limitations and ongoing debilitating pain to have severe long-term implications for their health and QOL.4, 8 Negative impacts for social functioning and vitality (ES=.66, ES=.54) were also found, but smaller though still impressive differences were found for the psychologic domains such as emotional role limitations (ES=.39) and mental health (ES=.33). This is also not surprising given that the social disadvantages arising from impairment and social participation restrictions are perceived as a major problem after SCI.8, 18 Additionally, given that poor mental health is a risk for a substantial minority with SCI,1, 10, 14 it was expected that psychologic aspects of QOL would also be reduced compared with the general population. Interestingly, some studies have not found significant differences between persons with SCI and the general population in the SF-36 mental health domain.4

Factors such as completeness of lesion, sex, age at injury, and time since injury have been inconsistently found to be related to lower QOL,2, 8, 18 and were not found to be associated with QOL in the present research. Women have been found to have a lower QOL after SCI,8 though this could be due to the unbalanced sex samples often used in research. Persons who are older at the time of the injury have been found to have lower QOL. Westgren and Levi8 found a poorer QOL in those over 20 years than those under 20 years of age at injury, and Stensman46 found this effect in those over 34 years at the time of the injury. It is thought older age reduces QOL after SCI due to lowered capacity (eg, reduced physiologic reserve, decreased vitality, aging body) to cope with the impairment.46 Obviously, more work needs to occur to clarify the influence of factors like age and time since injury. Level of neurologic impairment was found to be a significant factor interacting with QOL. People with higher lesions (tetraplegia) were found to have lower QOL, though differences occurred only in 2 SF-36 domains, these being physical functioning and bodily pain domains. It has not been clear from the literature that a higher lesion level will lower QOL relative to a lower lesion, though people with tetraplegia have been shown in recent research to have lower global QOL scores as well as the physical functioning domain.4, 8, 18 Unlike the findings of the present research, Haran et al4 found lower QOL related to bodily pain in people with paraplegia than in those with tetraplegia. Obviously, more work needs to be conducted on differences in QOL between paraplegia and tetraplegia.17

One of the major findings of this research was the substantial association found between QOL and a measure of self-efficacy specifically targeted to SCI. Prior research has produced data to suggest that expectations of control are associated with health quality in people with SCI35, 36, 37, 38, 39 and this was further supported by the findings of the present study. Persons with SCI with low levels of self-efficacy reported significantly lower levels of QOL than SCI persons with higher levels of self-efficacy and moreover, differences occurred in all 8 SF-36 domains. Table 2 shows the effect sizes for the SF-36 domain differences between persons with SCI having low versus high self-efficacy, as well as the SF-36 domains for the Australian general population. Scrutinizing these effect sizes, it becomes clear that self-efficacy beliefs are strongly associated with life quality. For those who believe they have low control over their behavior and life outcomes (ie, low self-efficacy), effect sizes ranged from moderate (mental health, role−emotional) to quite large effect sizes (the remaining 6 domains). In contrast, those SCI persons who believed they had some degree of control (high self-efficacy) had lower QOL for only 3 domains (physical functioning, role−physical, bodily pain), but the remaining 5 domains were no different from the QOL for the general population. This replicates research that has found that persons with SCI can have levels of QOL similar to able-bodied members of society,2 suggesting that certain factors may act to lower QOL in persons with SCI. Clearly, low self-efficacy can negatively impact QOL. Therefore, if this is the case, and research is suggesting it is, rehabilitation strategies should be designed to counter the influence of low self-efficacy. This would involve using cognitive behavior therapy, for instance, to challenge negative attitudes such as “I cannot deal with problems in my life” and “My life is ruined by not being able to avoid having bowel accidents.” If not challenged and changed, these types of self-statements or beliefs may result in increased risks of depressive mood and anxiety.1 Although physical and medical dimensions of SCI often attract most attention in the early stages of rehabilitation, it is becoming clearer that attitudinal issues are at least equally important in contributing to life satisfaction and well-being in the long-term. It would be a critical mistake to believe that a traumatic injury like SCI will necessarily result in low QOL just because of the onset of severe physical impairment.

Pain intensity seems also to be an influential factor in reducing life quality as shown in figure 3. Those with high pain intensity were found to have significantly lower QOL than those with lower levels of pain intensity for all the SF-36 domains other than physical functioning. Given that the pain intensity data was based on 1 item (Q.7) from the SF-36 scale itself, this finding will need replication using an independent pain measure. However, others have also found that pain amplifies the decrease in QOL associated with SCI.8, 26 The question may arise whether self-efficacy is independent of the pain measure, that is, is it just a surrogate measure of pain? This is unlikely given the low strength of the association found between the 2 variables shown in table 3. One can assume then that both self-efficacy and pain are independent factors, and this raises some pertinent issues. Inspecting the data in table 3 suggests that it is likely that if someone has low self-efficacy then they have a higher risk of experiencing debilitating pain (P=.67). It is of course not possible to argue causally; however, the odds are high (2.7 times) that a person with SCI possessing low self-efficacy will have a problem with pain (or perhaps vice versa). This may have important therapeutic implications as evidence emerges to suggest that cognitive-behavioral interventions are feasible to run and likely to be effective for individuals with SCI,47, 48 in similar fashion to other chronic pain conditions, where higher pain self-efficacy beliefs have been shown to predict reduced pain and avoidance behaviors.49 People who believe that they have no ability to bring about their desired outcomes in life (be it vocational, social, behavioral, physical, and so on) due to negative thinking, may place themselves at a higher risk of negative medical outcomes such as pain, or negative mood (depression) and chronic anxiety. In turn, this may lead to a decrease in social participation, which is known to reduce life satisfaction.40, 41 Certainly, the connections between all these factors need to be further explored with greater consideration given to them during the rehabilitation process.

A further point needs to be made. The negative impact on QOL and life satisfaction may well be cumulative. The combination of the 2 negative factors “low self-efficacy” and “pain intensity” was associated with an even greater reduction in QOL than the reduction in QOL associated with each of the 2 factors independently. This reduction was seen in all SF-36 domains except for physical functioning. This is a concerning finding and needs to be addressed in rehabilitation and in community-based interventions for people with SCI.

Conclusions 

Future studies will need to control for major variables such as age, completeness, time since injury, and so on, so that the influence of pain and self-efficacy is delineated.