- Email: [email protected]
Quality of life in cystic fibrosis: the impact of gender, general health perceptions and disease severity
Journal of Cystic Fibrosis 2 (2003) 206–213
Quality of life in cystic fibrosis: the impact of gender, general health perceptions and disease severity L. Geea, J. Abbotta,*, S.P. Conwayb, C. Etheringtonb, A.K. Webbc a
Department of Nursing, Faculty of Health, University of Central Lancashire, Preston PR1 2HE, UK b Adult Cystic Fibrosis Unit, Seacroft Hospital, Leeds LS14 6UH, UK c Adult Cystic Fibrosis Unit, Wythenshawe Hospital, Manchester M23 9LT, UK Received 13 September 2002; accepted 2 June 2003
Abstract Background: Disease progression in cystic fibrosis (CF) is marked by deterioration across a number of physiological systems. In addition, there is evidence that females have a worse prognosis than males. The current work assesses the impact of both these factors on health related quality of life (HRQoL). Methods: Two hundred and twenty-three adolescents and adults completed the cystic fibrosis quality of life (CFQoL) questionnaire with a further 185 approached and not responding by non-completion of the questionnaire. The CFQoL is divided into nine domains: physical, social, treatment, chest symptoms, emotional functioning, concerns for the future, relationships, body image, and career. Measurement of objective clinical status included, body mass index (BMI), and percentage of predicted forced expiratory volume in one second (FEV1 ). General health perceptions (GHP) were also measured. Results: Patients were sub-divided by gender and disease severity (mild )70% FEV1 , moderate 40–69% and severe -40%). Factorial analysis of variance indicated significant main effects for FEV1 (Fs587.98, PF0.001) and BMI (Fs 17.29, PF0.001) as a function of disease severity. Post hoc tests revealed significant two-group differences for FEV1 and BMI between disease severity groups. No differences were observed for gender across FEV1 or BMI. Differences emerged across most CFQoL domains for disease severity, with the exception of concerns for the future, which was consistently low throughout. Gender differences emerged for chest symptoms, emotional functioning, concerns for the future, body image and career. With the exception of body image, females exhibited poorer HRQoL. Pearson correlations indicated that females’ perception of health was more closely related to clinical status than males. Conclusions: Disease severity has an impact on HRQoL in adolescents and adults with CF. Some differences emerged between males and females, with females generally reporting poorer HRQoL. Evidence indicated that males and females perceived their health status differently, with females having a more accurate perception of objective clinical health status. 䊚 2003 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved. Keywords: Health related quality of life; General health perceptions; Cystic fibrosis
1. Introduction The natural course of cystic fibrosis (CF) is one of progressive deterioration in health. As the disease proceeds, decrements in physical functioning become more apparent placing greater limitations on the individual. Treatments also become both more aggressive and invasive. It may be anticipated that in parallel with these deteriorations in health status and increased treatment regimens that decrements in health related quality of life (HRQoL) would occur. How these changes, which have *Corresponding author. E-mail address: [email protected] (J. Abbott).
largely been brought about by increased longevity affect the HRQoL of adults with CF needs to be assessed w1x. Regarding gender, there is evidence to suggest that females with CF have a slightly worse prognosis, with a median mortality rate of 4 years less than males until the age of 20 beyond which male and female mortality rates are comparable w2,3x. Demko and colleagues w2x found that chronic infection of Pseudomonas aeruginosa in females occurred on average 1.7 years before colonisation in males. This outcome, was also highlighted by Sawyer et al. w4x who noted that women have more serious respiratory involvement than males and a lower survival rate from adolescence onwards. In the general
1569-1993/03/$30.00 䊚 2003 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved. doi:10.1016/S1569-1993Ž03.00093-6
L. Gee et al. / Journal of Cystic Fibrosis 2 (2003) 206–213
literature, there is evidence to suggest that men and women respond very differently to poor health w5–8x. For example, women have been observed to report higher levels of physical disease, greater pain and more subjective or emotional symptoms than men. Gender differences appear to reflect the way men and women respond to and perceive their health status rather than objective disease processes w5,7,9x. This suggests that men and women would differ in their reporting of health and illness. This was demonstrated in a study, which highlighted that women were more likely to perceive physical symptoms and act on them w7x. These findings are important in relation to HRQoL assessment, which requires subjective responding by the patient. In the past, differences between gender and different levels of disease severity in CF have been considered w10,11x, however, these factors have not previously been assessed using a CF specific quality of life measure. The present study examined HRQoL as a function of both disease severity and gender using the Cystic Fibrosis Quality of Life (CFQoL1) questionnaire w12x. Because the CFQoL is a patient derived measure, it is likely to be more sensitive to differences in disease severity and between men and women w13x. Any differences that do emerge will have repercussions in relation to clinical approaches with these different groups. The study also examined whether there was any variation in the relationships between health perceptions, clinical indicators and HRQoL between males and females. 2. Method 2.1. Study design A cross sectional questionnaire study of patients currently attending two regional cystic fibrosis specialist centres was undertaken. 2.2. Subjects All patients attending the regional adult cystic fibrosis centres in Manchester and Leeds in the United Kingdom, were considered eligible for the study. Over a 6-month period, all patients attending the outpatient clinic at their respective centre were approached to participate. 2.3. Measures Objective measurement of disease status included percentage of predicted forced expiratory volume in one second (FEV1) and body mass index (BMI). HRQoL was measured using the CFQoL questionnaire w12x. The CFQoL questionnaire is a patient derived disease spe1
Formatted copies of the CFQoL questionnaire, scoring equations and supplementary general health perception items are available on request from the corresponding author as hard copies or via e-mail.
207
cific HRQoL measure that has been fully validated. It consists of 52 items across nine domains of functioning: physical functioning, social functioning, treatment issues, chest symptoms, emotional functioning, concerns for the future, interpersonal relationships, body image, career issues. Scores are converted into values ranging from 0 (worst possible quality of life) to 100 (best possible quality of life). Scores of 50 or less reflect negative responding to items whilst scores of above 50 reflect positive responding to items. Negative responding to items suggests that the individual is experiencing difficulties within a particular domain. Further examination of the items within the domain can highlight where these problems lie. Item content was determined by feedback from patients, literature reviews and consultation with the multidisciplinary teams. Domain structures were determined by two rounds of factor analysis to arrive at the final questionnaire format. Cronbach alpha coefficients for internal reliability of each domain were high ranging from 0.72 to 0.91. Concurrent validity was assessed by comparing similar domains between the Short Form-36 item health status questionnaire w14x (SF-36) and the CFQoL. Correlations were shown to be highly significant and in the moderate to strong range (rs0.64 to rs0.74). Test re-test values were also strong with correlations ranging from rs0.74 to 0.96. The CFQoL was also demonstrated to display sensitivity to improvements brought about in disease status by intravenous antibiotics w12x. As outlined in Section 1, gender variations in health reporting appear to be based in the differences between the health perceptions of men and women. Given this link, participants were also given two additional questions alongside the CFQoL, which asked them about their general health perceptions (GHP). These were: in general, how is your health? Compared to others with cystic fibrosis, do you think your health is« Each question was answered on a five point scale from 1s poor, to 5swell above average. 2.4. Procedure The CFQoL questionnaire was distributed to patients when they attended a routine outpatients appointment. Return of questionnaires was via postal means. This method was chosen to minimise disruption to patients at clinic and to the running of the clinics. It was also felt that this method would give the patients time to reflect on whether they wanted to participate in the study and would remove the pressure inherent in a faceto-face request. Clinical and demographic data for sex, age, FEV1 and BMI were routinely assessed for all patients on the day they attended clinic. A pre-paid envelope was included for all patients to return the questionnaires. Patients were instructed to return their questionnaires as soon as possible after the clinic
208
L. Gee et al. / Journal of Cystic Fibrosis 2 (2003) 206–213
Table 1 Patient characteristics data
Number in group(n) Males(n) Females(n) Age in years Total sample Male Female % FEV1 Total sample Male Female BMI Total group Male Female
Mild
Moderate
Severe
Total sample
60 31 29
97 45 51
66 26 40
223 102 121
23.5 (6.6) "1.71 24.4 (6.9) "2.56 22.6 (6.2) "2.36
24.8 (7.3) "1.47 24.9 (6.4) "1.95 24.7 (8.0) "2.23
27.0 (6.9) "1.69 27.6 (5.5) "2.24 26.6 (7.6) "2.45
25.1 (7.1) "0.93 25.6 (6.4) "1.27 24.6 (7.6) "1.36
86.7 (11.3) "3.10 85.5 (9.8) "3.61 88.0 (12.8) "4.88
54.2 (9.0) "1.83 55.3 (9.4) "2.82 53.2 (8.8) "2.45
29.3 (7.2) "1.77 29.4 (8.1) "3.28 29.3 (6.6) "2.13
55.0 (23.5) "3.10 58.0 (23) "4.5 54.0 (23.8) "4.3
22.0 (2.78) "0.71 22.5 (2.5) "0.92 21.6 (3.0) "1.14
21.0 (2.2) "0.46 21.1 (2.3) "0.70 20.8 (2.6) "0.63
19.5 (2.1) "0.52 19.8 (1.8) "0.73 19.3 (2.3) "0.74
20.8 (2.5) "0.34 21.1 (2.4) "0.48 20.5 (2.6) "0.47
With the exception of the first three variables, values represent mean ("S.D.) and "95% confidence intervals around the mean. nsnumber in group.% FEV1spercentage of predicted forced expiratory volume in 1 s. BMIsbody mass index.
appointment preferably within a few days. The date of the outpatient appointment was recorded and patients were asked to also record the date they completed the questionnaire. During the study period, if a non-responding patient had a repeat attendance at clinic a second questionnaire was given to the patient and the above procedure was observed. All questionnaires included in the study were returned within a week of distribution. Clinical status and demographic information were also assessed for questionnaire non-responders and a comparison was made between groups to determine how representative the observed samples were. Data collection was continuous for a period of 6 months. In total 408 patients were approached and given questionnaires to complete and return. All responding patients completed the questionnaires during a stable phase of their disease. 2.5. Statistical analysis Scores on the CFQoL questionnaire were converted into values that ranged from 0 to 100. Descriptive statistics for clinical, demographic and HRQoL scores were normally distributed and are represented as means, standard deviations and confidence intervals. Disease groups were defined as a function of FEV1 and divided into three groups w15x: mild disease (FEV1)70%), moderate disease (FEV1 41–69%) and severe disease
(FEV1-40%). This division of mild, moderate and severe disease has been applied and adopted internationally as a categorisation of disease severity in both the medical and psychosocial literature and has been used as the standard for the epidemiological study in CF (ESCF) w16–18x. Differences between HRQoL scores for males and females and between disease severity groups were analysed using factorial analysis of variance (ANOVA). Post hoc analyses were applied using Tukeys Honestly Significant Difference (HSD) test. Relationships between GHP, BMI, FEV1 and CFQoL domains were analysed using Pearson product moment correlations. Statistical analyses were conducted using SPSS for windows version 6.0 w19x. Effect sizes for significant differences were calculated (quoted as ‘d’ in Section 3). Effect sizes were calculated using D-STAT meta-analytic software w20x. Magnitude of effect sizes are based on those quoted by Cohen w21x as: small effect sizes0.20, medium effect sizes0.50 and a large effect sizes0.80. 3. Results 3.1. Descriptive statistics Of the initial 408 patients who were approached and given a questionnaire to complete, 223 patients completed and returned questionnaires. Descriptive statistics for patient characteristics are presented in Table 1 as a
L. Gee et al. / Journal of Cystic Fibrosis 2 (2003) 206–213
function of disease severity for males and females separately and for the overall population. Descriptive statistics for domains of the CFQoL questionnaire, and GHP for gender, disease severity and the total population are presented in Table 2. The present sample represented a response rate of 55% with a further 185 not responding by virtue of failure to return questionnaires. Data for this sub-set have been presented in detail elsewhere w22x and so a brief summary of the outcome will be outlined for the present study. Of the 185 non-responders full demographic and clinical data were available for 103. Whilst a disparity was observed between the number of male and female non-responders with a higher percentage of males by comparison to females (68% and 46%, respectively), no differences were found between responders and non-responders for age, percentage of predicted FEV1 or BMI. For the present data, comparisons were also made between males and females in each disease severity group. No significant differences emerged between either males or females across disease severity groups for age, FEV1 or BMI. 3.2. Differences between demographic and clinical indicators as a function of gender and disease severity Two-way factorial ANOVA indicated significant main effects for gender=disease severity across FEV1 (Fs 395.90, PF0.001), BMI (Fs13.37, PF0.001) and age (Fs2.98, Ps0.03). Significant 1-way main effects were demonstrated across FEV1 (Fs587.98, PF0.001), BMI (Fs17.29, PF0.001) and age (Fs4.18, Ps0.01) for disease severity. FEV1 and BMI deteriorated in parallel with health status, whilst age increased as disease status deteriorated. Tukey’s HSD post hoc test for pair-wise analysis of groups indicated differences for both FEV1 and BMI across all disease group comparisons with effect sizes ranging from moderate to large (ds0.66– 1.49). Tukey’s HSD indicated that age differences were also significant across all group comparisons, with effect sizes in the small to moderate range (ds0.18–0.51). No differences in FEV1, BMI or age were observed between males and females. 3.3. CFQoL questionnaire scores as a function of disease severity and gender Applying factorial ANOVA on a domain-by-domain basis indicated that two-way main effects for all CFQoL domains across disease severity=gender were significant. The results also demonstrated highly significant one-way main effects for disease severity across all domains. Post Hoc analysis using Tukey’s HSD indicated significant pair-wise differences between mild and moderate groups for body image (ds0.62), interpersonal relationships (ds0.45) and concerns for the future
209
Table 2 Mean health related quality of life scores, standard deviations and 95% confidence intervals for the total sample, gender and disease severity Mild
Moderate
Severe
Total sample
Number in group Males 31 Females 29
45 52
26 40
102 121
Physical functioning Males 94.38 (9.2) "3.4 Females 87.37 (16.5) "6.3
86.57 (14.6) 73.84 (16.9) "4.4 "6.8 82.96 (20) 74 (21.3) "5.6 "6.8
85.7 (15.7) "3.0 81 (20.2) "3.6
Social functioning Males 91.93 (13.6) "5 Females 85.51 (22.5) "8.6
88.55 (14.9) 78 (25.4) "4.5 "10.2 84.23 (22.5) 78.25 (25.3) "6.3 "8.1
86.9 (18.4) "3.7 82.5 (23.5) "4.2
Treatment issues Males 87.09 (13.4) "5 Females 76.55 (25.1) "6.92
73.48 (23) "7 72.69 (25.2) "7
77.6 (22.6) "4.4 71.7 (25.4) "4.5
Chest symptoms Males 86.12 (19.8) "7.2 Females 65.17 (29.5) "11.2
68.66 (26.5) 58.46 (24.1) 71.3 (26) "8.0 "9.7 "5.1 65.76 (24.7) 51.25 (25.4) 60.82 (26.7) "6.7 "8.1 "4.8
Emotional responses Males 89.43 (11.6) "4.25 Females 80.43 (23.2) "8.8
81.77 (15.1) 78.26 (16.1) 83.2 (14.9) "4.5 "6.5 "2.9 77.06 (17.6) 73.87 (19.1) 76.8 (19.5) "5 "6.1 "3.52
Concerns for the future Males 60 (25.4) "9.3 Females 42.9 (24.1) "9.1
42.07 (21.7) 44.48 (23.3) "6.5 "9.39 42.17 (24) 41.58 (20.4) "6.7 "6.52
Interpersonal relationships Males 68 (22.1) "8.1 Females 72.75 (19.7) "7.5
56.22 (24.3) 61.92 (19.5) 61.2 (22.8) "7.3 "7.9 "4.4 63.73 (20.4) 56.05 (20.3) 63.3 (20.9) "5.7 "6.5 "3.7
73.58 (27.7) "11.2 67.16 (25.7) "8.2
48.1 (24.3) "4.7 42.1 (22.7) "4.0
Body image Males 76.77 (22.8) 50.81 (20.7) 52.82 (21.3) 59.2 (24.2) "8.3 "6.2 "8.6 "4.7 Females 77.47 (22.3) 72.17 (25.1) 61.83 (23.3) 70 (24.4) "8.5 "7.0 "7.5 "4.4 Career issues Males 75.64 (27.8) "10.1 Females 61.20 (32.2) "12.2
56.22 (28.9) 60.76 (24.9) 63.2 (28.5) "8.7 "10 "5.6 60.76 (26.3) 47.25 (25.1) 56.4 (27.9) "7.3 "8 "5.0
General health perceptions Males 76.9 (17.4) 67 (18.4) "6.4 "5.5 Females 75.27 (17.2) 61.96 (17.3) "6.5 "4.9
56.38 (15.5) "6.3 44.6 (18.9) "6
67.3 (18.8) "3.7 59.4 (21.1) "3.8
With the exception of the first variable, values represent mean ("S.D). and "95% confidence intervals around the mean.
210
L. Gee et al. / Journal of Cystic Fibrosis 2 (2003) 206–213
(ds0.39). Differences between moderate and severe disease groups were found for physical functioning (ds 0.57), social functioning (ds0.36) and chest symptoms (ds0.51). Whilst for mild and severe disease comparisons, statistically significant differences were found for the majority of domains (effect size range ds0.37– 0.99), with the exception of concerns for the future. Means indicated lower scores for increased severity groups highlighting poorer HRQoL as the disease progresses. Significant differences between males and females were found for body image (Fs10.79, Ps0.001, ds 0.44), chest symptoms (Fs8.85, Ps0.003, ds0.39), career concerns (Fs4.22, Ps0.013, ds0.27), emotional responses (Fs6.3, Ps0.013, ds0.33) and concerns for the future (Fs4.34, Ps0.038, ds0.28). Significant two-way interactions as a function of gender=disease severity were found for body image (Fs4.0, Ps0.02) and career issues (Fs3.06, Ps0.049). Males experienced poorer body image by comparison to females with the lowest scores for males being observed in the moderate disease group. Tukey’s HSD test yielded a significant difference between males and females for body image in the moderate disease group (ds0.91). No other significant differences were observed for body image. For career issues, females experienced more concerns in the mild and severe groups. This trend reversed in the moderate groups where males demonstrated more concerns regarding career issues. Although a significant interaction was observed, no statistically significant differences were found for career issues within disease severity groups between males and females. However, a moderate effect size was observed between males and females in the severe disease group (ds0.53). Marginally significant two-way interactions were found for chest symptoms (Fs2.43, Ps0.09) and concerns for the future (Fs2.17, Ps0.06). Females generally reported poorer chest symptoms. This was particularly marked in the mild disease group within which there was a significant difference between males and females (ds0.82) for chest symptom scores. No other significant differences were observed for chest symptoms between males and females. The interaction for concerns for the future indicated that across disease severity groups, females reported a poorer quality of life. This was particularly pronounced in the mild group in which there was a statistically significant difference between males and females (ds0.68).
Table 3a (a) Correlations (r) between FEV1 and CFQoL, domains in descending order of shared variance as a function of gender
3.4. General health perceptions as a function of gender
BMI and CFQoL domains for both males and females. For males significant relationships existed between BMI and body image (rs0.34, Ps0.001) which explained 11.6% of common variance (r 2=100) and chest symptoms (rs0.21, Ps0.02) explaining only 4% of variance. In common with males, a significant relationship
Further analyses using Pearson product moment correlations examined the relationships between CFQoL domains, FEV1, BMI and GHP as a function of gender. Only two significant correlations were observed between
(r)
(P)
% variance
Females Emotional functioning Relationships Physical functioning Body image Chest symptoms Career issues Treatment issues Concerns for the future Social functioning
0.60 0.28 0.25 0.25 0.21 0.18 0.17 nys nys
0.001 0.002 0.005 0.005 0.02 0.03 0.05 nys nys
36 8 6.25 6.25 4 3.25 3 nys nys
Males Physical functioning Body image Chest symptoms Treatment issues Emotional functioning Social functioning Concerns for the future Career issues Relationships
0.50 0.41 0.38 0.27 0.27 0.26 0.22 0.22 nys
0.001 0.001 0.001 0.005 0.005 0.007 0.02 0.02 nys
25 16 14.5 7.3 7.3 7 5 5 nys
rscorrelation coefficient. Pssignificance level. % variancespercentage of common variance explained. nyssnot significant. Table 3b (b) Correlations (r) between GHP and CFQoL domains in descending order of shared variance as a function of gender (r)
(P)
% variance
Females Physical functioning Chest symptoms Relationships Career issues Body image Social functioning Emotional functioning Treatment issues Concerns for the future
0.52 0.51 0.50 0.48 0.39 0.39 0.36 0.34 0.25
0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001
27 26 25 23 15 15 13 11.5 6.25
Males Physical functioning Chest symptoms Body image Career issues Emotional functioning Relationships Social functioning Treatment issues Concerns for the future
0.58 0.51 0.47 0.41 0.40 0.40 0.39 0.37 0.25
0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001
34 26 22 17 16 16 15 14 6.25
rscorrelation coefficient. Pssignificance level. % variancespercentage of common variance explained. nyssnot significant.
L. Gee et al. / Journal of Cystic Fibrosis 2 (2003) 206–213
was observed for females between BMI and body image (rs0.55, Ps0.001), which was stronger than the correlation observed for males and explained 30% of variance. The second significant but weak relationship for females was between BMI and concerns for the future (rs0.20, Ps0.02) explaining only 4% of common variance. Correlations between FEV1 and CFQoL domains are presented in Table 3a. Correlations were generally weak, with differences emerging in the descending order of shared variance as a function of gender. For females, the strongest relationship between FEV1 and CFQoL domains was for emotional functioning (rs0.60, Ps 0.001) with 36% of common variance explained. For males, the strongest correlation was between FEV1 and physical functioning (rs0.50, Ps0.001) with 25% of common variance explained. For correlations between GHP and CFQoL domains (Table 3b), the descending order of relationships did not vary greatly between males and females. The only difference worthy of comment was between GHP and relationship issues with a stronger correlation between variables for females, explaining 25% of variance by comparison to 16% of variance for males. Correlations between GHP and clinical indicators revealed that females demonstrated stronger relationships for GHP with both FEV1 (rs0.59, Ps0.001, 34% of variance), and BMI (rs0.34, Ps0.001, 11% of variance) by comparison to males (FEV1yrs0.41, Ps0.00, 17% variance, BMI–correlation not significant). 4. Discussion The results suggest that adolescents and adults with CF experience a progressive deterioration in the majority of CFQoL domains as their disease severity increases. This was with the exception of scores across the ‘concerns for the future’ domain, which remained low irrespective of disease severity and evidenced only a small effect size difference between the mild and moderate disease groups, an outcome that was likely to have been influenced by the significantly higher scoring males in the mild group. This suggests that concerns for the future are of particular salience to adults with CF throughout their lives. Additionally, it would seem that on a domain-by-domain basis for the CFQoL, differences between mild and moderate and moderate and severe groups varied. This suggests that at various points in the disease trajectory different HRQoL issues may emerge as problematic. Overall, between the mild and severe ends of the disease spectrum there were significant differences across the majority of domains. This was once again with the exception of ‘concerns for the future’. These findings reinforce the belief that interventions aimed at improving and maintaining the HRQoL
211
of adults with CF are paramount and should have an increasing emphasis as the individual’s disease status deteriorates. In particular, the study reflects the need to address patients’ concerns for the future. Regarding overall differences in CFQoL domains between males and females, where differences existed, females reported poorer HRQoL for chest symptoms, emotional functioning, concerns for the future and career issues. Gillen et al. w26x found that work disability in adults with CF was common, but not universal. They also outlined that few respondents to their study had received any kind of career guidance or counselling. The outcome of the present study would indicate that this type of guidance, whilst important for both genders, might be of greater importance when considering support for females with CF. For body image, males reported a poorer HRQoL. This is interesting given that there was no difference between males and females for FEV1 or BMI, and with males demonstrating a trend towards higher BMI than females. The finding that males have a poorer body image than females is in keeping with previous research w27x. The study by Abbott et al. indicated that by comparison to non-CF controls CF females were happier with their body image whilst CF males desired to be heavier. CF males also perceived their body shape as heavier than it actually was. This was by comparison to non-CF males who were more accurate when judging body shape. The finding may reflect cultural stereotypes, which lead women to desire thin bodies whereas males desire heavier and more muscular bodies. The idea of misperception of BMI also fits in with the current study, which indicated no relationship for males between general health perceptions and BMI and a weak relationship between BMI and body image, which only explained 11.6% of variance. However, Walters w28x found that women in her study tended to overestimate their weight by comparison to males who underestimated their weight. The study also concluded that perception of self as underweight was an important factor in determining nutritional behaviour and those patients who perceived themselves as underweight were more likely to be adherent to nutritional interventions. The difference in outcome between this study and the Abbott et al. paper may be based in the type of methodology employed. Whilst the Walters paper used BMI and self-perception as underweightyoverweight, the Abbott et al. study asked respondents to gauge which out of a series of silhouettes developed to represent various BMI’s was more like their own. This highlights an important distinction between satisfaction with body weight and perception of body weight. Willis and colleagues w29x have indicated in a qualitative study of men and women with CF, that of primary importance to adults with CF is the construction of their identity as either young women and young men, rather
212
L. Gee et al. / Journal of Cystic Fibrosis 2 (2003) 206–213
than adults with CF. Identification as people living with CF was secondary to their identification as men and women. This study also demonstrated that part of this role for women entailed a degree of passivity to their situation alongside an emphasis on having a slender body shape. This coupled with the finding that there was a greater discordance with adherence for women by comparison to men may suggest that not only do females prefer their slender body shape, but rather than being passive they may engage in non-adherence behaviours to enhance this ideal. Collins and colleagues w30x, who suggest that poorer compliance to dietary factors in females may be associated with the lower median survival rate of females by comparison to males, further support this notion. Interactions between gender and disease severity indicated that across some domains females had a poorer quality of life as a function of disease severity. This was despite having similar health status as indicated by objective clinical measures. This is important to consider when approaching treatment in females who appear more susceptible to lower HRQoL than males with the same objective clinical status. It is possible that the differences in CFQoL domain scores between males and females, was based on their perceptions of health rather than their actual health status. This would be consistent with research that states males and females responses to ill health are often based in perceptions rather than actual disease status w5,7,9,31x. Few notable differences emerged when considering relationships between FEV1, BMI, and GHP with CFQoL domains for males and females. The strongest relationship for females was between FEV1 and emotional functioning whilst for males it was between FEV1 and physical functioning. This outcome suggests that FEV1 has a more profound impact on emotional functioning in females whilst for males the greatest impact is on physical functioning. For the present study, potential weaknesses lie with the possible bias that may be introduced via the nonresponders. The representativeness of the responders by comparison to the non-responders for this sample has been discussed in detail elsewhere w22x. For the present study, which is largely concerned with gender differences, the percentage of non-responders may have influenced the higher number of female participants in the data set. However, based on previous analysis w22x, the conclusion was arrived at that the individuals who participated in the study were representative of the wider population sampled. Therefore, observed disparity between males and females was unlikely to have unduly influenced the outcome although this potential exists in any study. Research has also indicated that the present levels of responding (55%) are typical in CF w23x and also that the percentage of males and females responding are comparable to other research into HRQoL w24,25x in CF.
When considering the relationships between GHP and clinical indicators, it emerged that females’ health perceptions were more strongly associated with both FEV1 and BMI. However, relationships between clinical indicators and GHP were far from perfect suggesting that variables other than those considered in the present study may be responsible for overall health perceptions in both males and females. The question remains, what is influencing health perceptions if it is not clinical status? The answer to this question is beyond the scope of the present study, but areas which may provide a focus for further research include coping styles w32,33x, health locus of control, self efficacy, optimismypessimism or social support. The study by Staab and colleagues indicated that subjective health perceptions and ways of coping explained a significant amount of variance in overall HRQoL, above that which was accounted for by disease severity and time spent on therapy. This indicates that both of these factors are important contributors to overall HRQoL. Further research should be aimed at establishing what factors are likely to be involved in HRQoL responses in adolescents and adults with CF. References w1x Abbott J, Gee L. Contemporary psychosocial issues in cystic fibrosis: treatment adherence and quality of life. Disabil Rehabil 1998;20:262 –71. w2x Demko CA, Byard PJ, Davis PB. Gender differences in cystic fibrosis: Pseudomonas aeruginosa infection. J Clin Epidemiol 1995;48:1041 –9. w3x Davis PB. The gender gap in cystic fibrosis survival. J Gender Specific Med 1999;2(2):47 –51. w4x Sawyer SM, Rosier MJ, Phelan PD, Bowes G. The self-image of adolescents with cystic fibrosis. J Adolescent Health 1995;16:204 –8. w5x Mechanic D. Sex illness behaviour and the use of health services. Social Sci Med 1978;128:207 –14. w6x Merrill SS, Seeman TE, Kasl SV, Berkman LF. Gender differences in the comparison of self reported disability and performance measures. J Gerontol: Med Sci 1997;52a:M19 –M26. w7x Piccinelli M, Simon G. Gender and cross-cultural differences in somatic symptoms in association with emotional distress: an international study in primary care. Psychol Med 1997;27:433 –44. w8x Unruh AM. Gender variations in clinical pain experience. Pain 1997;65:123 –67. w9x Ruiz MT, Verbrugge LM. A two way view of gender bias in medicine. J Epidemiol Commun Health 1997;51:106 –9. w10x Congleton J, Hodson ME, Duncan-Skingle F. Quality of life in adults with cystic fibrosis. Thorax 1996;51:936 –40. w11x Congleton J, Hodson ME, Duncan-Skingle F. Quality of life in adults with cystic fibrosis. Thorax 1997;52:397 –400. w12x Gee LM, Abbott J, Conway S, Etherington C, Webb AK. The development of a disease specific health related quality of life measure for adults and adolescents with cystic fibrosis. Thorax 2000;55:946 –54. w13x Abbott J, Gee L. Quality of life in children and adolescents with cystic fibrosis: implications for optimising treatments and clinical trial design. Paediatr Drugs 2002;5(1):41 –56.
L. Gee et al. / Journal of Cystic Fibrosis 2 (2003) 206–213 w14x Ware JE, Sherbourne CD. The MOS short form health survey (SF-36): 1. Conceptual framework and item selection. Med Care 1992;30:473 –83. w15x Fiel SB, FitzSimmons S, Schidlow D. Evolving demographics of cystic fibrosis. Semin Respir Crit Care Med 1994;15:349 – 55. w16x Connolly MA, Johnson JA, Brown N, Montgomery M, Zuberbuhler P. Development of a disease specific quality of life instrument for people with cystic fibrosis. Institute of Health Economics Working Paper 00-11 2000. w17x Quittner AL, Buu A, Watrous M, Davis MA, Cystic Fibrosis Questionnaire (CFQ): a health related quality of life measure. The Cystic Fibrosis Foundation 2000. w18x Britto MT, Kotagal UR, Hornung RW, Atherton HD, Tsevat J, Wilmott RW. Impact of recent pulmonary exacerbations on quality of life in patients with cystic fibrosis. Chest 2002;121:64 –74. w19x SPSS Inc. SPSS for Windows release 6.0, 1993; Chicago, IL: SPSS Inc. w20x Johnson, BT. DSTAT: Software for the Meta-analytic Review of Research Literatures (version 1), 1989; Lawrence Erlbaum Associates. w21x Cohen J. Statistical power analysis for the behavioural sciences. In: Howell DC. Statistical methods for psychology (4th Ed.). Duxbury: p. 215. w22x Gee L, Abbott J, Conway S, Etherington C, Webb AK. Validation of the SF-36 for the assessment of health related quality of life in adults with cystic fibrosis. J Cystic Fibrosis 2002;1:137 –45. w23x Walters S., Association of Cystic Fibrosis Adults (UK) Survey 1994: Analysis and Report 1994; Association of Cystic Fibrosis Adults. w24x Goldbeck L, Schmitz TG. Comparison of three generic questionnaires measuring quality of life in adolescents and adults
w25x
w26x
w27x
w28x
w29x
w30x
w31x
w32x
w33x
213
with cystic fibrosis: The 36-item short form health survey, the quality of life profile for chronic diseases, and the questions on life satisfaction. Qual Life Res 2001;10:23 –36. Quittner A.L., Buu A., Watrous M., Davis M.A., Cystic Fibrosis Questionnaire (CFQ): a health related quality of life measure 2000; The Cystic Fibrosis Foundation. Gillen M, Lallas D, Brown C, Yelin E, Blanc P. Work disability in adults with cystic fibrosis. Am J Respir Crit Care Med 1995;152:153 –6. Abbott J, Conway S, Etherington C, Fitzjohn J, Gee L, Morton A, et al. Perceived body image and eating behaviour in young adults with cystic fibrosis and their healthy peers. J Behav Med 2000;23:501 –17. Walters S. Sex differences in weight perception and nutritional behaviour in adults with cystic fibrosis. J Hum Nutr Dietet 2001;14:83 –91. Willis E, Miller R, Wyn J. Gendered embodiment and survival for young people with cystic fibrosis. Soc Sci Med 2001;53:1163 –74. Collins CE, O’Loughlin EV, Henry R. Discrepancies between males and females with cystic fibrosis in dietary intake and pancreatic enzyme use. J Pediatr Gastroenterol Nutr 1998;26(3):258 –62. Abbott J, Dodd M, Webb AK. Different perceptions of disease severity and self-care between patients with cystic fibrosis, their close companions and physician. Thorax 1995;50:794 –6. Abbott J, Dodd M, Gee L, Webb K. Ways of coping with cystic fibrosis: implications for treatment adherence. Disabil Rehabil 2001;23:315 –24. Staab D, Wenninger K, Gerbert N, Rupprath K, Bisson S, Trettin M, et al. Quality of life in patients with cystic fibrosis and their patients: what is important besides disease severity? Thorax 1998;53:727 –31.
Quality of life in cystic fibrosis: the impact of gender, general health perceptions and disease severity L. Geea, J. Abbotta,*, S.P. Conwayb, C. Etheringtonb, A.K. Webbc a
Department of Nursing, Faculty of Health, University of Central Lancashire, Preston PR1 2HE, UK b Adult Cystic Fibrosis Unit, Seacroft Hospital, Leeds LS14 6UH, UK c Adult Cystic Fibrosis Unit, Wythenshawe Hospital, Manchester M23 9LT, UK Received 13 September 2002; accepted 2 June 2003
Abstract Background: Disease progression in cystic fibrosis (CF) is marked by deterioration across a number of physiological systems. In addition, there is evidence that females have a worse prognosis than males. The current work assesses the impact of both these factors on health related quality of life (HRQoL). Methods: Two hundred and twenty-three adolescents and adults completed the cystic fibrosis quality of life (CFQoL) questionnaire with a further 185 approached and not responding by non-completion of the questionnaire. The CFQoL is divided into nine domains: physical, social, treatment, chest symptoms, emotional functioning, concerns for the future, relationships, body image, and career. Measurement of objective clinical status included, body mass index (BMI), and percentage of predicted forced expiratory volume in one second (FEV1 ). General health perceptions (GHP) were also measured. Results: Patients were sub-divided by gender and disease severity (mild )70% FEV1 , moderate 40–69% and severe -40%). Factorial analysis of variance indicated significant main effects for FEV1 (Fs587.98, PF0.001) and BMI (Fs 17.29, PF0.001) as a function of disease severity. Post hoc tests revealed significant two-group differences for FEV1 and BMI between disease severity groups. No differences were observed for gender across FEV1 or BMI. Differences emerged across most CFQoL domains for disease severity, with the exception of concerns for the future, which was consistently low throughout. Gender differences emerged for chest symptoms, emotional functioning, concerns for the future, body image and career. With the exception of body image, females exhibited poorer HRQoL. Pearson correlations indicated that females’ perception of health was more closely related to clinical status than males. Conclusions: Disease severity has an impact on HRQoL in adolescents and adults with CF. Some differences emerged between males and females, with females generally reporting poorer HRQoL. Evidence indicated that males and females perceived their health status differently, with females having a more accurate perception of objective clinical health status. 䊚 2003 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved. Keywords: Health related quality of life; General health perceptions; Cystic fibrosis
1. Introduction The natural course of cystic fibrosis (CF) is one of progressive deterioration in health. As the disease proceeds, decrements in physical functioning become more apparent placing greater limitations on the individual. Treatments also become both more aggressive and invasive. It may be anticipated that in parallel with these deteriorations in health status and increased treatment regimens that decrements in health related quality of life (HRQoL) would occur. How these changes, which have *Corresponding author. E-mail address: [email protected] (J. Abbott).
largely been brought about by increased longevity affect the HRQoL of adults with CF needs to be assessed w1x. Regarding gender, there is evidence to suggest that females with CF have a slightly worse prognosis, with a median mortality rate of 4 years less than males until the age of 20 beyond which male and female mortality rates are comparable w2,3x. Demko and colleagues w2x found that chronic infection of Pseudomonas aeruginosa in females occurred on average 1.7 years before colonisation in males. This outcome, was also highlighted by Sawyer et al. w4x who noted that women have more serious respiratory involvement than males and a lower survival rate from adolescence onwards. In the general
1569-1993/03/$30.00 䊚 2003 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved. doi:10.1016/S1569-1993Ž03.00093-6
L. Gee et al. / Journal of Cystic Fibrosis 2 (2003) 206–213
literature, there is evidence to suggest that men and women respond very differently to poor health w5–8x. For example, women have been observed to report higher levels of physical disease, greater pain and more subjective or emotional symptoms than men. Gender differences appear to reflect the way men and women respond to and perceive their health status rather than objective disease processes w5,7,9x. This suggests that men and women would differ in their reporting of health and illness. This was demonstrated in a study, which highlighted that women were more likely to perceive physical symptoms and act on them w7x. These findings are important in relation to HRQoL assessment, which requires subjective responding by the patient. In the past, differences between gender and different levels of disease severity in CF have been considered w10,11x, however, these factors have not previously been assessed using a CF specific quality of life measure. The present study examined HRQoL as a function of both disease severity and gender using the Cystic Fibrosis Quality of Life (CFQoL1) questionnaire w12x. Because the CFQoL is a patient derived measure, it is likely to be more sensitive to differences in disease severity and between men and women w13x. Any differences that do emerge will have repercussions in relation to clinical approaches with these different groups. The study also examined whether there was any variation in the relationships between health perceptions, clinical indicators and HRQoL between males and females. 2. Method 2.1. Study design A cross sectional questionnaire study of patients currently attending two regional cystic fibrosis specialist centres was undertaken. 2.2. Subjects All patients attending the regional adult cystic fibrosis centres in Manchester and Leeds in the United Kingdom, were considered eligible for the study. Over a 6-month period, all patients attending the outpatient clinic at their respective centre were approached to participate. 2.3. Measures Objective measurement of disease status included percentage of predicted forced expiratory volume in one second (FEV1) and body mass index (BMI). HRQoL was measured using the CFQoL questionnaire w12x. The CFQoL questionnaire is a patient derived disease spe1
Formatted copies of the CFQoL questionnaire, scoring equations and supplementary general health perception items are available on request from the corresponding author as hard copies or via e-mail.
207
cific HRQoL measure that has been fully validated. It consists of 52 items across nine domains of functioning: physical functioning, social functioning, treatment issues, chest symptoms, emotional functioning, concerns for the future, interpersonal relationships, body image, career issues. Scores are converted into values ranging from 0 (worst possible quality of life) to 100 (best possible quality of life). Scores of 50 or less reflect negative responding to items whilst scores of above 50 reflect positive responding to items. Negative responding to items suggests that the individual is experiencing difficulties within a particular domain. Further examination of the items within the domain can highlight where these problems lie. Item content was determined by feedback from patients, literature reviews and consultation with the multidisciplinary teams. Domain structures were determined by two rounds of factor analysis to arrive at the final questionnaire format. Cronbach alpha coefficients for internal reliability of each domain were high ranging from 0.72 to 0.91. Concurrent validity was assessed by comparing similar domains between the Short Form-36 item health status questionnaire w14x (SF-36) and the CFQoL. Correlations were shown to be highly significant and in the moderate to strong range (rs0.64 to rs0.74). Test re-test values were also strong with correlations ranging from rs0.74 to 0.96. The CFQoL was also demonstrated to display sensitivity to improvements brought about in disease status by intravenous antibiotics w12x. As outlined in Section 1, gender variations in health reporting appear to be based in the differences between the health perceptions of men and women. Given this link, participants were also given two additional questions alongside the CFQoL, which asked them about their general health perceptions (GHP). These were: in general, how is your health? Compared to others with cystic fibrosis, do you think your health is« Each question was answered on a five point scale from 1s poor, to 5swell above average. 2.4. Procedure The CFQoL questionnaire was distributed to patients when they attended a routine outpatients appointment. Return of questionnaires was via postal means. This method was chosen to minimise disruption to patients at clinic and to the running of the clinics. It was also felt that this method would give the patients time to reflect on whether they wanted to participate in the study and would remove the pressure inherent in a faceto-face request. Clinical and demographic data for sex, age, FEV1 and BMI were routinely assessed for all patients on the day they attended clinic. A pre-paid envelope was included for all patients to return the questionnaires. Patients were instructed to return their questionnaires as soon as possible after the clinic
208
L. Gee et al. / Journal of Cystic Fibrosis 2 (2003) 206–213
Table 1 Patient characteristics data
Number in group(n) Males(n) Females(n) Age in years Total sample Male Female % FEV1 Total sample Male Female BMI Total group Male Female
Mild
Moderate
Severe
Total sample
60 31 29
97 45 51
66 26 40
223 102 121
23.5 (6.6) "1.71 24.4 (6.9) "2.56 22.6 (6.2) "2.36
24.8 (7.3) "1.47 24.9 (6.4) "1.95 24.7 (8.0) "2.23
27.0 (6.9) "1.69 27.6 (5.5) "2.24 26.6 (7.6) "2.45
25.1 (7.1) "0.93 25.6 (6.4) "1.27 24.6 (7.6) "1.36
86.7 (11.3) "3.10 85.5 (9.8) "3.61 88.0 (12.8) "4.88
54.2 (9.0) "1.83 55.3 (9.4) "2.82 53.2 (8.8) "2.45
29.3 (7.2) "1.77 29.4 (8.1) "3.28 29.3 (6.6) "2.13
55.0 (23.5) "3.10 58.0 (23) "4.5 54.0 (23.8) "4.3
22.0 (2.78) "0.71 22.5 (2.5) "0.92 21.6 (3.0) "1.14
21.0 (2.2) "0.46 21.1 (2.3) "0.70 20.8 (2.6) "0.63
19.5 (2.1) "0.52 19.8 (1.8) "0.73 19.3 (2.3) "0.74
20.8 (2.5) "0.34 21.1 (2.4) "0.48 20.5 (2.6) "0.47
With the exception of the first three variables, values represent mean ("S.D.) and "95% confidence intervals around the mean. nsnumber in group.% FEV1spercentage of predicted forced expiratory volume in 1 s. BMIsbody mass index.
appointment preferably within a few days. The date of the outpatient appointment was recorded and patients were asked to also record the date they completed the questionnaire. During the study period, if a non-responding patient had a repeat attendance at clinic a second questionnaire was given to the patient and the above procedure was observed. All questionnaires included in the study were returned within a week of distribution. Clinical status and demographic information were also assessed for questionnaire non-responders and a comparison was made between groups to determine how representative the observed samples were. Data collection was continuous for a period of 6 months. In total 408 patients were approached and given questionnaires to complete and return. All responding patients completed the questionnaires during a stable phase of their disease. 2.5. Statistical analysis Scores on the CFQoL questionnaire were converted into values that ranged from 0 to 100. Descriptive statistics for clinical, demographic and HRQoL scores were normally distributed and are represented as means, standard deviations and confidence intervals. Disease groups were defined as a function of FEV1 and divided into three groups w15x: mild disease (FEV1)70%), moderate disease (FEV1 41–69%) and severe disease
(FEV1-40%). This division of mild, moderate and severe disease has been applied and adopted internationally as a categorisation of disease severity in both the medical and psychosocial literature and has been used as the standard for the epidemiological study in CF (ESCF) w16–18x. Differences between HRQoL scores for males and females and between disease severity groups were analysed using factorial analysis of variance (ANOVA). Post hoc analyses were applied using Tukeys Honestly Significant Difference (HSD) test. Relationships between GHP, BMI, FEV1 and CFQoL domains were analysed using Pearson product moment correlations. Statistical analyses were conducted using SPSS for windows version 6.0 w19x. Effect sizes for significant differences were calculated (quoted as ‘d’ in Section 3). Effect sizes were calculated using D-STAT meta-analytic software w20x. Magnitude of effect sizes are based on those quoted by Cohen w21x as: small effect sizes0.20, medium effect sizes0.50 and a large effect sizes0.80. 3. Results 3.1. Descriptive statistics Of the initial 408 patients who were approached and given a questionnaire to complete, 223 patients completed and returned questionnaires. Descriptive statistics for patient characteristics are presented in Table 1 as a
L. Gee et al. / Journal of Cystic Fibrosis 2 (2003) 206–213
function of disease severity for males and females separately and for the overall population. Descriptive statistics for domains of the CFQoL questionnaire, and GHP for gender, disease severity and the total population are presented in Table 2. The present sample represented a response rate of 55% with a further 185 not responding by virtue of failure to return questionnaires. Data for this sub-set have been presented in detail elsewhere w22x and so a brief summary of the outcome will be outlined for the present study. Of the 185 non-responders full demographic and clinical data were available for 103. Whilst a disparity was observed between the number of male and female non-responders with a higher percentage of males by comparison to females (68% and 46%, respectively), no differences were found between responders and non-responders for age, percentage of predicted FEV1 or BMI. For the present data, comparisons were also made between males and females in each disease severity group. No significant differences emerged between either males or females across disease severity groups for age, FEV1 or BMI. 3.2. Differences between demographic and clinical indicators as a function of gender and disease severity Two-way factorial ANOVA indicated significant main effects for gender=disease severity across FEV1 (Fs 395.90, PF0.001), BMI (Fs13.37, PF0.001) and age (Fs2.98, Ps0.03). Significant 1-way main effects were demonstrated across FEV1 (Fs587.98, PF0.001), BMI (Fs17.29, PF0.001) and age (Fs4.18, Ps0.01) for disease severity. FEV1 and BMI deteriorated in parallel with health status, whilst age increased as disease status deteriorated. Tukey’s HSD post hoc test for pair-wise analysis of groups indicated differences for both FEV1 and BMI across all disease group comparisons with effect sizes ranging from moderate to large (ds0.66– 1.49). Tukey’s HSD indicated that age differences were also significant across all group comparisons, with effect sizes in the small to moderate range (ds0.18–0.51). No differences in FEV1, BMI or age were observed between males and females. 3.3. CFQoL questionnaire scores as a function of disease severity and gender Applying factorial ANOVA on a domain-by-domain basis indicated that two-way main effects for all CFQoL domains across disease severity=gender were significant. The results also demonstrated highly significant one-way main effects for disease severity across all domains. Post Hoc analysis using Tukey’s HSD indicated significant pair-wise differences between mild and moderate groups for body image (ds0.62), interpersonal relationships (ds0.45) and concerns for the future
209
Table 2 Mean health related quality of life scores, standard deviations and 95% confidence intervals for the total sample, gender and disease severity Mild
Moderate
Severe
Total sample
Number in group Males 31 Females 29
45 52
26 40
102 121
Physical functioning Males 94.38 (9.2) "3.4 Females 87.37 (16.5) "6.3
86.57 (14.6) 73.84 (16.9) "4.4 "6.8 82.96 (20) 74 (21.3) "5.6 "6.8
85.7 (15.7) "3.0 81 (20.2) "3.6
Social functioning Males 91.93 (13.6) "5 Females 85.51 (22.5) "8.6
88.55 (14.9) 78 (25.4) "4.5 "10.2 84.23 (22.5) 78.25 (25.3) "6.3 "8.1
86.9 (18.4) "3.7 82.5 (23.5) "4.2
Treatment issues Males 87.09 (13.4) "5 Females 76.55 (25.1) "6.92
73.48 (23) "7 72.69 (25.2) "7
77.6 (22.6) "4.4 71.7 (25.4) "4.5
Chest symptoms Males 86.12 (19.8) "7.2 Females 65.17 (29.5) "11.2
68.66 (26.5) 58.46 (24.1) 71.3 (26) "8.0 "9.7 "5.1 65.76 (24.7) 51.25 (25.4) 60.82 (26.7) "6.7 "8.1 "4.8
Emotional responses Males 89.43 (11.6) "4.25 Females 80.43 (23.2) "8.8
81.77 (15.1) 78.26 (16.1) 83.2 (14.9) "4.5 "6.5 "2.9 77.06 (17.6) 73.87 (19.1) 76.8 (19.5) "5 "6.1 "3.52
Concerns for the future Males 60 (25.4) "9.3 Females 42.9 (24.1) "9.1
42.07 (21.7) 44.48 (23.3) "6.5 "9.39 42.17 (24) 41.58 (20.4) "6.7 "6.52
Interpersonal relationships Males 68 (22.1) "8.1 Females 72.75 (19.7) "7.5
56.22 (24.3) 61.92 (19.5) 61.2 (22.8) "7.3 "7.9 "4.4 63.73 (20.4) 56.05 (20.3) 63.3 (20.9) "5.7 "6.5 "3.7
73.58 (27.7) "11.2 67.16 (25.7) "8.2
48.1 (24.3) "4.7 42.1 (22.7) "4.0
Body image Males 76.77 (22.8) 50.81 (20.7) 52.82 (21.3) 59.2 (24.2) "8.3 "6.2 "8.6 "4.7 Females 77.47 (22.3) 72.17 (25.1) 61.83 (23.3) 70 (24.4) "8.5 "7.0 "7.5 "4.4 Career issues Males 75.64 (27.8) "10.1 Females 61.20 (32.2) "12.2
56.22 (28.9) 60.76 (24.9) 63.2 (28.5) "8.7 "10 "5.6 60.76 (26.3) 47.25 (25.1) 56.4 (27.9) "7.3 "8 "5.0
General health perceptions Males 76.9 (17.4) 67 (18.4) "6.4 "5.5 Females 75.27 (17.2) 61.96 (17.3) "6.5 "4.9
56.38 (15.5) "6.3 44.6 (18.9) "6
67.3 (18.8) "3.7 59.4 (21.1) "3.8
With the exception of the first variable, values represent mean ("S.D). and "95% confidence intervals around the mean.
210
L. Gee et al. / Journal of Cystic Fibrosis 2 (2003) 206–213
(ds0.39). Differences between moderate and severe disease groups were found for physical functioning (ds 0.57), social functioning (ds0.36) and chest symptoms (ds0.51). Whilst for mild and severe disease comparisons, statistically significant differences were found for the majority of domains (effect size range ds0.37– 0.99), with the exception of concerns for the future. Means indicated lower scores for increased severity groups highlighting poorer HRQoL as the disease progresses. Significant differences between males and females were found for body image (Fs10.79, Ps0.001, ds 0.44), chest symptoms (Fs8.85, Ps0.003, ds0.39), career concerns (Fs4.22, Ps0.013, ds0.27), emotional responses (Fs6.3, Ps0.013, ds0.33) and concerns for the future (Fs4.34, Ps0.038, ds0.28). Significant two-way interactions as a function of gender=disease severity were found for body image (Fs4.0, Ps0.02) and career issues (Fs3.06, Ps0.049). Males experienced poorer body image by comparison to females with the lowest scores for males being observed in the moderate disease group. Tukey’s HSD test yielded a significant difference between males and females for body image in the moderate disease group (ds0.91). No other significant differences were observed for body image. For career issues, females experienced more concerns in the mild and severe groups. This trend reversed in the moderate groups where males demonstrated more concerns regarding career issues. Although a significant interaction was observed, no statistically significant differences were found for career issues within disease severity groups between males and females. However, a moderate effect size was observed between males and females in the severe disease group (ds0.53). Marginally significant two-way interactions were found for chest symptoms (Fs2.43, Ps0.09) and concerns for the future (Fs2.17, Ps0.06). Females generally reported poorer chest symptoms. This was particularly marked in the mild disease group within which there was a significant difference between males and females (ds0.82) for chest symptom scores. No other significant differences were observed for chest symptoms between males and females. The interaction for concerns for the future indicated that across disease severity groups, females reported a poorer quality of life. This was particularly pronounced in the mild group in which there was a statistically significant difference between males and females (ds0.68).
Table 3a (a) Correlations (r) between FEV1 and CFQoL, domains in descending order of shared variance as a function of gender
3.4. General health perceptions as a function of gender
BMI and CFQoL domains for both males and females. For males significant relationships existed between BMI and body image (rs0.34, Ps0.001) which explained 11.6% of common variance (r 2=100) and chest symptoms (rs0.21, Ps0.02) explaining only 4% of variance. In common with males, a significant relationship
Further analyses using Pearson product moment correlations examined the relationships between CFQoL domains, FEV1, BMI and GHP as a function of gender. Only two significant correlations were observed between
(r)
(P)
% variance
Females Emotional functioning Relationships Physical functioning Body image Chest symptoms Career issues Treatment issues Concerns for the future Social functioning
0.60 0.28 0.25 0.25 0.21 0.18 0.17 nys nys
0.001 0.002 0.005 0.005 0.02 0.03 0.05 nys nys
36 8 6.25 6.25 4 3.25 3 nys nys
Males Physical functioning Body image Chest symptoms Treatment issues Emotional functioning Social functioning Concerns for the future Career issues Relationships
0.50 0.41 0.38 0.27 0.27 0.26 0.22 0.22 nys
0.001 0.001 0.001 0.005 0.005 0.007 0.02 0.02 nys
25 16 14.5 7.3 7.3 7 5 5 nys
rscorrelation coefficient. Pssignificance level. % variancespercentage of common variance explained. nyssnot significant. Table 3b (b) Correlations (r) between GHP and CFQoL domains in descending order of shared variance as a function of gender (r)
(P)
% variance
Females Physical functioning Chest symptoms Relationships Career issues Body image Social functioning Emotional functioning Treatment issues Concerns for the future
0.52 0.51 0.50 0.48 0.39 0.39 0.36 0.34 0.25
0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001
27 26 25 23 15 15 13 11.5 6.25
Males Physical functioning Chest symptoms Body image Career issues Emotional functioning Relationships Social functioning Treatment issues Concerns for the future
0.58 0.51 0.47 0.41 0.40 0.40 0.39 0.37 0.25
0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001
34 26 22 17 16 16 15 14 6.25
rscorrelation coefficient. Pssignificance level. % variancespercentage of common variance explained. nyssnot significant.
L. Gee et al. / Journal of Cystic Fibrosis 2 (2003) 206–213
was observed for females between BMI and body image (rs0.55, Ps0.001), which was stronger than the correlation observed for males and explained 30% of variance. The second significant but weak relationship for females was between BMI and concerns for the future (rs0.20, Ps0.02) explaining only 4% of common variance. Correlations between FEV1 and CFQoL domains are presented in Table 3a. Correlations were generally weak, with differences emerging in the descending order of shared variance as a function of gender. For females, the strongest relationship between FEV1 and CFQoL domains was for emotional functioning (rs0.60, Ps 0.001) with 36% of common variance explained. For males, the strongest correlation was between FEV1 and physical functioning (rs0.50, Ps0.001) with 25% of common variance explained. For correlations between GHP and CFQoL domains (Table 3b), the descending order of relationships did not vary greatly between males and females. The only difference worthy of comment was between GHP and relationship issues with a stronger correlation between variables for females, explaining 25% of variance by comparison to 16% of variance for males. Correlations between GHP and clinical indicators revealed that females demonstrated stronger relationships for GHP with both FEV1 (rs0.59, Ps0.001, 34% of variance), and BMI (rs0.34, Ps0.001, 11% of variance) by comparison to males (FEV1yrs0.41, Ps0.00, 17% variance, BMI–correlation not significant). 4. Discussion The results suggest that adolescents and adults with CF experience a progressive deterioration in the majority of CFQoL domains as their disease severity increases. This was with the exception of scores across the ‘concerns for the future’ domain, which remained low irrespective of disease severity and evidenced only a small effect size difference between the mild and moderate disease groups, an outcome that was likely to have been influenced by the significantly higher scoring males in the mild group. This suggests that concerns for the future are of particular salience to adults with CF throughout their lives. Additionally, it would seem that on a domain-by-domain basis for the CFQoL, differences between mild and moderate and moderate and severe groups varied. This suggests that at various points in the disease trajectory different HRQoL issues may emerge as problematic. Overall, between the mild and severe ends of the disease spectrum there were significant differences across the majority of domains. This was once again with the exception of ‘concerns for the future’. These findings reinforce the belief that interventions aimed at improving and maintaining the HRQoL
211
of adults with CF are paramount and should have an increasing emphasis as the individual’s disease status deteriorates. In particular, the study reflects the need to address patients’ concerns for the future. Regarding overall differences in CFQoL domains between males and females, where differences existed, females reported poorer HRQoL for chest symptoms, emotional functioning, concerns for the future and career issues. Gillen et al. w26x found that work disability in adults with CF was common, but not universal. They also outlined that few respondents to their study had received any kind of career guidance or counselling. The outcome of the present study would indicate that this type of guidance, whilst important for both genders, might be of greater importance when considering support for females with CF. For body image, males reported a poorer HRQoL. This is interesting given that there was no difference between males and females for FEV1 or BMI, and with males demonstrating a trend towards higher BMI than females. The finding that males have a poorer body image than females is in keeping with previous research w27x. The study by Abbott et al. indicated that by comparison to non-CF controls CF females were happier with their body image whilst CF males desired to be heavier. CF males also perceived their body shape as heavier than it actually was. This was by comparison to non-CF males who were more accurate when judging body shape. The finding may reflect cultural stereotypes, which lead women to desire thin bodies whereas males desire heavier and more muscular bodies. The idea of misperception of BMI also fits in with the current study, which indicated no relationship for males between general health perceptions and BMI and a weak relationship between BMI and body image, which only explained 11.6% of variance. However, Walters w28x found that women in her study tended to overestimate their weight by comparison to males who underestimated their weight. The study also concluded that perception of self as underweight was an important factor in determining nutritional behaviour and those patients who perceived themselves as underweight were more likely to be adherent to nutritional interventions. The difference in outcome between this study and the Abbott et al. paper may be based in the type of methodology employed. Whilst the Walters paper used BMI and self-perception as underweightyoverweight, the Abbott et al. study asked respondents to gauge which out of a series of silhouettes developed to represent various BMI’s was more like their own. This highlights an important distinction between satisfaction with body weight and perception of body weight. Willis and colleagues w29x have indicated in a qualitative study of men and women with CF, that of primary importance to adults with CF is the construction of their identity as either young women and young men, rather
212
L. Gee et al. / Journal of Cystic Fibrosis 2 (2003) 206–213
than adults with CF. Identification as people living with CF was secondary to their identification as men and women. This study also demonstrated that part of this role for women entailed a degree of passivity to their situation alongside an emphasis on having a slender body shape. This coupled with the finding that there was a greater discordance with adherence for women by comparison to men may suggest that not only do females prefer their slender body shape, but rather than being passive they may engage in non-adherence behaviours to enhance this ideal. Collins and colleagues w30x, who suggest that poorer compliance to dietary factors in females may be associated with the lower median survival rate of females by comparison to males, further support this notion. Interactions between gender and disease severity indicated that across some domains females had a poorer quality of life as a function of disease severity. This was despite having similar health status as indicated by objective clinical measures. This is important to consider when approaching treatment in females who appear more susceptible to lower HRQoL than males with the same objective clinical status. It is possible that the differences in CFQoL domain scores between males and females, was based on their perceptions of health rather than their actual health status. This would be consistent with research that states males and females responses to ill health are often based in perceptions rather than actual disease status w5,7,9,31x. Few notable differences emerged when considering relationships between FEV1, BMI, and GHP with CFQoL domains for males and females. The strongest relationship for females was between FEV1 and emotional functioning whilst for males it was between FEV1 and physical functioning. This outcome suggests that FEV1 has a more profound impact on emotional functioning in females whilst for males the greatest impact is on physical functioning. For the present study, potential weaknesses lie with the possible bias that may be introduced via the nonresponders. The representativeness of the responders by comparison to the non-responders for this sample has been discussed in detail elsewhere w22x. For the present study, which is largely concerned with gender differences, the percentage of non-responders may have influenced the higher number of female participants in the data set. However, based on previous analysis w22x, the conclusion was arrived at that the individuals who participated in the study were representative of the wider population sampled. Therefore, observed disparity between males and females was unlikely to have unduly influenced the outcome although this potential exists in any study. Research has also indicated that the present levels of responding (55%) are typical in CF w23x and also that the percentage of males and females responding are comparable to other research into HRQoL w24,25x in CF.
When considering the relationships between GHP and clinical indicators, it emerged that females’ health perceptions were more strongly associated with both FEV1 and BMI. However, relationships between clinical indicators and GHP were far from perfect suggesting that variables other than those considered in the present study may be responsible for overall health perceptions in both males and females. The question remains, what is influencing health perceptions if it is not clinical status? The answer to this question is beyond the scope of the present study, but areas which may provide a focus for further research include coping styles w32,33x, health locus of control, self efficacy, optimismypessimism or social support. The study by Staab and colleagues indicated that subjective health perceptions and ways of coping explained a significant amount of variance in overall HRQoL, above that which was accounted for by disease severity and time spent on therapy. This indicates that both of these factors are important contributors to overall HRQoL. Further research should be aimed at establishing what factors are likely to be involved in HRQoL responses in adolescents and adults with CF. References w1x Abbott J, Gee L. Contemporary psychosocial issues in cystic fibrosis: treatment adherence and quality of life. Disabil Rehabil 1998;20:262 –71. w2x Demko CA, Byard PJ, Davis PB. Gender differences in cystic fibrosis: Pseudomonas aeruginosa infection. J Clin Epidemiol 1995;48:1041 –9. w3x Davis PB. The gender gap in cystic fibrosis survival. J Gender Specific Med 1999;2(2):47 –51. w4x Sawyer SM, Rosier MJ, Phelan PD, Bowes G. The self-image of adolescents with cystic fibrosis. J Adolescent Health 1995;16:204 –8. w5x Mechanic D. Sex illness behaviour and the use of health services. Social Sci Med 1978;128:207 –14. w6x Merrill SS, Seeman TE, Kasl SV, Berkman LF. Gender differences in the comparison of self reported disability and performance measures. J Gerontol: Med Sci 1997;52a:M19 –M26. w7x Piccinelli M, Simon G. Gender and cross-cultural differences in somatic symptoms in association with emotional distress: an international study in primary care. Psychol Med 1997;27:433 –44. w8x Unruh AM. Gender variations in clinical pain experience. Pain 1997;65:123 –67. w9x Ruiz MT, Verbrugge LM. A two way view of gender bias in medicine. J Epidemiol Commun Health 1997;51:106 –9. w10x Congleton J, Hodson ME, Duncan-Skingle F. Quality of life in adults with cystic fibrosis. Thorax 1996;51:936 –40. w11x Congleton J, Hodson ME, Duncan-Skingle F. Quality of life in adults with cystic fibrosis. Thorax 1997;52:397 –400. w12x Gee LM, Abbott J, Conway S, Etherington C, Webb AK. The development of a disease specific health related quality of life measure for adults and adolescents with cystic fibrosis. Thorax 2000;55:946 –54. w13x Abbott J, Gee L. Quality of life in children and adolescents with cystic fibrosis: implications for optimising treatments and clinical trial design. Paediatr Drugs 2002;5(1):41 –56.
L. Gee et al. / Journal of Cystic Fibrosis 2 (2003) 206–213 w14x Ware JE, Sherbourne CD. The MOS short form health survey (SF-36): 1. Conceptual framework and item selection. Med Care 1992;30:473 –83. w15x Fiel SB, FitzSimmons S, Schidlow D. Evolving demographics of cystic fibrosis. Semin Respir Crit Care Med 1994;15:349 – 55. w16x Connolly MA, Johnson JA, Brown N, Montgomery M, Zuberbuhler P. Development of a disease specific quality of life instrument for people with cystic fibrosis. Institute of Health Economics Working Paper 00-11 2000. w17x Quittner AL, Buu A, Watrous M, Davis MA, Cystic Fibrosis Questionnaire (CFQ): a health related quality of life measure. The Cystic Fibrosis Foundation 2000. w18x Britto MT, Kotagal UR, Hornung RW, Atherton HD, Tsevat J, Wilmott RW. Impact of recent pulmonary exacerbations on quality of life in patients with cystic fibrosis. Chest 2002;121:64 –74. w19x SPSS Inc. SPSS for Windows release 6.0, 1993; Chicago, IL: SPSS Inc. w20x Johnson, BT. DSTAT: Software for the Meta-analytic Review of Research Literatures (version 1), 1989; Lawrence Erlbaum Associates. w21x Cohen J. Statistical power analysis for the behavioural sciences. In: Howell DC. Statistical methods for psychology (4th Ed.). Duxbury: p. 215. w22x Gee L, Abbott J, Conway S, Etherington C, Webb AK. Validation of the SF-36 for the assessment of health related quality of life in adults with cystic fibrosis. J Cystic Fibrosis 2002;1:137 –45. w23x Walters S., Association of Cystic Fibrosis Adults (UK) Survey 1994: Analysis and Report 1994; Association of Cystic Fibrosis Adults. w24x Goldbeck L, Schmitz TG. Comparison of three generic questionnaires measuring quality of life in adolescents and adults
w25x
w26x
w27x
w28x
w29x
w30x
w31x
w32x
w33x
213
with cystic fibrosis: The 36-item short form health survey, the quality of life profile for chronic diseases, and the questions on life satisfaction. Qual Life Res 2001;10:23 –36. Quittner A.L., Buu A., Watrous M., Davis M.A., Cystic Fibrosis Questionnaire (CFQ): a health related quality of life measure 2000; The Cystic Fibrosis Foundation. Gillen M, Lallas D, Brown C, Yelin E, Blanc P. Work disability in adults with cystic fibrosis. Am J Respir Crit Care Med 1995;152:153 –6. Abbott J, Conway S, Etherington C, Fitzjohn J, Gee L, Morton A, et al. Perceived body image and eating behaviour in young adults with cystic fibrosis and their healthy peers. J Behav Med 2000;23:501 –17. Walters S. Sex differences in weight perception and nutritional behaviour in adults with cystic fibrosis. J Hum Nutr Dietet 2001;14:83 –91. Willis E, Miller R, Wyn J. Gendered embodiment and survival for young people with cystic fibrosis. Soc Sci Med 2001;53:1163 –74. Collins CE, O’Loughlin EV, Henry R. Discrepancies between males and females with cystic fibrosis in dietary intake and pancreatic enzyme use. J Pediatr Gastroenterol Nutr 1998;26(3):258 –62. Abbott J, Dodd M, Webb AK. Different perceptions of disease severity and self-care between patients with cystic fibrosis, their close companions and physician. Thorax 1995;50:794 –6. Abbott J, Dodd M, Gee L, Webb K. Ways of coping with cystic fibrosis: implications for treatment adherence. Disabil Rehabil 2001;23:315 –24. Staab D, Wenninger K, Gerbert N, Rupprath K, Bisson S, Trettin M, et al. Quality of life in patients with cystic fibrosis and their patients: what is important besides disease severity? Thorax 1998;53:727 –31.