The first aim of the study was to investigate the extent to which pre-dialysis patients experience feelings of autonomy and self-esteem, and participate in the work domain. Secondly, we wished to explore the content of patients' illness and treatment perceptions, and whether these perceptions are related to patients' perceived autonomy, state self-esteem and labor participation.
The mean age of the study group (64 years) and the gender distribution (64% male) corresponds with pre-dialysis patients and patients starting dialysis in the Netherlands [33, 34]. By comparing the mean scores on the autonomy measures of the total group with the answer scale, the results indicate that patients feel less autonomous because of their health condition or otherwise. In spite of this, most patients reported a high level of self-esteem. The autonomy and self-esteem levels of the pre-dialysis patients are slightly higher than the reported levels by patients on dialysis .
Looking at the mean illness and treatment perceptions of pre-dialysis patients it is noticed that patients are quite worried about their illness (M = 6.9) and believe that they themselves have rather little control over their illness (M = 4.7). To compare, patients on dialysis reported mean levels of 6.3 on the 'concern' dimension and 4.9 on the 'personal control' dimension . In a study of Broadbent et al. people with diabetes and people with asthma reported higher mean levels of personal control (M = 6.7) . Feelings of personal control are important for dialysis patients' quality of life [35, 9]. Personal control over the illness refers to the feeling that one can influence the course of the illness and one can fit the disease and treatment into daily life. In order to manage their illness pre-dialysis patients obviously are dependent on treatment. However, this does not indicate that there are no possibilities for personal control. It is of great importance that pre-dialysis patients practice self care behaviors, such as following diets and performing daily exercise in order to optimise their health condition . However, patients in this stage of the illness got the news that they have to start with renal replacement therapy in the near future, which indicates that despite of their self care activities they apparently were not able to remain sufficient renal function. This knowledge might have a negative effect on patients' personal control beliefs.
On the whole, the correlation analyses demonstrated that as patients hold more positive beliefs about their illness and their current treatment, they perceive more autonomy (both global and health related) and have a higher self-esteem. In light of these findings it is important to point out the difference between the construct of personal control and the construct of autonomy, since autonomy is often incorrectly equated with ideas of internal locus of control [37, 38]. Beliefs of personal control reflect individuals' beliefs regarding the extent to which one feels that one can control or influence an outcome, for example one's illness. However, people are autonomous when they act in accord with their authentic interests or integrated values and desires [17, 37–39]. To make the distinction more explicit, a person can experience control over carrying out a walking program, but not feel intrinsically motivated, and thus do not act in accordance with his/her own values.
The regression analyses revealed that the illness and treatment perceptions explain a substantial amount of variance in predicting both global autonomy and state self-esteem after controlling for background characteristics. These results illustrate that less perceived disruption by the treatment upon life is a significant predictor of state self-esteem. The findings furthermore suggest that less perceived impact of the treatment upon life is an important determinant of global autonomy as well. Treatment in the pre-dialysis phase in most cases includes taking pharmacotherapy and following a diet. Although these treatments are far less disruptive than dialysis treatment, the findings show that treatment already is a significant theme in this stage of the illness. Illness representations are considered to be constantly updated as new experiences and knowledge are acquired . In this transition phase of treatments, in which patients receive information on all available renal replacement therapies, it therefore can be expected that patients are more occupied with treatment in general, both their current treatment as well as their future treatment.
It should be noted that a large amount of variance remained unexplained. This indicates that other factors are of influence as well, for example the extent to which people in the patient's close environment, like the patient's partner or care providers, support the patient. Moreover, health related perceived autonomy could not be predicted by the illness perceptions and treatment perceptions. An explanation for this finding might be that patients are inclined to interpret 'health' as 'physical health'. At this stage of the illness, the renal disease - in most cases - will however not be associated with severe physical symptoms, which is also reflected by the mean score on the 'identity' dimension.
Because of the relatively old age of the study group (M = 64 years), only 45 patients (42%) were of working age (18-64 years). Fifty-one percent of the patients aged between 18 and 64 years performed paid work for at least 12 h per week, which is a higher percentage compared to dialysis patients; 24% , though considerably lower than that of the general Dutch population between the ages of 15-64 years; 65% . Thus, as suggested by Van Manen et al. , drop out of the labor market already occurs before patients start with dialysis treatment. Furthermore, the results show that, despite of their health condition, patients of working age place relatively high importance on carrying out a paid job. These findings point to importance and necessity of work related assistance in an early stage of the illness process. We wish to mention here that the average age of the working age group (18-64 years) was rather high (50 years) and 53% of the working age group was 50 years or older. To put this into perspective, in 2006, 32% of the Dutch people aged 20-64 years were 50-64 years . Notwithstanding that, our results suggest that labour participation in pre-dialysis patients is indeed lower than in the general population. Because of the small number of patients aged 18-64 years, we could not investigate the relationships between patients' perceptions of their illness and current treatment on the one hand and employment on the other hand more thoroughly. The findings, however, do show some trends: employed patients perceive their treatment as less disruptive and their illness as better controllable by self care and medical care than unemployed patients.
A limitation of this study is the replacement of the missing values on the variable comorbidity with the mean value of the total study group. Mean substitution preserves the mean of a variable's distribution; however, mean substitution typically distorts other characteristics of a variable's distribution (i.e., variance, median) . In spite of this we decided to substitute the missing values by the mean in order to be able to make maximal use of the data of all our cases. Another issue to note is the use of single-item measures in order to minimize the burden on respondents. Single-item measures are sometimes seen as less psychometrically sound than multiple-items. However, several studies show that single-item measures and their multiple counterparts are comparable [42, 43]. Moreover, Gardner et al.  demonstrate that a well-developed single item measure can be appropriate in avoiding common methods variance, which is often a problem with psychological measures that require respondent self-reports of attitudes, beliefs, perceptions, and the like. Furthermore, it is noteworthy that the present study had a cross-sectional design which means that no conclusions can be drawn regarding the causality of the observed relationships. Notwithstanding this limitation, our results suggest that the beliefs pre-dialysis patients hold about their illness and treatment are important factors for patients' sense of (global) autonomy and self-esteem. Finally, it should be noted that the study sample was rather small (N = 109), as well as the working age sample (N = 45). Consequently there was little statistical power to demonstrate relationships between perceptions and labor participation in particular. Future research should take this issue into account. It would be worthwhile to investigate these relationships once again in a larger sample of pre-dialysis patients.