Aim: Primary lymphedema requires continuous conservative treatment during the patient's life, which may affect their health-related quality of life (HRQOL). Physical and psychosocial characteristics related to lymphedema are associated with their HRQOL. This study aimed to assess HRQOL in those patients undergoing conservative treatment and to determine the factors associated with their HRQOL.
Methods: This was a cross-sectional, observational study that included 83 patients with primary lymphedema at an outpatient clinic for lymphedema. HRQOL was assessed with the Medical Outcomes Study 36-Item Short Form Survey (SF-36) and EuroQol 5-Dimension (EQ-5D). Lymphedema status/complications and psychosocial status were evaluated by medical records, physical assessment, and self-administrative questionnaires.
Results: The general health perception score in the SF-36 was lower than the age- and sex-stratified national norms (P = 0.001). In the EQ-5D, the proportions of patients with some/moderate or severe problems were 42.2% for pain/discomfort, 21.7% for usual activities, and 20.5% for anxiety/depression. By multiple regression analyses, a lower physical component summary score in the SF-36 was associated with a higher lymphedema stage (P = 0.021), cellulitis within 30 days (P = 0.003), exercise (P = 0.010), and more substance use coping (P = 0.012). A lower mental component summary score was associated with skin lesions over edematous limbs (P = 0.008), less humor coping (P = 0.005), and more self-blame coping (P = 0.014).
Conclusion: Patients with primary lymphedema have problems in health perception, discomfort, usual activities, and anxiety/depression. Preventing cellulitis and skin lesions and humor coping may be able to improve HRQOL.
Lymphedema is a pathological condition of the lymphatic system with interstitial accumulation of protein-rich fluid (Warren, Brorson, Borud & Slavin, 2007). In the classification of lymphedema (Rockson, 2001), primary lymphedema, which accounts for 8% of new diagnoses of lymphedema (Sitzia et al., 1998), occurs because of a genetically determined abnormality of lymph drainage (Lee et al., 2010). This is pathophysiologically different from secondary lymphedema such as cancer-related or filariasis-related lymphedema (Warren et al., 2007).
Primary lymphedema has a significant impact on patients' quality of life (Williams, Moffatt & Franks, 2004) because it develops mainly from birth to adolescence by congenital or uncertain causes (Kerchner, Fleischer & Yosipovitch, 2008) and requires conservative treatment over a lifetime because of incurable and intractable conditions. However, there are limited numbers of studies on health-related quality of life (HRQOL) regarding patients with primary lymphedema. Previous studies of HRQOL of non-cancer-related lymphedema did not distinguish patients with primary lymphedema from those with secondary lymphedema (Cheville, Almoza, Courmier & Basford, 2010; Franks et al., 2006; Sitzia & Sobrido, 1997). Although another study targeted a limited group of patients with primary lymphedema undergoing ablative surgery for advanced edema conditions (Ogunbiyi, Modarai, Smith & Burnand, 2009), the HRQOL of the patients with primary lymphedema undertaking common conservative treatment remains unknown.
Furthermore, it is important to identify the factors related to HRQOL from a variety of aspects because patients with primary lymphedema often have both physical and mental impairments in their lives. Patients with primary lymphedema are likely to have a larger limb volume and have trouble with cellulitis than those with secondary lymphedema (Casley-Smith, 1995; Lam, Wallace, Burbidge, Franks & Moffatt, 2006; Sitzia et al., 1998). In addition, because patients have experienced life-long events because of edema or its complications, they have coping characteristics such as adaptation or denial as reported by a qualitative study (Williams et al., 2004), where they may require psychological or social support.
Despite the diversity of potential associated factors of HRQOL, there are few studies on patients with primary lymphedema. Although one study investigated demographics, comorbidities, and lymphedema-related factors (Cheville et al., 2010), psychosocial variables such as social support and coping strategies have not been assessed. Therefore, both physical and psychosocial factors need to be comprehensively recognized for these patients.
The purposes of this study were: (i) to assess HRQOL in patients with primary lymphedema who have conservative treatment; and (ii) to determine the physical and psychosocial factors associated with their HRQOL.
Setting and participants
This cross-sectional, observational study was conducted from November 2009 to November 2010 at an outpatient clinic for patients with lymphedema in Tokyo. This clinic was selected from the initiative facilities across Japan dealing with all types of lymphedema and the mean number of patients per week was approximately 150. In the clinic, conservative treatment including compression garments or bandages are prescribed for most patients and other lymphedema managements such as manual lymph drainage are provided if required.
All patients who met the following inclusion criteria were recruited: (i) they were diagnosed with primary lymphedema by a physician specializing in lymphedema, based on medical history and physical observations; and (ii) they visited the clinic at least two times and underwent conservative treatment. Patients aged below 16 years or those unable to fill in the questionnaire on their own were excluded. The study protocol was approved by the Ethical Committee of the Graduate School of Medicine, The University of Tokyo (#2763-). Written informed consent was obtained from all patients.
After examination by a specialist in lymphedema, data on patients' demographic details, and lymphedema status including edematous sites, edema staging, and history of complications were collected from the medical records. The physical status related to lymphedema consisting of Stemmer sign, skin lesions, and limb circumference were then evaluated by a trained researcher. Finally, patients completed a self-administered, structured questionnaire in the clinic which included questions relevant to lymphedema onset, psychosocial status including self-care and social support, and scales for coping style and HRQOL scales. When the participants had limited time to answer the questionnaire, they were asked to mail it back to the researcher in a prepaid envelope within a week.
Demographic data included age, sex, body mass index, and the number of admissions to the clinic.
Edematous sites were classified into upper, lower, or both, and unilateral or bilateral. Edema staging was evaluated by the International Society of Lymphology (ISL) lymphedema staging, which was divided into five categories: stage 0, where there is subclinical conditions despite impaired lymph transport; stage I, where there is an early accumulation of fluid that subsides with limb elevation; stage II, where there is limb elevation alone that rarely reduces swelling and pitting is manifest; late stage II, where there may or may not be pitting as tissue fibrosis supervenes; and stage III, which encompasses lymphostatic elephantiasis where pitting is absent and trophic skin changes such as acanthosis, fat deposits, and warty overgrowths develop (International Society of Lymphology, 2003). An incidence of cellulitis within 30 days and a history of hospital admission for cellulitis were evaluated as a history of complications.
Stemmer sign, skin lesions, and the percent difference of limb circumference were evaluated as physical status related to lymphedema. Stemmer sign, which is defined as the inability to pinch a fold of skin (Moffatt, 2006), was examined at the base of the middle finger or second toe. A skin lesion was defined as a skin defect over the edematous area in the present study. Limb circumference was measured at four sites of the extremities using a tape measure. The sites of measurement were at 5 cm and 10 cm distal and proximal to the midpoint of the lateral epicondyle for the upper extremities, and 10 cm and 20 cm for the lower extremities in the same manner. The site of percent difference of limb circumference was determined at the most different length between affected limb and unaffected limb in the four sites of measurement. Modified Yoshizumi's formula was selected for the calculation: percent difference of limb circumstances (%) = (difference between limb circumferences/circumferences of unaffected limb [cm]) × 100 (Yoshizumi et al., 1992)
The age of onset of edema was evaluated and classified into congenital edema (≤2 years old) or late onset (>2 years old). The time since onset of edema was also evaluated.
The social status, self-care, having a job, support from their family and belonging to the patients' association of lymphedema were evaluated. Self-care included compression garments, bandaging, skin care, self-lymphatic massage, exercise, and intermittent pneumatic compression.
Coping style was assessed by the Brief-COPE, a short version of the Coping Orientation to Problems Experienced (COPE) scales (Carver, 1997; Carver, Scheier & Weintraub, 1989). The questionnaire consists of 14 subscales, with two items each, and it measures conceptually differentiable coping reactions. Scores range 2–8 for each subscale, with higher scores representing more use of such coping styles. In the Japanese version of the Brief-COPE, criteria and construct validity have been confirmed (Otsuka, Suzuki, Takada, Tomotake & Nakata, 2008). Cronbach's alpha for each subscale in the present study ranged 0.52–0.63.
Two types of HRQOL scales, the Medical Outcomes Study 36-Item Short Form Survey (SF-36) version 2.0 and EuroQol 5-Dimension (EQ-5D), were used.
The SF-36 was developed to measure generic health status, and consists of eight subscales: physical functioning (PF), role-physical (RP), bodily pain (BP), general health perception (GH), vitality (VT), social functioning (SF), role-emotional (RE), and mental health (MH) (Ware & Sherbourne, 1992). The raw score on each scale ranges 0–100 with a lower score indicating poorer health status. The Japanese version of SF-36 was used in this study, which is validated for the Japanese general population and lymphedema patients (Fukuhara, Bito, Green, Hsiao & Kurokawa, 1998; Sakuda, Miyakoshi, Kataoka, Sakaguchi & Satou, 2007). First, the z-scores of each subscale were calculated as a norm-based score (NBS) from the Japanese population according to the SF-36 manual (Fukuhara & Suzukamo, 2001). Factor loadings for the Japanese population were then used to calculate the score of the physical component summary (PCS) and mental component summary (MCS) (Fukuhara & Suzukamo, 2001). The mean of NBS was set at 50 and every 10 points indicated 1 standard deviation (SD). Cronbach's alpha for each subscale and summary score in the present study ranged 0.85–0.87.
The EQ-5D is a generic HRQOL profile instrument that was developed for measuring utility (The EuroQol Group, 1990). It contains five domains: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. A single answer with three possible response options (1 = no problem, 2 = some/moderate problems, 3 = extreme problems) is required. The EQ-5D has been found to be sensitive to the effect of lymphedema on HRQOL (Cheville et al., 2010). The Japanese version was validated for the Japanese general population (Tsuchiya et al., 2002). Scores from the five domains were combined into a single utility score between −0.111 (worst possible state) and 1.000 (best possible state) based on the Japanese weighting system (Japanese EuroQol Translation Team, 1998).
Descriptive data are expressed as N (%) for categorical variables and mean (SD) for continuous variables because these variables were normally distributed. Differences between groups were evaluated by Student's t-test, χ2-test, or Fisher's exact probability test. The statistical significance level was set at P = 0.050. All statistical analyses were performed using Statistical Analysis System software, version 9.2 (SAS Institute, Cary, NC, USA).
The differences between the NBS of the SF-36 and age- and sex-stratified national norm (ASNN) scores were tested. The frequencies of patients below 2 SD of the ASNN for the SF-36, or those who had some/moderate or severe problems for each subscale of the EQ-5D, were calculated. HRQOL scores among subgroups of age (16–29, 30–59, and ≥60 years old), age of onset (0–2, 3–18, 19–35, and ≥36 years old), and the time since onset of edema (≤5, 6–14, and ≥15 years) were assessed with anova with the post-hoc Dunnet's test.
Multiple regression analysis using backward selection with P-values for inclusion set at 0.100 was applied to identify associated factors with PCS, MCS, and the EQ-5D utility index scores. Independent variables that were correlated with each HRQOL score at P < 0.200 by Spearman's rank correlation coefficient were included into the multiple regression analysis. If strong correlations between independent variables were found, one of the variables was entered into the model to avoid multicollinearity. Age, sex, and age of onset were then adjusted. The fitness of the model was assessed by adjusted R2. The association between the percentage difference of circumference and dependent variables was assessed in subgroups with unilateral edema by Spearman's rank correlation coefficient.
The associations between independent variables remaining in multiple regression analysis and scores of HRQOL subscales were then analyzed.
Eighty-nine patients with primary lymphedema were recruited. Two patients refused to participate and three had invalid answers for more than half of the questionnaire. Among 11 patients who filled in the questionnaire at home, one patient did not return the questionnaire. Therefore, a total of six patients were excluded and 83 patients were subjected to analyses.
Demographics and lymphedema status are shown in Table 1. The mean age was 42.3 years (SD, 14.2). Sixty (72.3%) patients were female. Seventy-eight (94.0%) patients had lymphedema only in their lower limbs and 67 (80.7%) had unilateral edema. With regard to ISL lymphedema staging, the most frequent stage was stage II (n = 45, 54.2%). Twenty-six (31.3%) patients suffered from cellulitis within 30 days. Seventy-six (91.6%) patients were classified as late-onset edema. Seventy-five patients (90.4%) used compression garments for their self-care and 22 (26.5%) used bandaging. Fifty-seven (68.7%) patients received some types of support from their family regarding their edema and six (7.2%) patients belonged to the patients' association of lymphedema. Among the Brief-COPE subscales (Table 2), active coping, planning, and acceptance showed the highest mean score of 6.0, whereas substance use showed the lowest mean score of 3.2.
Table 1. Demographics and lymphedema status of patients
|Age (years), mean (SD)||42.3 (14.2)|
|Sex, n (%)|| |
| Male||23 (27.7)|
| Female||60 (72.3)|
|BMI (kg/m2), mean (SD)||22.0 (3.8)|
|No. of admissions to the clinic, median (range)||11 (2–52)|
|Site of edema, N (%)|| |
| Upper limb||3 (3.6)|
| Lower limb||78 (94.0)|
| Both upper and lower limb||2 (2.4)|
|Unilateral/bilateral edema, N (%)|| |
| Unilateral||67 (80.7)|
| Bilateral||16 (19.3)|
|ISL lymphedema staging, N (%)|| |
| Stage 0||0 (0.0)|
| Stage I||2 (2.4)|
| Stage II||45 (54.2)|
| Late stage II||23 (27.7)|
| Stage III||13 (15.7)|
|Cellulitis within 30 days, N (%)||26 (31.3)|
|History of hospital admission for cellulitis, N (%)||22 (26.5)|
|Stemmer sign, N (%)||52 (62.7)|
|Skin lesions over the edematous limbs, N (%)||11 (13.3)|
|Limb circumference difference (%)†, mean (SD)||19.0 (12.5)|
|Time since onset of edema (years), mean (SD)||15.5 (13.2)|
|Classification of primary lymphedema based on onset, N (%)|
| Congenital||7 (8.4)|
| Late onset||76 (91.6)|
| Compression garments, N (%)||75 (90.4)|
| Bandaging, N (%)||22 (26.5)|
| Skin care, N (%)||26 (31.3)|
| Self-lymphatic massage, N (%)||48 (57.8)|
| Exercise, N (%)||31 (37.4)|
| Intermittent pneumatic compression, N (%)||11 (13.3)|
|Having a job, N (%)||47 (56.6)|
|Support from family, N (%)||57 (68.7)|
|Belonging to the patients' association of lymphedema, N (%)||6 (7.2)|
Table 2. Coping styles of patients with primary lymphedema
|Active coping||6.0 (1.0)|
|Substance use||3.2 (1.5)|
|Use of emotional support||5.4 (1.2)|
|Use of instrumental support||5.4 (1.2)|
|Behavioral disengagement||3.8 (1.0)|
|Positive reframing||5.6 (1.2)|
Scores for SF-36 and EQ-5D are shown in Table 3. In the subscales of the SF-36, the mean GH score was lower than the ASNN score (P = 0.001). The mean (SD) PCS and MCS scores were 50.8 (10.7) and 49.1 (9.1), respectively, which were not significantly different from the ASNN score. Among the SF-36 subscales, the number of patients reporting a score below 2 SD of the ASNN score was highest in RP (7.2%). The mean (SD) score of the EQ-5D utility index was 0.853 (0.144). In the EQ-5D subscales, the highest number of patients who had some/moderate or severe problems was 35 (42.2%) for pain/discomfort, followed by 18 (21.7%) for usual activities, 17 (20.5%) for anxiety/depression, seven (8.4%) for mobility, and three (3.6%) for self-care.
Table 3. Domain scores of SF-36 and EQ-5D
|SF-36|| || || |
| Physical functioning (PF)||52.0 (8.1)||0.302||3 (3.6)|
| Role-physical (RP)||49.3 (10.4)||0.310||6 (7.2)|
| Bodily pain (BP)||52.1 (11.0)||0.117||4 (4.8)|
| General health perception (GH)||46.7 (9.9)||0.001||4 (4.8)|
| Vitality (VT)||49.2 (9.2)||0.710||1 (1.2)|
| Social functioning (SF)||49.6 (9.9)||0.898||3 (3.6)|
| Role-emotional (RE)||50.0 (10.4)||0.943||5 (6.0)|
| Mental health (MH)||50.7 (9.0)||0.278||2 (2.4)|
| Physical component summary (PCS)||50.8 (10.7)||0.977||5 (6.0)|
| Mental component summary (MCS)||49.1 (9.1)||0.954||2 (2.4)|
|EQ-5D|| || || |
| Mobility||–||–||7 (8.4)|
| Self-care||–||–||3 (3.6)|
| Usual activities||–||–||18 (21.7)|
| Pain/discomfort||–||–||35 (42.2)|
| Anxiety/depression||–||–||17 (20.5)|
| Utility index||0.853 (0.144)||–||–|
Only the PCS score had a significant difference in age (PCS, P = 0.016; MCS, P = 0.192; EQ-5D, P = 0.321), where patients aged under 30 years old had a higher PCS score than those who were aged over 60 years (P = 0.009). There were no significant differences in HRQOL scores between subgroups by age of onset of edema (PCS, P = 0.240; MCS, P = 0.678; EQ-5D, P = 0.812), and by duration of edema and HRQOL scores (PCS, P = 0.918; MCS, P = 0.315; EQ-5D, P = 0.435).
Factors associated with HRQOL
In univariate analysis, independent variables correlated with the HRQOL score at P < 0.200 were as follows: for PCS, ISL stage, cellulitis within 30 days, skin care, self-massage, exercise, and six Brief-COPE subscales; for MCS, ISL stage, belonging to the lymphedema patients' association, skin lesions over edematous limbs, and eight Brief-COPE subscales for MCS; and for the EQ-5D utility index, ISL stage, cellulitis within 30 days, hospital admission for cellulitis, skin lesions, support from their family, self-massage, exercise, and five Brief-COPE subscales. The authors focused on the difference between unilateral and bilateral edema and examined the correlation between unilateral/bilateral edema and the scores of HRQOL scales. However, these correlations were not significant.
Multiple regression analysis (Table 4) showed that a lower PCS score was associated with a higher ISL stage (P = 0.021), cellulitis within 30 days (P = 0.003), exercise (P = 0.010), and a higher score of substance use coping (P = 0.012). A lower MCS score was associated with skin lesions over edematous limbs (P = 0.008), a lower score of humor coping (P = 0.005), and a higher score of self-blame coping (P = 0.014). A lower EQ-5D utility index score was associated with a higher ISL stage (P = 0.010), exercise (P = 0.002), and a higher score of substance use coping (P = 0.013). The P-value for partial regression coefficients of these variables remained significant in adjusted models and their directions were the same as those in unadjusted models. In the subgroups of patients with unilateral lymphedema, there were no significant correlations between the percent difference of limb circumference and HRQOL scales.
Table 4. Multiple regression analysis of factors associated with HRQOL scores
|ISL Stage (0: stage 0, I and II; 1: late II and III)||−5.07||0.021||0.005||–||–||–||−0.08||0.010||0.025|
|Cellulitis within 30 days (0, none; 1, yes)||−6.97||0.003||0.008||–||–||–||–||–||–|
|Skin lesions on the edematous limbs (0, none; 1, yes)||–||–||–||−6.67||0.008||0.005||–||–||–|
|Self-care|| || || || || || || || || |
| Exercise (0, none; 1, yes)||−5.88||0.010||0.021||–||–||–||−0.10||0.002||0.002|
| Substance use||−1.79||0.012||0.039||–||–||–||−0.02||0.013||0.023|
|Adjusted R2|| ||0.211||0.210||–||0.289||0.315||–||0.208||0.181|
Score of HRQOL subscales in relation to the presence of cellulitis or skin lesions
Subscales of SF-36 and EQ-5D were further investigated with regard to cellulitis and skin lesions (Table 5). The scores for RP (P = 0.035), BP (P = 0.018), SF (P = 0.014), and RE (P = 0.045) were lower in patients with cellulitis within 30 days than in those without cellulitis. The scores for BP (P = 0.047) and GH (P = 0.003) were lower in patients with skin lesions than in those without skin lesions. There were no significant differences in all domains of the EQ-5D, except that more patients with skin lesions tended to have problems with pain/discomfort than those without skin lesions.
Table 5. Domain scores of SF-36 and EQ-5D between with and without cellulitis or skin lesions
|SF-36, mean (SD)|| || || || || || |
| Physical functioning (PF)||51.3 (7.7)||52.4 (8.3)||0.600||53.1 (5.3)||51.9 (8.5)||0.629|
| Role-physical (RP)||44.9 (13.6)||51.2 (7.8)||0.035||48.0 (11.8)||49.5 (10.1)||0.641|
| Bodily pain (BP)||47.2 (13.3)||54.3 (9.1)||0.018||46.3 (11.1)||53.1 (10.8)||0.047|
| General health perception (GH)||46.7 (9.3)||46.7 (10.2)||0.992||39.1 (8.3)||48.0 (9.6)||0.003|
| Vitality (VT)||49.3 (8.0)||49.2 (9.8)||0.947||45.1 (10.8)||49.9 (8.8)||0.097|
| Social functioning (SF)||45.7 (11.2)||51.4 (8.8)||0.014||45.6 (11.6)||50.3 (9.6)||0.127|
| Role-emotional (RE)||46.1 (12.7)||51.8 (8.6)||0.045||44.5 (14.2)||51.0 (9.4)||0.154|
| Mental health (MH)||51.4 (6.8)||50.5 (9.8)||0.634||45.6 (12.4)||51.6 (8.0)||0.129|
|EQ-5D, n (%)|| || || || || || |
| Mobility||3 (11.5)||4 (7.0)||0.672b||2 (16.7)||5 (7.0)||0.266b|
| Self-care||0 (0.0)||3 (5.3)||0.548b||0 (0.0)||3 (4.2)||1.000b|
| Usual activities||8 (30.8)||10 (17.5)||0.175||3 (25.0)||15 (21.1)||0.717|
| Pain/discomfort||14 (53.9)||21 (36.8)||0.146||8 (66.7)||27 (38.0)||0.063|
| Anxiety/depression||7 (26.9)||10 (17.5)||0.326||4 (33.3)||13 (18.3)||0.255|
This is the first study the authors are aware of to investigate HRQOL and associated factors of patients with primary lymphedema who have conservative treatment. Additionally, the present study is the first to have collected more than 80 patients with primary lymphedema for QOL research and the collection rate of the questionnaires was more than 90%. There were several novel findings in this study. First, these patients had problems in some domains of HRQOL such as general health perceptions, pain/discomfort, usual activities, and anxiety/depression, although summary scores of HRQOL were similar to the ASNN score. Second, with regard to associated factors, the authors found that cellulitis was associated with the PCS score in the SF-36, and both skin lesions and positive coping style such as humor were associated with the MCS score.
The finding that most of the NBS of the SF-36 were similar to the Japanese ASNN score except for the GH score was a surprising result, because a poor HRQOL of patients with lymphedema has been reported in a previous study (Franks et al., 2006). This may be due to the fact that the target subjects in the present study were different from the previous study. Additionally, the present study targeted physically independent outpatients, less patients had complications of skin lesions, and more patients received support from their family than those in the previous study including patients living alone with leg ulcers (Franks et al., 2006).
On the other hand, the proportion of patients having some/moderate or severe problems was relatively large in the subscales of pain/discomfort, usual activities, and anxiety/depression in the EQ-5D in addition to having a lower score of GH in the SF-36. Because the BP score in the SF-36 was not significantly different from the ASNN score, the pain/discomfort result indicated that patients complained of discomfort rather than bodily pain, reflecting a sense of heaviness or languidness due to painless swelling (Lee et al., 2010). Because a few patients had mobility problems in the EQ-5D, the result for usual activities indicated that patients with primary lymphedema complained more about impairment of daily activities than for mobility. Impairment of anxiety and depression has been previously found in these patients (Williams et al., 2004). It is important to determine the relevance of discomfort for edema, impairment of usual activities, and anxiety/depression, which may deteriorate general health perception for these patients.
This study also evaluated the factors associated with HRQOL. ISL stage, cellulitis within 30 days, exercise as self-care, and substance use coping were associated with the physical aspect of HRQOL, and skin lesions, denial, humor, and self-blame coping were associated with the mental aspect. ISL stage, cellulitis within 30 days, exercise as self-care, and substance use coping were associated with utility. The association of lymphedema staging and negative coping style with HRQOL has previously been shown to be associated with HRQOL in other types of lymphedema (Cheville et al., 2010; Passik, Newman, Brennan & Tunkel, 1995). It appears that cellulitis, skin lesions, and humor coping style as associated factors have been found for the first time in patients with primary lymphedema.
Cellulitis within 30 days was significantly associated with a lower PCS score and RP, BP, SF, and RE sub-scores of the SF-36, but not with the MCS score, which may indicate that cellulitis has a large impact on the physical aspect of HRQOL rather than the mental aspect. However, a previous study on lymphedema without cancer did not find a relationship between a history of cellulitis and HRQOL (Cheville et al., 2010). This disagreement between studies may be attributed to the timespan to evaluate cellulitis because the present study assessed a recent incidence within 30 days, which could have a more direct impact than the history from onset (Cheville et al., 2010). In addition, the fact that most patients had edema in the lower leg could enhance the association of cellulitis because cellulitis in the lower legs is characterized by swelling and erythema with pain (Cox, Colver & Paterson, 1998). This site-specific characteristic could result in a decrease in RP and SF scores in the SF-36 through restriction of the lower limbs by having cellulitis. The result could be unique to patients with primary lymphedema because a previous study for patients with breast cancer-related lymphedema mainly in the upper limbs reported that acute inflammatory episodes, which is a similar condition to cellulitis, is related to psychological distress as well as disability (Chachaj et al., 2010).
The presence of skin lesions was negatively associated with the MCS score and BP and GH scores of the SF-36, but not with the PCS score, indicating that skin lesions can lead to bodily pain, which could have an impact on the mental aspect of HRQOL such as health perceptions rather than physical impairments. This finding is supported by a report stating that chronic wound-related pain is intensely connected to psychological well-being as well as physical well-being (Woo & Sibbald, 2008). In addition, pain itself gives lymphedema patients the feeling that it is exceeding their control and there is something wrong with their body (Passik & McDonald, 1998).
The present study showed that coping with humor has a positive association with the mental aspect of HRQOL among patients with primary lymphedema contrary to dysfunctional coping such as denial, substance use, and self-blame, which are generally regarded as negative factors of psychological distress for patients with breast upper limb lymphedema (Passik et al., 1995). Although a qualitative study showed that some patients with primary lymphedema are positively resigned to having lymphedema among other patients having negative feelings (Williams et al., 2004), the association between positive coping and HRQOL has not been assessed. Because humor coping style is linked with positive emotion in cancer patients (Shapiro, McCue, Heyman, Dey & Haller, 2010), it has been suggested that, even in patients with primary lymphedema, positive coping style such as humor might be associated with adjustment of their lives with illness and might contribute to an improvement of HRQOL.
The present study also showed that exercise was related to a lower PCS score, which was an unexpected result. Because the authors could not collect the details of exercise, it is possible that heavy exercise, which should be avoided for lymphedema (Moffatt, 2006), may lead to a poorer HRQOL. Conversely, patients might exercise due to a poor physical status.
There are limitations to this study. This is a cross-sectional study and thus the authors cannot mention cause and effect relationships. Also, the duration of edema, age of onset, and the percent difference of the limbs were not factors associated with HRQOL for primary lymphedema, although they were predicted to be important factors for primary lymphedema. This may be because this study used the general HRQOL scale, which cannot evaluate specific components of HRQOL reflecting body image or uncertainty of diagnosis for primary lymphedema (Williams et al., 2004). Therefore, a specific HRQOL scale for primary lymphedema needs to be developed and further prospective studies may reinforce the HRQOL studies of patients with lymphedema.
Clinically, cellulitis and skin lesions can be prevented and coping style can be intervened in daily life, while it is difficult to improve advanced lymphedema stage. Self-management such as monitoring or care, patient education, and psychosocial support are important for all types of lymphedema (Moffatt, 2006). However, it has been reported that patients with primary lymphedema are less likely to perform self-monitoring than those with cancer-related lymphedema (Lam et al., 2006). In addition, the present study found that not many patients carried out skin care as self-care, neither did they belong to the patients' association of lymphedema. Creating a supportive environment through consultation with health professionals or self-help groups may encourage patients with primary lymphedema to access preventive care for complications and positive coping strategies, which might improve their HRQOL even at a progressed lymphedema stage.
The present study found that patients with primary lymphedema undertaking conservative treatment had problems in domains of HRQOL such as general health perception, discomfort, usual activities, and anxiety/depression; however, they were not very different from the Japanese standard HRQOL scores. Moreover, the study revealed that both physical and psychosocial factors were associated with their HRQOL; recent cellulitis was associated with physical aspects, and skin lesions, and positive humor coping were associated with mental aspects.