need short discussion
stevensmary
discussion_2_article.pdf
Psychometric assessment of the Life Satisfaction Questionnaire (LSQ) and a comparison of a randomised sample of Swedish women and those su�ering from breast cancer
Marianne Carlsson 1 , Elisabeth Hamrin
2 & Ragny Lindqvist
1
1Department of Public Health and Caring Sciences, Caring Sciences Unit, Uppsala University, Sweden; 2Department of
Medicine and Care, Division of Nursing Science, University of LinkoÈping, Sweden
Accepted in revised form 29 August 1998
Abstract. In a previous study, the Life Satisfaction Questionnaire (LSQ) was developed especially for women with breast cancer. The aim of the present study was to assess the psychometric properties of the LSQ in a randomized sample of Swedish women and to compare the perceived quality of life with that of women su�ering from breast cancer. Another aim was to relate the result to educational background. A total of 257 women aged 20±80, randomly selected from the Swedish population register, answered the LSQ. The result was compared with that from 362 women aged 27±78 su�ering from breast cancer. The construct
validity was calculated by a principal component analysis, and the reliability by Cronbach a-coe�- cients. It was concluded that the LSQ has acceptable validity and reliability. The result also showed that the women with breast cancer rated their perceived qual- ity of life higher than women in general with respect to the quality of personal relations and the quality of daily activities, but lower for physical symptoms. In the comparison with respect to educational back- ground, women with university education rated their quality of life higher than women with other types of education. The testing of the LSQ will continue.
Key words: Breast cancer, Psychometric assessment, Quality of life
Introduction
The term `quality of life' refers to a person's perceived quality of her/his physical, psychological, social and existential functioning [1]. Several instruments have been developed to measure quality of life in di�erent groups of people. Some measure quality of life in general, e.g. the NHP (Nottingham Health Pro®le) [2], the SIP (Sickness Impact Pro®le) [3], the Swed- Qual [4], the MOS SF-36 (Medical Outcome Studies Short Form) [5, 6], and the QLI (Quality of Life In- dex) by Ferrans and Powers [7]. Others are more speci®c, for instance, for cancer patients, the QLI (Quality of Life Index) by Spitzer et al. [8], the FLIC (Functional Living Index Cancer) [9], the CIPS (Cancer Inventory of Problem Scale) [10], the CARES (Cancer Rehabilitation Evaluation System) [11], the FACT (Functional Assessment of Cancer Therapy) [12], and the EORTC QLQ-C30 (European Organisation for Research and Treatment in Cancer, Quality of Life Questionnaire) [13].
A literature review by Carlsson and Hamrin [14] showed that the diagnosis of breast cancer evokes grief, anger and intense fear in women. But the ma- jority do not experience long-term emotional distress. Risk factors for psychological distress seemed to in- clude many medical symptoms, little or no social support from family and signi®cant others, little or
no hope of a recovery, a desire to give up, and great concern about death. There were also some studies where the outcome failed to prove any relationship between breast cancer and psychosocial variables.
In a previous study, the LSQ-32 (Life Satisfaction Questionnaire) was developed to assess the quality of life of Swedish women with breast cancer [15]. The instrument was developed by use of a principal component analysis [16]. The results showed that women with breast cancer estimated their quality of life as quite positive, especially concerning the quality of family relations and everyday life. Since there were no data available for women in general, no certain conclusion could be drawn. On the other hand, there are studies of di�erent factors of importance for women to normalize their life situation after a breast cancer diagnosis.
According to Omne-Ponte n [17], living with a spouse seemed to protect women from developing psychosocial problems post surgery. And several studies [18±20] have shown that social support is an important factor for psychosocial adjustment and for survival for women with breast cancer. Di�erent sources of support have been studied (partner, family, friends, relatives, and medical professionals), which all turned out to be important in di�erent ways.
LaFortune Fredette [21] interviewed 14 women who had lived at least 5 years after a diagnosis of
Quality of Life Research 8: 245±253, 1999. Ó 1999 Kluwer Academic Publishers. Printed in the Netherlands.
breast cancer. The women described how they tried to adjust to the new situation by using multiple coping strategies, with information seeking, work, spiritual- ity, and family being predominant.
In a study by Hilton [22], 55 women with newly diagnosed early stage breast cancer and their families were interviewed in order to delineate how they coped with the illness. The result showed that normalization was an important process for these families, staying normal and getting back to normal. One important strategy, among others, was the reframing of nega- tives to be more positive. One outcome of this was satisfaction with the family state.
Collins et al. [23] investigated how the perception of life satisfaction changed after a cancer diagnosis. Cancer patients were asked 3 years after their diag- nosis how their life values had changed. The results showed that their attitude to the outside world and the future had changed negatively, whereas their evaluation of daily activities and personal relations had changed in a positive direction.
The purpose of a study by Moch [24] was to ex- plicate the concept of health within illness by de- scribing women's experience of breast cancer. The result showed that three di�erent themes were im- portant, namely changing relatedness to others, identifying meaning in the experience, and adding new perspectives about life.
O'Connor et al. [25] interviewed 30 patients in or- der to identify the process of personal search for meaning conducted by patients who had been re- cently diagnosed with cancer. The result showed that many respondents re¯ected on their relationships with other people and described them as ``closer'' as a result of the cancer experience. Furthermore, some of the respondents also described a general change in attitude towards life as a result.
In a study by Mackenbach et al. [26] a comparison was made between 11 countries in western Europe (Denmark, England/Wales, Finland, France, Ger- many, Italy, Netherlands, Norway, Spain, Sweden, and Switzerland). Self-reported morbidity by level of education, occupational class and/or level of income was obtained for 1985 to 1992. Inequalities in health were found in all countries. A relative index of in- equality was calculated and the result showed that, for women, the inequality of perceived health by education was greatest in Sweden compared with the other 11 countries.
In conclusion, there are studies indicating that a serious diagnosis like cancer can make people change their life values and life styles. Di�erent coping strategies could be used in order to adjust to the disease and try to normalize the life situation. Several studies have also shown that social support and a close relationship with others are important factors for psychosocial adjustment. In addition, education has been shown to be an important factor for the perception of health.
The aims of the present study became twofold. One aim was to assess the psychometric properties of the LSQ in a randomized sample of Swedish women and to compare the perceived quality of life with that of women su�ering from breast cancer. Another aim was to relate the perceived quality of life to educa- tional background.
Research questions
1. What factors are identi®ed in the LSQ in a ran- domised sample of Swedish women and what are the internal consistency-coe�cients (Cronbach a) of the factors?
2. Is the factor structure of the LSQ the same in the randomised sample of Swedish women (RSW) as in the sample of women su�ering from breast cancer (BCW)?
3. Are there any di�erences in perceived quality of life between the two samples, RSW and BCW?
4. Are there any di�erences in perceived quality of life in Swedish women with di�erent levels of education?
Method
The instrument
The questionnaire originally consisted of 32 items (version I) [15]; in the present version 34 items (ver- sion II) were to be answered on a 7-point Likert-type scale [27]. The questions are formulated in the fol- lowing way. Example (1): ``How much have you been troubled by tiredness during the last week?''. The answer should be on a 7-point scale from (1) ``very much'' to (7) ``not at all''. Example (2): ``How inter- esting do you think your activities during the last week have been?''. The answers could vary from (1) ``not interesting at all'' to (7) ``very interesting''.
In the ®rst test of the instrument (LSQ-32), 560 women with breast cancer in all stages were asked to participate. All were registered in a regional cancer register in south-eastern Sweden. They constituted all women from two of the hospitals, and comprised about 12% of all women in the regional cancer reg- ister. They were judged by a chief physician not to be terminally ill and to be able to answer the questions. (The response rate was 74%, n = 414, but only 362 of the questionnaires were considered su�ciently answered. Of those, however, 97 had left out single items, which were calculated by use of mean substi- tution. Most women (90%) were in an early stage of the disease and only 10% had cancer in an advanced stage. The women were between 27 and 76 years of age (M = 60.7, SD = 10.23). Most of them (76%) were married or lived with a partner; 24% lived alone. Their educational background varied from compulsory school 68%, secondary school 17%, and university education 15%.
246
In the version of the questionnaire (LSQ-34) used in the present study, the item ``necessity of activities'' was excluded due to comments from the respondents. (The respondents meant that it was di�cult to tell the necessity of an activity.) The following three items were added: ``perceived overall health'', ``creativity in activities'', and ``meaningfulness of activities''. In the study by Carlsson and Hamrin [15], the women with breast cancer also ®lled out the EORTC QLQ-C30 [13]. In the present study, the answers to the question of perceived physical health in EORTC QLQ-C30 were used in order to make a comparison possible.
Subjects
Six hundred women aged 20±80 were randomly se- lected from the Swedish population register (Semo Group Info Data AB, 1995). The questionnaire, LSQ-34, was mailed to the women and two reminders were sent. The second reminder consisted only of demographic data.
A total of 383 women returned the questionnaires. Of those, only 111 answered the demographic ques- tionnaire and 15 returned the questionnaire unan- swered. Seventeen questionnaires were returned with address unknown and 200 women did not answer at all. This gives a response rate of 66%, of which 19% only answered the demographic questionnaire. Two hundred and ®fty-seven women answered the two questionnaires completely.
The average age of these 257 women was 46.1 and the standard deviation was 16.1 years. Most of them (70%) were married or lived with a partner; 30% lived alone. Their educational background varied from compulsory school 28%, secondary school 38%, and university education 34%.
A comparison between the sample studied and o�cial statistics [28] was made regarding age, country of origin and level of education. This was done in order to check how representative the ®nal sample was. The comparison is shown in Table 1.
The work rate of the study sample was 76% and, according to o�cial statistics, the comparable ®gure is 74% [28].
The women in the present study were considered representative with regard to age, country of origin, and work rate. Concerning level of education, the women in the present study had university education to a somewhat higher extent.
Procedure
In order to investigate the construct validity of the LSQ, the data from the 257 women was factor ana- lysed, as a principal component analysis with varimax rotation [16]. The analysis was interpreted in the following way: for every item the highest factor loading was identi®ed and the item was allocated to that factor. Only factor loadings > 0:40 were re-
garded as signi®cant. Since there were some items with high loadings in two factors (items 3, 9, and 34), a second-order factor analysis was performed as well. The result from that analysis was compared with the result from that of women su�ering from breast cancer [15]. Other statistical procedures were de- scriptive statistics (mean values and standard devia- tions), and, for comparisons, Student's one group t- test, v2 analyses, and analyses of variance (ANOVA). For post hoc tests, Fisher PLSD (p < 0:05), was used. Cronbach alpha coe�cients [29] were used to test the reliability. The StatView [30] computerized program was used.
Results
The factor structure of the LSQ-34 in a randomized sample of Swedish women (RSW)
The result of the factor analysis is shown in Table 2.
The analysis gave six factors, which provided a logical explanation of 60% of the variance. The fac- tors were called: ``Quality of close friend relation (Factor I)'', ``Physical symptoms (Factor II)'', ``Quality of everyday life (Factor III)'', ``Quality of family relation (Factor IV)'', ``Socio-economic situ- ation (Factor V)'', and ``Sickness impact (Factor- VI)''. The internal consistency (reliability) coe�cients are shown in Table 4.
The factor structure of the LSQ in the RSW-group compared with the factor structure in the BCW-group
The factor analysis with LSQ-32 (the BCW-group) also gave six factors explaining 64% of the variance,
Table 1. Comparison in percentage between the study sample (n = 257) and Swedish o�cial statistics
Study
sample
O�cial
statistics
Age 18±24 8 11 25±44 37 34 45±64 32 30
65± 23 25
Country of origin Sweden 91 89 Other
countries
9 11
Level of education Compulsory
school
28 31 a
Secondary school
38 44
University 34 22
a Three percent were coded as missing value concerning
level of education.
247
Table 2. Factor loadings and communalities of the six factors solution of the LSQ-34 for women in the RSW-group (n = 257)
Item Factor I
Quality of close friend relation
Factor II
Physical symptoms
Factor III
Quality of everyday life
Factor IV
Quality of family relation
Factor V
Socio- economic situation
Factor VI
Sickness impact
Communality
1. Tiredness )0.105 0.235 0.006 0.131 0.381 0.512 0.491 2. Un®t )0.026 0.122 0.216 0.186 0.053 0.732 0.635 3. Di�culty sleeping 0.021 0.417 0.068 0.164 0.028 0.473 0.430 4. Loss of appetite )0.44 0.575 0.076 0.099 0.266 0.109 0.387 5. Diarrhoea )0.089 0.562 )0.087 0.159 )0.085 0.274 0.439 6. Constipation 0.187 0.508 0.117 )0.002 0.145 0.199 0.367 7. Dizziness 0.002 0.718 0.147 )0.042 0.124 0.156 0.678 8. Palpitation of the
heart
0.088 0.582 )0.036 0.261 )0.087 0.337 0.536
9. Di�culty breathing 0.239 0.407 )0.133 )0.042 )0.041 0.456 0.452 10. Muscular weakness 0.114 0.210 0.135 0.034 )0.56 0.751 0.643 11. Pain 0.148 0.198 0.205 )0.056 0.033 0.669 0.555 12. Nausea 0.076 0.725 0.034 )0.029 0.141 0.103 0.564 13. Perceived health 0.142 0.421 0.105 0.058 0.208 0.661 0.692
14. Satisfaction with life situation
0.173 0.303 0.263 0.135 0.522 0.134 0.500
15. Satisfaction with
economic situation
)0.097 0.019 0.099 0.112 0.705 0.118 0.543
16. Satisfaction with housing conditions
0.108 0.319 )0.020 0.124 0.541 )0.127 0.438
17. Level of activity )0.039 0.148 0.553 0.245 0.183 0.261 0.491 18. Activity: pleasant 0.281 0.056 0.664 0.156 0.303 0.232 0.708 19. Activity: interesting 0.236 )0.078 0.712 0.132 0.247 0.138 0.666 20. Activity: creative 0.104 0.031 0.712 )0.029 0.211 )0.053 0.566 21. Activity:
independent 0.146 0.144 0.696 0.167 )0.261 0.056 0.625
22. Activity: useful 0.092 0.014 0.772 0.225 )0.80 0.131 0.679 23. Activity:
meaningful 0.161 0.051 0.809 0.243 0.007 0.104 0.754
24. Family relation: emotionally satisfying
0.188 0.171 0.141 0.795 0.150 )0.014 0.739
25. Family relation: pleasant
0.279 0.014 0.099 0.815 0.166 0.096 0.789
26. Family relation:
meaningful
0.229 0.088 0.267 0.780 0.071 0.046 0.748
27. Family relation: independent
0.260 0.089 0.300 0.580 )0.033 0.086 0.511
28. Family relation: full of variations
0.231 0.018 0.194 0.708 0.143 0.167 0.640
29. Close friend relation:
emotionally satisfying
0.816 0.039 0.051 0.203 0.095 0.023 0.721
30. Close friend relation: pleasant
0.808 0.009 0.143 0.252 0.065 0.070 0.746
31. Close friend relation:
how meaningful
0.847 0.083 0.146 0.235 )0.005 0.078 0.807
32. Close friend relation: independent
0.715 0.094 0.266 0.141 )0.097 0.070 0.624
33. Close friend relation: full of variations
0.732 0.020 0.188 0.212 0.043 0.169 0.647
34. Overall quality
of life
0.165 0.072 0.324 0.361 0.433 0.402 0.617
The items ®nally included in respective factors are in bold type.
248
and the reliability was calculated as the alpha-coe�- cients for the di�erent factors [15]. (Table 4). The eigenvalues and variance proportions for the six factor-solutions after the varimax rotation, in the BRC-group and the RSW-group, respectively, are shown in Table 3. The factor ``Physical symptoms'', explained 10% of the variance in the RSW-group and 16% in the BRC-group. The variance proportions of the other factors were about equal.
All items except three (numbers 3, 9, and 10) loaded in the same factor in the two analyses. Item number 3 (di�culty sleeping), number 9 (di�culty breathing), and number 10 (muscular weakness) loaded in the factor called ``Sickness impact'' in the RSW-group and in the factor called ``Physical symptoms'' in the BCW-group. In Table 5 the factor loadings, com- munalities, eigenvalues, and variance proportions of the second-order factor analysis for the RSW-group is presented. The second-order factor analysis in the RSW-group gave three factors that provided logical explanations of 79% of the variance. The factors were ``Personal relation factor'' (originally factors `Quality of family relation' and `Quality of close friend rela- tion'), ``Physical factor'' (originally factors `Physical symptoms', and `Sickness impact'), and ``Socio-eco- nomic factor'' (originally factors `Quality of daily activities', and `Socio-economic situation').
The second-order factor analysis for LSQ-32 in the BCW-group gave three factors that provided logical explanations of 81% of the variance. The factors were ``Existential factor'' (orignally factors `Quality of family relation', `Quality of daily activities', and `Quality of close friend relation') ``Physical factor'' (originally factors `Physical symptoms', and `Sickness impact') and ``Socio-economic factor'' (originally factor `Socio-economic situation') [15].
The perceived quality of life in the RSW-group compared with that in the BCW-group
A comparison showed that there were di�erences between the women in the RSW-group and those in
the BCW-group in all six factors measuring di�erent aspects of quality of life. (The maximum point for each factor was recalculated to 100). In ®ve of the six factors the women in the BCW-group scored higher, whereas in one factor (Physical symptoms) the women in the RSW-group scored higher. (see Table 6).
In Table 7, the BRC-group is divided into three groups according to the stage of the disease. In the ®rst group, 201 women in early stage of the disease were included (stages CIS and I). In the second group, 123 women with breast cancer in stage II were included, and in the last group, 38 women with breast cancer in advanced stage were included (stages III och IV).
Perceived quality of life in women with di�erent educational backgrounds
The comparison between women with di�erent edu- cational backgrounds regarding perceived quality of life showed signi®cant di�erences. Women with university education scored higher in several of the comparisons. See Tables 8 and 9.
For the women in the RSW-group, signi®cant di�erences were found in four of the factors. In the factor ``Quality of close friend relation'', women with university education scored higher than those with compulsory school education. In the factors ``Physi- cal symptoms'' ``Sickness impact'', and ``Quality of everyday life'' the women with university education scored higher than those with compulsory school education and those with secondary high school education.
For the women in the BCW-group signi®cant dif- ferences were found in four of the factors. In the factors ``Quality of family relation'' and ``Quality of everyday life'', the women with university education scored higher than those with compulsory school education. In the factors ``Physical symptoms'' and ``Socio-economic situation'' the women with univer- sity education scored higher than those with com- pulsory school education and those with secondary high school education.
Table 3. Eigenvalues and variance proportions in the six factors solution for the RSW-group (n = 257) and the BRC- group (n = 362), respectively, after varimax rotation
Factor Group
RSW BRC
Eigenvalue Variance proportion
Eigenvalue Variance proportion
I Quality of close friend relation 3.8257 0.112 3.5068 0.110
II Physical symptoms 3.2724 0.100 5.2748 0.164 III Quality of everyday life 4.2601 0.125 3.4465 0.108 IV Quality of family relation 3.5772 0.105 3.6301 0.113
V Socio-economic situation 1.9959 0.059 2.1335 0.067 VI Sickness impact 3.4293 0.101 2.5773 0.081 Total variance 0.601 0.643
249
Analysis of missing cases
The response rate for RSW was 45%. But 66% of the women answered a demographic questionnaire, and a comparison between the women who answered both the demographic questionnaire and the LSQ-34 and the women who only answered the demographic questionnaire showed some di�erences. The women who did not answer the LSQ-34 were older (M = 46.1, SD 16.28, for the women who answered both questionnaires, and 50.9, SD = 17.60, for the women who only answered the demographic ques- tionnaire) (t = )2.675; df = 351, p = 0.0078). They also di�ered signi®cantly regarding educational background. Of the women who answered both questionnaires, 28% had compulsory school, 39% secondary school, and 33% university education; among the women who only answered the demo- graphic questionnaire, 43% had compulsory school, 35% secondary school, and 22% university education (v2 � 10:318; df � 2; p � 0:0057�.
Discussion
In this study a new tool, the Life Satisfaction Ques- tionnaire, LSQ, which was developed especially for
women with breast cancer, has been tried on a ran- domised sample of Swedish women in order to compare with those with breast cancer. We also wanted to relate the result to educational back- ground.
The response rate for RSW was rather low (45%), which could be a problem for the external validity. The factor structure of the LSQ-34 (RSW) turned out to be almost the same as the factor structure for LSQ- 32 (BCW); both analyses provided a logical expla- nation with six factors. But there were signi®cant di�erences between the RSW-group and the BCW- group regarding their perceived quality of life. The women in the RSW-group estimated their quality of life to be lower than the women in the BCW-group. Finally, with both samples, the comparison regarding educational background showed that women with university education estimated their quality of life to be higher than did the women with other educational backgrounds.
Both ®rst-order factor analyses gave six logical factors explaining 60% and 64%, respectively, of the variance. All items except three loaded in the same factor in the two analyses. The items ``di�culty sleeping'', ``di�culty breathing'', ``muscular weak- ness'' loaded in ``Sickness impact'' in the RSW-group and in ``Physical symptoms'' in the BCW-group. However, the factors ``Sickness impact'' and ``Physi- cal symptoms'' turned out to be one factor in the second-order factor analysis.
The a-coe�cients for the six factors were accept- able for ®ve of the factors (0.75±0.92). The a-coe�- cients for the sixth factor (``Socio-economic situation'') were 0.62 and 0.69, respectively. This factor consisted of only four items, which might ex- plain the low value. Since the two factor analyses gave the same factors and the a-coe�cients were satisfactory, the construct validity as well as the re- liability of the LSQ for the two samples of women could be considered acceptable.
When the ®rst-order factors were considered, all di�erences but one indicated that the women in the RSW-group estimated their quality of life to be lower compared with the women in the BCW-group. The
Table 4. Cronbach a-coe�cients for the six factors of the LSQ-34 (RSW) and the LSQ-32 (BCW), respectively
Factor RSW (n =257) BCW (n = 362)
a a
Quality of close
friend relation
0.90 0.86
Physical symptoms 0.75 0.90 Quality of daily activities
0.92 0.83
Quality of family relation
0.87 0.88
Satisfaction with
socio-economic conditions
0.62 0.69
Sickness impact 0.85 0.83
Table 5. Factor loadings, communalities, eigenvalues, and variance proportions of the second-order factor analysis for
LSQ-34 (RSW) (n = 257)
Original factor Personal relations Physical function Daily living Communality
Quality of close friend relation 0.924 0.130 0.068 0.875 Physical symptoms 0.086 0.904 0.121 0.839
Quality of everyday life 0.447 0.076 0.682 0.670 Quality of family relation 0.684 0.123 0.476 0.710 Socio-economic situation 0.050 0.327 0.841 0.817
Sickness impact 0.148 0.842 0.240 0.789 Eigenvalue/magnitude 1.5532 1.7007 1.4758 Variance proportion 0.259 0.283 0.246 0.784
a
a cumulative variance proportion.
250
women in the RSW-group experienced fewer physical symptoms than the women in the BCW-group. On the other hand, however, the women in the RSW- group experienced more sickness impact than the women in the BCW-group. The study by Hilton [22] suggested that normalization, and di�erent ways to keep normal and to return to normal, was important for women with breast cancer. One way to do that was to reframe negative cognitions into more positive ones. The participants in that study were women with newly diagnosed early stage breast cancer, which
meant that they were not terminally ill. In the BCW- group, 90% had breast cancer in an early stage and less advanced stage of the disease and 10% in an advanced stage [15]. That could be one explanation for why the women in the BCW-group estimated their quality of life to be better than the women in the RSW-group; they behaved like the women in the study by Hilton [22] who tried to normalize and, through that process, reframed negative cognitions. The women in the BRC-group who su�ered from breast cancer in an advanced stage di�ered signi®-
Table 6. Comparisons according to means between RSW (n = 257) and BCW (n = 362) regarding di�erent aspects of perceived quality of life. (The maximum point for each factor is recalculated to 100). Statistics: Student's one-group t-test
Factor RSW BCW t-values p-values
n = 257 n = 362 Mean (SD) Mean (SD)
Quality of family relation 77.68 (15.07) 83.65 (13.75) )8.27 0.0001 Quality of close friend relation 74.31 (14.49) 82.93 (14.80) )11.082 0.0001 Physical symptoms 89.37 (11.06) 82.09 (13.17) 10.517 0.0001 Sickness impact 70.36 (16.98) 76.16 (15.55) )7.101 0.0001 Quality of everyday life 69.14 (14.93) 83.65 (13.75) )20.191 0.0001 Socio-economic situation 75.04 (14.29) 77.07 (12.77) )3.027 0.0026
Table 7. Comparisons between RSW (n = 257) and BCW, divided into three categories according to the stage of the disease, regarding di�erent aspects of perceived quality of life. The maximum point for each factor is recalculated to 100.
Statistics: ANOVA
Factor RSW (n = 257) Mean (SD)
BCW early stage (n = 201) Mean (SD)
p-value
BCW less advanced stage (n = 123)
Mean (SD) p-value
BCW advanced stage (n = 38) Mean (SD)
p-value
Quality of family relation 77.68 (15.07) 92.35 (15.23) 92.45 (11.91) 92.04 (10.46) p = 0.0001 p = 0.0001 p = 0.0001
Quality of close friend relation 74.31 (14.49) 83.80 (15.38) 82.70 (13.19) 79.11 (16.29) p = 0.0001 p = 0.0001 n.s.
Physical symptoms 89.37 (11.06) 80.78 (13.26) 83.50 (12.61) 83.09 (13.95) p = 0.0001 p = 0.0001 p = 0.0078
Sickness impact 70.36 (16.98) 76.57 (15.16) 76.00 (15.45) 74.56 (18.07) p = 0.0001 p = 0.0001 n.s.
Quality of everyday life 69.14 (14.93) 82.46 (13.60) 85.44 (13.26) 84.21 (15.66)
p = 0.0001 p = 0.0001 p = 0.0001 Socio-economic situation 75.04 (14.29) 77.16 (12.31) 77.50 (13.37) 75.38 (13.37)
p = 0.0154 p = 0.0479 n.s.
Table 8. Perceived quality of life in women in the RSW-group (n = 257) with di�erent levels of education. Statistics: ANOVA
Level of education Quality of
family relation
Quality of close
friend relation
Physical
symptoms
Sickness
impact
Quality of
everyday life
Socio-economic
situation M (SD) M (SD) M (SD) M (SD) M (SD) M (SD)
Compulsory school 75.89 (16.4) 70.77 (15.9) 88.37 (10.6) 68.38 (16.1) 65.17 (14.3) 75.53 (15.2) Secondary school 77.43 (14.9) 74.23 (14.5) 87.88 (12.3) 67.06 (17.6) 66.53 (14.4) 74.49 (13.7) University 79.72 (14.1) 77.30 (12.7) 91.84 (9.6) 75.70 (15.9) 75.34 (14.3) 76.90 (13.7) F-value 1.0924 4.0564 3.3968 6.916 12.5189 1.2027
p-value 0.337 0.0184 0.035 0.0012 0.0001 0.3021
251
cantly from the RSW-group in three factors; they su�ered from more physical symptoms and experi- enced higher quality in their everyday life and in the relationships with their families.
Other studies have shown that valuing relation- ships with other people and daily activities more [23± 25] could be coping strategies that might be possible explanations for the di�erences between the women in the RSW-group and the women in the BCW- group. That the women with cancer (BCW) estimated their quality of life regarding the quality of family relationships and of close friend relationships higher than women did in general supports the conclusions from other studies [17±20] that social support is an important factor for the psychosocial adjustment of women with breast cancer.
The second-order factor analyses for the RSW- group and the BCW-group did not turn out exactly the same. In both analyses the most logical solution contained three factors. The factor, ``Physical fac- tor''. was identical, but the other two di�ered. The ®rst-order factor, ``Quality of daily activities'', loaded in the ``Socio-economic factor'' for women in the RSW-group and in the ``Existential factor'' for women in the BCW-group. This is an interesting di�erence that could be explained with the help of the results from Moch [24] and O'Connor et al. [25], whose results showed that in the personal search for meaning in the cancer experience, new perspectives about life could be added and general changes in attitude towards life could develop. Possible expla- nations could be that women with cancer re-evaluate ordinary daily activities and that they perform acitivities they do not like to a lesser extent than before the cancer disease.
The result of the present study was in agreement with the results from the study by Mackenbach et al. [26]. Women with university education estimated their quality of life to be higher than women with compulsory school education and secondary educa- tion. This was true for the women in the RSW-group as well for those in the BCW-group. Why women with university education estimated their quality of life to be higher than other women is interesting, but in the present study, it is not possible to speculate about the reasons for that.
In conclusion, the result of the present study shows that the LSQ has acceptable reliability and validity. The factor structure of the LSQ-34 turned out to be almost the same as the factor structure for LSQ-32, both analyses providing a logical explanation with six factors. It was also concluded that it is possible to experience good quality of life in spite of having breast cancer. The women in the BCW-group even estimated their quality of life to be higher than the women in the RSW-group, possibly because the cancer diagnosis made them re-evaluate their life situation and reframe negative cognitions in order to normalize the experience. University education turned out to be positively correlated with high values of perceived quality of life. The testing of the LSQ will continue.
References
1. Anderson RT, Aaronson NK, Wilkin D. Critical re-
view of the international assessments of health-related quality of life. Qual Life Res 1993; 2: 369±395.
2. Wiklund I, Romanus B, Hunt SM. Self-assessed disability in patients with arthrosis of the hip joint.
Reliability of the Swedish version of the Nottingham Health Pro®le. Int Disabil Stud 1987; 10(4): 159± 163.
3. Bergner M, Bobbitt RA, Cartner WB, Gibson BS. The Sickness Impact Pro®le: Development and ®nal revi- sion of a health status measure. Med Care 1981; 19(8):
787±805. 4. Brorson B, Ifver J, Hays RD. The Swedish health-re-
lated quality of life survey (SWED-QUAL). Qual of Life Re 1993; 2: 33±45.
5. Ware Jr JE. SF-36: Manual and Interpretation Guide. Boston: Health Institute, New England Medical Cen- ter, 1993.
6. Sullivan M, Karlsson J, Ware Jr JE. The Swedish SF-36 Health Survey ± I. Evaluation of data quality, scaling assumptions, reliability and construct validity across
general populations in Sweden. Soc Sci Med 1995; 41(10): 1349±1358.
7. Ferrans CE, Powers MJ. Psychometric assessment of
the Quality of Life Index. Research in Nursing and Health, 1992; 15: 29±38.
8. Spitzer WO, Dobson AJ, Hall J et al. Measuring the quality of life of cancer patients. J Chron Dis 1981; 34:
585±597.
Table 9. Perceived quality of life in women in the BCW-group (n = 362) with di�erent levels of education. Statistics: ANOVA
Level of education Quality of
family relation
Quality of close
friend relation
Physical
symptoms
Sickness
impact
Quality of
everyday life
Socio-economic
situation M (SD) M (SD) M (SD) M (SD) M (SD) M (SD)
Compulsory school 82.38 (13.4) 83.18 (15.0) 80.99 (12.9) 75.15 (15.5) 82.38 (13.4) 75.21 (13.6) Secondary school 85.71 (14.3) 81.57 (14.7) 81.82 (13.9) 77.43 (16.7) 85.71 (14.3) 79.47 (13.6) University 87.06 (14.3) 83.34 (13.9) 87.32 (12.4) 79.31 (14.2) 87.06 (14.0) 82.73 (11.9) F-value 3.4811 0.313 5.3471 1.8656 3.4811 9.5026
p-value 0.0318 0.7314 0.0051 0.1563 0.0318 0.0001
252
9. Schipper H, Clinch J, McMurray A, Levitt M. Mea- suring the quality of life of cancer patients: The func-
tional living index ± Cancer: Development and validation. J Clin Oncol 1984; 2(5): 472±483.
10. Cosgarelli-Schag CA, Heinrich RL. Development of a comprehensive quality of life measurement tool:
CARES. Oncology 1990; 4: 135±138. 11. Cosgarelli-Schag CA, Heinrich RL, Ganz PA. Cancer
inventory of problem situations: An instrument for
assessing cancer patients' rehabilitation needs. J Psy- chosoc Oncol 1983; 14: 11.24.
12. Cella DF, Tulsky DS, Gray G et al. The Functional
Assessment of Cancer Therapy Scale: Development and validation of the general measure. J Clin Onc 1993; 3: 570±579.
13. Aaronson NK, Ahmedzai S, Bullinger M et al. for the EORTC Study Group on Quality of Life. The Euro- pean Organization for Research and Treatment of Cancer QLQ-C30: A Quality-of-life instrument for use
in international clinical trials in oncology. J Nat Cancer Inst 1993; 85(5): 365±375
14. Carlsson M, Hamrin E. Psychological and psychosocial
aspects of breast cancer and breast cancer treatment. Cancer Nursing 1994; 17(5): 418±428.
15. Carlsson M, Hamrin E. Measurement of quality of life
in women with breast cancer. Development of a Life Satisfaction Questionnaire (LSQ-32) and a comparison with the EORTC QLQ-C30. Quality of Life Research 1996; 5: 265±274
16. Gorsuch R. Factor Analysis. New Jersey: Lawrence Erlbaum Associates, 1983.
17. Omne-Ponte n M, Holmberg L, Burns T, Adami HO,
BergstroÈ m R. Determinants of the psychosocial out- come after operation for breast cancer. Results of a prospective comparative interview study following
mastectomy and breast conservation. Eur J Cancer 1992; 28A: 1062±1067.
18. Irvine D, Brown B, Crooks D, Roberts J, Browne G.
Psychological adjustment in women with breast cancer. Cancer 1991; 67: 1097±1117.
19. Neuling SJ, Winefelt HR. Social support and recovery after surgery for breast cancer: frequency and correlates
of supportive behaviours by family, friends, and sur- geon. Soc Sci Med 1988; 27: 385±392.
20. Waxler-Morrison N, Hislop TG, Mears B, Kan L. E�ects of social relationship on survival for women with breast cancer: a prospective study. Soc Sci Med 1991; 33: 177±183.
21. LaFortune Fredette S. Breast cancer survivors: con- cerns and coping. Cancer Nursing 1995; 18(1): 35±46.
22. Hilton A. Getting back to normal: The family experi-
ence during early stage breast cancer. Oncol Nurs Fo- rum, 1996; 4: 605±614.
23. Collins RL, Taylor SE, Skokan LA. A better world or
a shattered vision? Changes in life perspectives fol- lowing victimization. Social Cognition 1990; 8(3): 263± 285.
24. Moch SD. Health within the experience of breast can- cer. J Adv Nurs 1990; 15: 1426±1435.
25. O'Connor AP, Wicker CA, Germino BB. Under- standing the cancer patient's search for meaning.
Cancer Nursing 1990; 13(3): 167±175. 26. Mackenbach JP, Kunst AE, Cavelaars AE, Groenhof
F, Geurts JJ, and the EU Working Group on Socio-
economic inequalities in Health. Socioeconomic in- equalities in morbidity and mortality in western Europe. Lancet 1997; 349: 1655±1659.
27. Likert R. A technique for the measurement of atti- tudes. New York: Archives of Psychology, No 140, 1932.
28. Statistics Sweden, 1997. O�cial Swedish Statistics.
29. Cronbach LJ. Coe�cient alpha and the internal struc- ture of test, Psychometrika 1951; 16: 297±334.
30. Haycock K, Roth J, Gagnon J, Finzer W, Soper C,
Sager S, Rocco T, Bauer L. StatView the ultimate in- tegrated data analysis & presentation system. ABA- CUS Concepts, Inc, Berkeley, CA. 1992.
Address for correspondence: Marianne Carlsson, Ph. D,
Associate Professor, Department of Public Health and Caring Sciences, Caring Sciences Unit, Uppsala University, Sweden
Fax: 46 18 471 34 90
253
Copyright of Quality of Life Research is the property of Springer Science & Business Media B.V. and its
content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's
express written permission. However, users may print, download, or email articles for individual use.
