Summary
Background and objectives
This study aimed to investigate the influence of social support and other psychosocial factors on mortality, adherence to medical care recommendations, and physical quality of life among hemodialysis patients.
Design, setting, participants, & measurements
Data on 32,332 hemodialysis patients enrolled in the Dialysis Outcomes and Practice Patterns Study (1996 to 2008) in 12 countries were analyzed. Social support and other psychosocial factors related to ESRD and its treatment were measured by patient self-reports of health interference with social activities, isolation, feeling like a burden, and support from family and dialysis staff. Cox regression and logistic regression were used to examine associations of baseline social support and other psychosocial factors with all-cause mortality and with other measured outcomes at baseline, adjusting for potential confounders.
Results
Mortality was higher among patients reporting that their health interfered with social activities, were isolated, felt like a burden, and were dissatisfied with family support. Poorer family support and several psychosocial measures also were associated with lower adherence to the prescribed hemodialysis length and the recommended weight gain between sessions. Some international differences were observed. Poorer self-reported social support and other psychosocial factors were associated with poor physical quality of life.
Conclusions
Poorer social support and other psychosocial factors are associated with higher mortality risk, lower adherence to medical care, and poorer physical quality of life in hemodialysis patients. More research is needed to assess whether interventions to improve social support and other psychosocial factors will lengthen survival and enhance quality of life.
Introduction
The treatment of ESRD by dialysis creates considerable stress on patients, including potential changes in family relations and social life. Psychosocial factors have been associated with patients' subsequent adjustment to dialysis (1). Many studies have addressed the role of social support in predicting the course of chronic disease. Social support has been defined as the perception that an individual is a member of a complex network in which one can give and receive affection, aid, and obligation (2). Social support can be received from family members, friends, colleagues, and medical personnel. An important distinction is made between a person's number of relationships and a person's perception of the supportive value of social interactions. The former is usually referred to as the social network; the latter is referred to as perceived social support. Although there is no clear agreement on its definition, greater social support has been linked to better health outcomes in community and clinical samples (3).
Among persons with ESRD, perceived social and family support has been associated with lower mortality risk (4,5). In recent research, perceived discrepancy between expected and received social support was associated with higher mortality, whereas frequency of interaction with members of one's social network was not associated with mortality (6). Quality of life (QoL) and adherence to medical care are also important outcomes to consider and have been associated with mortality (7–11). Several studies have examined the association between social support and patient adherence to medical care. Two studies found association between family/friends support and adherence with recommended fluid intake (12,13). Other studies have failed to show any link between social support and biochemical indicators of adherence and missed or shortened dialysis sessions (5,14). Inconsistency in these results may be explained by small sample sizes and/or inadequate control for medical and treatment variables. Other possible explanations are the lack of a consensus standard to define and measure adherence (15), the use of different measures of social support, and different cutoff values for sensitivity and specificity in each study.
Associations between social support and QoL have also been observed (16), although not consistently among dialysis patients. Tell et al. (17) found a strong association between perceived social support and several measures of health-related quality of life (HRQoL) in dialysis patients, whereas Vazquez et al. (18) did not.
The Dialysis Outcomes and Practice Patterns Study (DOPPS) is an international prospective cohort study of hemodialysis practices and outcomes (19). In addition to assessing clinical characteristics associated with mortality and other numerous clinical outcomes, the DOPPS permits the assessment of relevant patient-centered care issues such as adherence with treatment recommendations and HRQoL, as captured by the Kidney Disease Quality of Life Scale (KDQoL) (20). KDQoL measures different aspects of the concept of HRQoL and includes items that capture social support and other psychosocial factors. The main purpose of this study was to investigate associations between these self-reported measures and the following outcomes: mortality, adherence to medical care, and physical QoL.
Materials and Methods
Study Design and Data Sources
We analyzed enrollment data on patients in DOPPS I (1996 to 2001), II (2002 to 2004), and III (2005 to 2008). The DOPPS is a prospective cohort study of adult hemodialysis patients randomly selected for study participation from representative dialysis facilities in 12 countries. The DOPPS sampling plan and study design have been published (19,21).
We examined associations between several measures of social support and other psychosocial factors (assessed by patient self-reported opinions) with patient outcomes. For these analyses, data were available from 308 dialysis facilities in DOPPS I (n = 12,465 patients from seven countries: France, Germany, Italy, Japan, Spain, the United Kingdom, and the United States), from 320 dialysis facilities in DOPPS II (n = 10,551 patients from the same seven countries with the addition of Australia and New Zealand, Belgium, Canada, and Sweden), and from 297 dialysis facilities in DOPPS III (n = 9316 patients from the 12 countries in DOPPS II). Baseline data for patient demographics, comorbidities, and clinical characteristics were obtained at study entry for each study phase.
Measurement of Social Support and Other Psychosocial Factors
At DOPPS enrollment, participants were asked to complete a self-administered patient questionnaire that included questions from the KDQoL Short Form (KDQoL-SF) (22). From the KDQoL-SF, the study presented here used four key questions to evaluate social support and three questions to evaluate other psychosocial factors. Each question was rated on a Likert scale and dichotomized for the purposes of our analysis (Table 1).
Table 1.
Scoring categorization for patient self-reported items
| Questione | Patient Responses |
Category B Referred to in Text as | |
|---|---|---|---|
| Category A (Reference): Higher Level of Social/Family Relationsd | Category B: Lower Level of Social/Family Relationsd | ||
| Social support | |||
| Concerning your family and friends, how satisfied are you withb | |||
| 1. The amount of time you are able to spend with your family and friends?b | Somewhat satisfied/very satisfied | Very dissatisfied/somewhat dissatisfied | Dissatisfied with family time |
| 2. The support you receive from your family and friends?b | Somewhat satisfied/very satisfied | Very dissatisfied/somewhat dissatisfied | Dissatisfied with family support |
| How true or false are each of the following statements?c | |||
| 3. Dialysis staff encourages me to be as independent as possible.c | Definitely true/mostly true | Do not know/mostly false/definitely false | Dissatisfied with staff encouragement to be independent |
| 4. Dialysis staff supports me in coping with my kidney disease.c | Definitely true/mostly true | Do not know/mostly false/definitely false | Dissatisfied with staff support |
| Other psychosocial factors | |||
| 5. During the past 4 weeks, how much of the time has your physical health or emotional problems interfered with your social activities?a | A little of the time/none of the time | All of the time/most of the time/some of the time | Health has interfered with social activities |
| 6. I feel like a burden on my family.a | Do not know/mostly false/definitely false | Definitely true/mostly true | Felt like a burden to family |
| 7. How much of the time during the past 4 weeks did you isolate yourself from people around you?b | None of the time/a little of the time/some of the time | A good bit of the time/most of the time/all of the time | Isolated from people |
DOPPS I, II and III;
DOPPS I;
DOPPS I and III.
As dichotomized for use in regression models.
Questions 1 to 7 are from KDQoL-SF.
Two questions concerned support from family and friends (“dissatisfied with family time”; “dissatisfied with family support”), and two questions concerned encouragement and support from dialysis staff (“dissatisfied with staff encouragement to be independent”; “dissatisfied with staff support”). Psychosocial factors were measured with three questions regarding patients' perception of the consequences of kidney disease on social relations (“health has interfered with social activities”; “felt like a burden to family”; and “isolated from people”; Table 1).
Marital and living status were not primary predictors in this study because these variables have been investigated previously in the DOPPS (11,20); however, they were treated as potential confounders.
Measurement of Adherence, Nutrition, and Physical QoL
Facility staff abstracted from patient medical records five baseline indicators of patient nonadherence to medical care: (1) skipping one or more hemodialysis sessions during the 30-day period before study enrollment, (2) shortening the hemodialysis session by >10 minutes during the 30-day period before enrollment, (3) interdialytic weight gain (IDWG) of >5.7% of dry weight, (4) serum phosphate level >7.5 mg/dl, and (5) serum potassium concentration >6.0 mEq/L. Detailed information about the importance of these measures has been published previously (11). Serum albumin level was also abstracted from patient medical records as a marker of nutrition/inflammation. For analysis, serum albumin level was dichotomized (≥3.5 versus <3.5g/dl).
Physical QoL was derived from four KDQoL-SF subscales (physical functioning, role-physical, bodily pain, and general health) originally developed for the Short-Form Health Survey. Physical QoL scores were based on a scale of 1 to 100 with the higher score indicating better QoL. For analysis, the baseline physical QoL score was dichotomized at the median score (>35.5 versus ≤35.5).
Statistical Methods
Standard descriptive statistics were used to describe prevalent cross-sections of participants at study entry. Cox regression was used to estimate the hazard ratios of associations of social support and other psychosocial factors with mortality and withdrawal from dialysis. For the mortality outcome, we examined death due to (1) all causes, (2) cardiac disease, (3) cachexia, and (4) infections, with the latter three outcomes modeled as the primary cause of death. Logistic regression was used to estimate associations of social support and other psychosocial factors with indicators of adherence measures and physical QoL. All models included the following as covariates (potential confounders): age, sex, race, years with ESRD, marital status, living status, and 13 summary comorbidity classes including physician-diagnosed depression in psychiatric disorder (listed in Table 2), serum albumin, and single-pool Kt/V. Models were stratified by country and study phase and accounted for clustering of outcomes within facilities. Study participants were followed until the earliest of death, transplant, change of treatment modality, withdrawal from dialysis, transfer to another facility, or study end. Time at risk was defined as the period from study enrolment until the end of patient follow-up. All statistical analyses were performed using SAS software, version 9.1 (SAS Institute, Cary, NC).
Table 2.
Baseline demographic and clinical characteristics of DOPPS participants overall and by geographic region
| Characteristic | Overall (n = 22,451)c | North America (n = 6807) | Europe-Australia/New Zealand (n = 10,144) | Japan (n = 5500) | Percent Missingb |
|---|---|---|---|---|---|
| Demographics | |||||
| age, years (mean, SD) | 61.5 (14.5) | 60.8 (15.3) | 62.5 (14.8) | 60.4 (12.6) | – |
| male (%) | 57.5 | 53.6 | 58.3 | 60.8 | – |
| African American (%) | 10.7 | 32.8 | 1.5 | 0.0 | – |
| time on dialysis (mean, SD) | 5.1 (5.7) | 3.5 (3.9) | 4.9 (5.6) | 7.6 (6.8) | – |
| Comorbidities (%) | |||||
| coronary heart disease | 43.6 | 56.8 | 43.5 | 27.5 | – |
| congestive heart failure | 30.9 | 43.9 | 30.7 | 15.0 | – |
| other cardiac conditions | 35.4 | 35.6 | 39.0 | 28.6 | – |
| hypertension | 77.1 | 87.2 | 77.0 | 64.8 | – |
| peripheral vascular disease | 24.2 | 28.4 | 27.6 | 12.7 | – |
| cerebrovascular disease | 15.7 | 17.5 | 16.2 | 12.7 | – |
| diabetes | 33.3 | 48.5 | 26.0 | 28.2 | – |
| lung disease | 10.4 | 14.8 | 12.0 | 1.9 | – |
| cancer, excluding skin | 10.9 | 11.7 | 12.6 | 6.5 | – |
| gastrointestinal bleeding | 5.6 | 7.6 | 5.2 | 3.7 | – |
| neurologic disease | 9.0 | 11.6 | 8.9 | 5.8 | – |
| psychiatric disease | 15.5 | 22.7 | 17.4 | 2.9 | – |
| recurrent cellulitis | 7.2 | 10.4 | 7.5 | 2.7 | – |
| Education (%)a | 4 | ||||
| high school or less | 83.5 | 74.9 | 88.1 | 85.8 | |
| attended college | 16.5 | 25.1 | 11.9 | 14.2 | |
| Occupational status (%)a | 2 | ||||
| employed | 16.6 | 10.4 | 11.9 | 32.8 | |
| retired | 40.9 | 38.6 | 56.0 | 15.8 | |
| disabled | 14.7 | 28.2 | 11.5 | 3.9 | |
| unemployed | 9.5 | 11.3 | 5.0 | 15.8 | |
| homemaker, never employed | 14.8 | 8.8 | 13.3 | 25.1 | |
| unknown | 3.5 | 2.8 | 2.3 | 6.7 | |
| Living status (%)a | 2 | ||||
| live alone | 16.4 | 19.1 | 18.1 | 10.0 | |
| live with family/friends | 79.6 | 74.8 | 78.4 | 87.8 | |
| live in nursing home | 4.0 | 6.1 | 3.5 | 2.2 | |
| Marital status (%)a | – | ||||
| single | 17.1 | 21.0 | 16.5 | 13.1 | |
| married | 60.9 | 49.9 | 61.7 | 73.0 | |
| widowed | 15.2 | 18.2 | 16.3 | 9.4 | |
| divorced | 6.9 | 10.9 | 5.4 | 4.5 | |
| Nonadherence measure (%) | |||||
| skipped ≥1 hemodialysis session/month | 4.1 | 8.4 | 1.9 | 3.3 | 7 |
| shortened session by ≥10 minutes | 12.0 | 20.5 | 10.7 | 5.0 | 7 |
| IDWG >5.7% of dry weight | 10.1 | 12.1 | 6.5 | 13.8 | 11 |
| serum phosphate >7.5 mg/dl | 11.9 | 13.7 | 11.2 | 11.1 | 4 |
| serum potassium >6 mEq/L | 10.9 | 5.0 | 15.7 | 9.2 | 4 |
| serum albumin <3.5 g/dl | 21.3 | 22.4 | 23.0 | 16.8 | 12 |
| single-pool Kt/V < 1.2 | 24.2 | 17.9 | 25.2 | 30.1 | 24 |
Columns in designated category add to 100%.
≤1% missing values except as listed. Percentages shown in other columns are among patients without missing data.
Restricted to a representative, prevalent cross-section of patients enrolled at the start of DOPPS phase I, II, and III; replacement placements (n = 9881), enrolled during study follow-up, are excluded from this table but included in subsequent models.
The percentage missing for each variable (Tables 2 and 3) ranged from 1% to 24%. To handle missing data analytically, IVEware software was used to perform multiple imputation using the sequential regression imputation method (23). Demographics, socioeconomic history, 13 summary comorbidities, indicators of patient nonadherences (all variables reported in Table 2), and physical QoL were used to impute missing baseline values of patient self-reported items (social support and other psychosocial factors) and covariate adjustments. Missing outcomes (physical QoL, indicators of patient nonadherence) were not imputed. Five samples of imputed data were generated. For each analysis, five estimates specific to each data set were generated and pooled to obtain the final estimate, with variance estimated using Rubin's formula (24).
Table 3.
Percent distribution of social support and other psychosocial factors overall and by geographic region
| Percent of Patients |
|||||
|---|---|---|---|---|---|
| Overall (n = 32,332) | North America (n = 11,159) | Europe-Australia/New Zealand (n = 14,566) | Japan (n = 6607) | Missinga | |
| Dissatisfied with family time | 23.8 | 29 | 20.6 | 16.2 | 9 |
| Dissatisfied with family support | 12.0 | 13.4 | 10.4 | 11.2 | 9 |
| Dissatisfied with staff encouragement to be independent | 21.4 | 20 | 18.7 | 29.3 | 13 |
| Dissatisfied with staff support | 14.1 | 12.9 | 11.5 | 21.1 | 12 |
| Health has interfered with social activities | 53.5 | 54.1 | 58.6 | 41 | 5 |
| Felt like a burden to family | 45.6 | 44.5 | 40.7 | 58.2 | 6 |
| Isolated from people | 12.0 | 12.3 | 13.5 | 9 | 8 |
Percentages shown in other columns are among patients without missing data.
Results
Patient Characteristics
Baseline clinical characteristics are presented in Table 2. The mean age was 61.5 years and the average time on dialysis was 5.1 years. The median follow-up time in the survival analysis was 1.76 years (maximum 5.4 years). More than half (60.9%) of the patients were married and 79.6% lived with family or friends. Concerning adherence indicators, 4.1% of patients skipped one or more hemodialysis sessions in a month, 12.0% shortened hemodialysis session by ≥10 minutes in a month, 10.1% had excessive IDWG, 11.9% had hyperphosphatemia, and 10.9% had hyperkalemia. 21.3% of patients had an albumin level <3.5 g/dl.
It is interesting to note that patients from Japan had distinct characteristics regarding marital, living, and occupational status compared with patients in North America and Europe-Australia/New Zealand. A high percentage of Japanese patients were married and lived with family and friends. Most patients were employed or a homemaker, whereas in the other regions most patients were retired.
Distributions of Social Support and Other Psychosocial Factors
The percentage distribution of patient self-report of social support and other psychosocial factors by region are presented in Table 3. Notably, across DOPPS phases (I to III), nearly half of the patients felt like a burden to family and felt that health has interfered with social activities. Regional specificities were observed. In Japan, more patients felt like a burden to family, but a smaller fraction felt that health interfered with social activities compared with the two other regions. Moreover, in Japan, patients expressed less dissatisfaction with family time and support, although they were more dissatisfied with staff encouragement and support.
All-Cause Mortality
All-cause mortality was higher in patients who reported feeling that their health has interfered with social activities (hazard ratio [HR] = 1.33; 95% confidence interval [CI] 1.26 to 1.40), isolated from people (HR = 1.28; 95% CI 1.12 to 1.45), felt like a burden to family (HR = 1.30; 95% CI 1.23 to 1.37), dissatisfied with family time (HR = 1.13; 95% CI 1.03 to 1.25), and dissatisfied with family support (HR = 1.14; 95% CI 1.02 to 1.28) (Table 4). The association of all-cause mortality with being dissatisfied with family support appeared to be substantially stronger in Japan (HR = 2.1; 95% CI 1.25 to 3.53) than in North America or in the Europe-Australia/New Zealand study region (Figure 1). For all of the other factors, the relationship of each social support and psychosocial factor with all-cause mortality was consistent and similar across regions and was consistent with the overall results stratified by country.
Table 4.
Associations of patient self-reported social support and other psychosocial factors with mortality
| HR (95% CI) |
||||
|---|---|---|---|---|
| All-Cause Mortality | Cause-Specific Mortality |
|||
| Cardiacg | Cachexiah | Infectioni | ||
| Dissatisfied with family timeb | 1.13 (1.03 to 1.25)f | 1.13 (0.98 to 1.31) | 1.29 (0.59 to 2.84) | 1.23 (0.93 to 1.64) |
| Dissatisfied with family supportb | 1.14 (1.02 to 1.28)f | 1.12 (0.91 to 1.39) | 1.43 (0.53 to 3.84) | 1.12 (0.77 to 1.64) |
| Dissatisfied with staff encouragement to be independentc | 1.04 (0.96 to 1.13) | 1.05 (0.93 to 1.19) | 1.19 (0.69 to 2.07) | 1.11 (0.90 to 1.37) |
| Dissatisfied with staff supportc | 1.08 (0.98 to 1.19) | 1.03 (0.87 to 1.21) | 1.21 (0.51 to 2.87) | 1.05 (0.77 to 1.45) |
| Health has interfered with social activitiesa | 1.33 (1.26 to 1.40)d | 1.28 (1.16 to 1.41)d | 1.57 (1.01 to 2.45)f | 1.35 (1.16 to 1.58)d |
| Felt like a burden to familya | 1.30 (1.23 to 1.37)d | 1.23 (1.12 to 1.35)d | 1.45 (1.02 to 2.08)f | 1.36 (1.17 to 1.58)d |
| Isolated from peopleb | 1.28 (1.12 to 1.45)e | 1.13 (0.94 to 1.36) | 1.73 (0.76 to 3.97) | 1.25 (0.87 to 1.81) |
Pooled across geographic regions. Region-specific mortality is in Figure 1. All models were adjusted for age, gender, race, time on dialysis, marital status, living status, 13 summary comorbidity classes, serum albumin, and single-pool Kt/V; stratified by country and phase; and accounted for facility clustering. Reference categories are in Table 1 (category A).
In DOPPS I, II, and III: n = 32,046;
In DOPPS I: n = 12,356;
In DOPPS I and III: n = 21,592.
P < 0.0001;
P = 0.0001 to <0.01;
P = 0.01 to <0.05.
Cardiac: primary cause of death reported as acute myocardial infarction, hyperkalemia, pericarditis including cardiac tamponade, atherosclerotic heart disease, cardiomyopathy, cardiac arrhythmia, cardiac arrest of unknown cause, valvular heart disease, pulmonary edema due to exogenous fluid, or congestive heart failure.
Cachexia: primary cause of death reported as cachexia.
Infections: primary cause of death reported as septicemia due to vascular access or peritonitis or peripheral vascular disease (gangrene) or other causes, pulmonary infection (bacterial, fungal, or other), viral infection (CMV or other), tuberculosis, AIDS, or infections of other cause.
Figure 1.
Association of patient self-reported social support and other psychosocial factors with all-cause mortality, by region. Reference categories are listed in Table 1 (category A). DOPPS I, II, and III: n = 32,046. DOPPS I: n = 12,356. DOPPS I and III: n = 21,592. All models were adjusted for age, gender, race, time on dialysis, marital status, living status, 13 summary comorbidity classes, serum albumin, and single-pool Kt/V; stratified by phase; and accounted for facility clustering. aIn DOPPS I, II, and III: n = 32,046; cIn DOPPS I and III: n = 21,592. dP < 0.0001; eP = 0.0001 to <0.01; fP = 0.01 to <0.05.
Cause-Specific Mortality
Health interference with social activities and feeling like a burden to family were significantly associated with cardiac, cachexia, and infection-related mortalities. The highest magnitude was observed for mortality due to cachexia (Table 4).
Withdrawal from Dialysis
Associations with withdrawal from dialysis are presented by region (North America and Europe-Australia/New Zealand) in Figure 2. Japan was excluded when shown by region because it has very few withdrawal events. Health interference with social activities and feeling like a burden to family were associated with higher rates of withdrawal in North America and Europe-Australia/New Zealand. Dissatisfaction with family time was only associated with a higher rate of withdrawal in North America.
Figure 2.
Association of patient self-reported social support and other psychosocial factors with withdrawal from dialysis by region. Reference categories are listed in Table 1 (category A). DOPPS I, II, and III: n = 32,046. DOPPS I: n = 12,356. DOPPS I and III: n = 21,592. All models were adjusted for age, gender, race, time on dialysis, marital status, living status, 13 summary comorbidity classes, serum albumin, and single-pool Kt/V; stratified by phase; and accounted for facility clustering. Because very few withdrawal events are reported, Japan was not shown. aIn DOPPS I, II, and III: n = 32,046; bIn DOPPS I: n = 12,356. cIn DOPPS I and III: n = 21,592. dP < 0.0001; eP = 0.0001 to <0.01; fP = 0.01 to <0.05.
Indicators of Nonadherence
Lower levels of patient self-reported social support and other psychosocial factors were associated with excessive IDWG and higher likelihood to shorten hemodialysis session and to have hyperkalemia (Table 5). Regional variations were noted. In North America, dissatisfaction with family time was associated with higher likelihood to skip a hemodialysis session (HR = 1.39; 95% CI 1.39 to 1.83). Feeling like a burden to family was associated with higher likelihood to shorten hemodialysis session (HR = 1.14; 95% CI 1.01 to 1.28). In Europe-Australia/New Zealand, feeling like a burden to family was associated with higher likelihood to shorten hemodialysis session (HR = 1.13; 95% CI 1.13 to 1.26) and to have hyperkalemia (HR = 1.15; 95% CI 1.03 to 1.27). In Japan, patients' isolation and dissatisfaction with family time and support were associated with excessive IDWG (HR = 2.01; 95% CI 1.28 to 3.14; HR = 1.51; 95% CI 1.07 to 2.14; HR = 1.50; 95% CI 1.00 to 2.24).
Table 5.
Associations of social support and other psychosocial factors with indicators of patient nonadherence
| Odds Ratios by Nonadherence Indicator (95% CI) |
||||||
|---|---|---|---|---|---|---|
| IDWG >5.7% of Dry Weight | Skipped ≥1 HD Session/Month | Shorten Session by ≥10 Minutes | Phosphate > 7.5 mg/dl | Potassium > 6 mEq/L | Albumin <3.5 g/dl | |
| Dissatisfied with family timeb | 1.18 (0.96 to 1.45) | 1.13 (0.93 to 1.37) | 1.06 (0.92 to 1.22) | 0.93 (0.82 to 1.07) | 1.09 (0.93 to 1.28) | 1.10 (0.97 to 1.24) |
| Dissatisfied with family supportb | 1.27 (1.03 to 1.58)f | 0.91 (0.71 to 1.17) | 1.21 (0.99 to 1.47)g | 0.98 (0.82 to 1.16) | 1.14 (0.95 to 1.37) | 0.96 (0.82 to 1.12) |
| Dissatisfied with staff encouragement to be independentc | 1.15 (1.01 to 1.30)f | 1.10 (0.93 to 1.29) | 1.03 (0.93 to 1.15) | 1.00 (0.89 to 1.11) | 0.95 (0.85 to 1.08) | 0.97 (0.88 to 1.05) |
| Dissatisfied with staff supportc | 1.11 (0.96 to 1.29) | 0.99 (0.81 to 1.21) | 1.06 (0.94 to 1.20) | 1.01 (0.89 to 1.14) | 0.97 (0.85 to 1.10) | 0.93 (0.84 to 1.04) |
| Health has interfered with social activitiesa | 1.12 (1.01 to 1.23)f | 1.05 (0.93 to 1.18) | 1.06 (0.98 to 1.15) | 1.06 (0.98 to 1.14) | 1.05 (0.96 to 1.14) | 1.23 (1.16 to 1.31)d |
| Felt like a burden to familya | 1.01 (0.92 to 1.11) | 1.07 (0.96 to 1.19) | 1.12 (1.04 to 1.21)e | 0.98 (0.91 to 1.05) | 1.10 (1.01 to 1.20)f | 1.22 (1.16 to 1.30)d |
| Isolated from peopleb | 1.22 (0.99 to 1.50)g | 1.00 (0.79 to 1.27) | 1.03 (0.87 to 1.22) | 0.95 (0.79 to 1.15) | 0.98 (0.79 to 1.22) | 1.18 (1.02 to 1.37)f |
All models were adjusted for age, gender, race, time on dialysis, marital status, living status, 13 summary comorbidity classes, serum albumin, single-pool Kt/V, country, and phase and accounted for facility clustering. Reference categories are in Table 1 (category A).
In DOPPS I, II, and III: n = 25,839 to 30,020;
In DOPPS I: n = 9206 to 10,807;
In DOPPS I and III: n = 17,354 to 20,309.
P < 0.0001;
P = 0.0001 to <0.01;
P = 0.01 to <0.05;
P = 0.05 to 0.1.
Albumin Levels
Lower psychosocial factor scores were associated with low serum albumin levels (Table 5). For each region, health interference with social activities and feeling like a burden to family were associated with low serum albumin levels.
Physical QoL
Lower levels of all seven patient self-reported social support and other psychosocial factors were associated with poor physical QoL (P < 0.0001) (Table 6). The poor physical QoL results within each region were consistent with the overall estimates across regions.
Table 6.
Associations of social support and other psychosocial factors with physical QoL
| OR by Physical QoL Score (95% CI) Physical Component Summary Score >35.5 versus ≤35.5 | |
|---|---|
| Dissatisfied with family timeb | 0.54 (0.48 to 0.61)d |
| Dissatisfied with family supportb | 0.74 (0.64 to 0.86)d |
| Dissatisfied with staff encouragement to be independentc | 0.73 (0.67 to 0.80)d |
| Dissatisfied with the staff supportc | 0.76 (0.69 to 0.84)d |
| Health has interfered with social activitiesa | 0.25 (0.24 to 0.27)d |
| Felt like a burden to familya | 0.40 (0.38 to 0.43)d |
| Isolated from peopleb | 0.39 (0.34 to 0.46)d |
All models were adjusted for age, gender, race, time on dialysis, marital status, living status, 13 summary comorbidity classes, serum albumin, single-pool Kt/V, country, and phase and accounted for facility clustering. Reference categories are in Table 1 (category A).
In DOPPS I, II, and III: n = 25,759 to 26,650;
In DOPPS I: n = 9213 to 9664;
In DOPPS I and III: n = 17,612 to 17,279.
P < 0.0001.
Sensitivity Analyses
The primary analyses adjusted for diagnosis of any psychiatric disorder. Physician-diagnosed depression comprised 78% of these psychiatric disorders. The associations of social support and other psychosocial factors with mortality were similar when adjusted for physician-diagnosed depression instead of psychiatric disorder. In another analysis adjusted for patient-reported depressive symptoms (using Center for Epidemiologic Studies Depression Scale [CES-D]), the associations of social support and other psychosocial factors with mortality were similar, although modestly attenuated. For the mortality model, no interactions between social support/psychosocial factors and psychiatric disorder, depression, or CES-D score were observed.
Discussion
This study shows strong associations between poorer social support, other psychosocial factors and a higher mortality rate, greater likelihood of nonadherence with medical care, and worse physical QoL. A strength of our study is that it is based on a much larger representative population than previous reports because it has international and multicultural dimensions.
Family and staff support measures were associated with higher mortality rates, a finding similar to that described previously by Christensen et al. (4) in a much smaller sample of 78 hemodialysis patients. Reporting that health has interfered with social activities, feeling isolated from people, feeling like a burden to family, and dissatisfaction with family time and support were associated with higher risk of all-cause mortality. These results are consistent with those from Kimmel et al. (5) who found that patients' social support and patients' perception of illness intrusiveness predict all-cause mortality. The study presented here shows two interesting results. First, regarding primary cause of mortality, social support measures seem to be less strongly associated with mortality than other psychosocial factors. Second, a cultural specificity was observed in Japan: being dissatisfied with family support was associated with an especially high mortality rate.
Previous studies have shown associations of social support with all-cause death but not with death due to specific causes. The study presented here gives additional insight into associations between psychosocial factors and mortality by demonstrating associations with a wide range of causes including cardiac causes, cachexia, and infection. Poorer psychosocial factors, as assessed by the interference of health/emotional problems with social activities and the feeling of being a burden to the family, were also associated with higher risk of withdrawal from dialysis.
Moreover, this study underlines the association between psychosocial factors and albumin level, an important biologic marker of nutritional status and possible inflammation. Health has interfered with social activities, feeling isolated from people, and felt like a burden to family were all associated with lower serum albumin levels. Previous studies have shown that low albumin and other markers of malnutrition/inflammation are associated with elevated overall and cardiovascular mortality (25). Additional studies are needed to assess whether hypoalbuminemia is possibly on the causal pathway between social factors and mortality.
Lower levels of family support and other psychosocial factors were associated with higher risk of nonadherence with medical care. These patterns of associations are consistent with those of previous studies (12,13). In the research presented here, cultural specificities were observed. In North America, dissatisfaction with family time and feeling like a burden to family were associated with nonadherence to prescribed hemodialysis sessions. In Europe-Australia/New Zealand, feeling like a burden to family was associated with nonadherence to hemodialysis session length and hyperkalemia. In Japan, patients' isolation and dissatisfaction with family time and support were associated with excessive IDWG. It is surprising to notice that staff encouragement and support were not associated with patient adherence and may not be able to overcome the possible stronger influence of other psychosocial factors and family support.
Few studies have examined associations between social relations and QoL in hemodialysis patients. The research presented here shows that higher social support and other psychosocial factors are associated with better physical QoL; these links were also found for each patient self-reported measure. Moreover, these associations were similar in North America, Europe-Australia/New Zealand, and Japan. These findings are consistent with observations from other studies (17,25) and emphasize the importance of family and staff support for the QoL of hemodialysis patients across cultures. A greater focus on increasing social support may improve QoL for hemodialysis patients and contribute to reducing rates of death and hospitalization in this patient population (10). We chose not to study the relations between social support and mental QoL because this component of QoL consists of social support items and is thus intrinsically correlated.
Several limitations in our study warrant mention. First, social support was measured with limited items from the KDQoL. Although future studies addressing similar questions should ideally use well established instruments that have proven reliable and valid, the reliability and validity of social support instruments that capture factors specific to hemodialysis and the lives of hemodialysis patients have not been established. Second, this research focused on one dimension of social support (i.e., perceptions about relationships with others). Other dimensions of social support (e.g., social network size, frequency of contact with family and friends, participation in group activities, etc.) merit additional study. Third, it was not possible to assess whether the levels of social support and other psychosocial factors preceded or followed the initiation of dialysis, which should be viewed as a limitation especially for associations with adherence indicators and other outcomes assessed cross-sectionally at the initiation of the study. Social support and other psychosocial factors may vary over time and may be important to consider at the initiation of dialysis therapy. Nevertheless, our findings indicate that crude psychosocial indicators assessed at study entry inclusion can predict subsequent clinical outcomes in hemodialysis patients, particularly death and withdrawal from dialysis. A fourth limitation is confounding by health status because healthier patients may perceive higher levels of support, whereas sicker patients may believe they are a greater burden to others. Finally, the items measuring psychosocial factors may also indicate state of physical health. All indicators of health cannot be accounted for in observational studies.
Our research underscores the links between people's resources and their resiliency (26), with social support being considered here as a resource. In relation to the conservation of resources theory (27), people seek to obtain, retain, and protect resources. Therefore, the lack of resources tends to undermine patients' positive psychologic outcomes. Like a spiral of loss, illness can mobilize a patient's personal and social resources (e.g., physical vigor, emotional robustness, cognitive ability, family love and availability). When confronted with stressful situations such as hemodialysis, the ability to seek and obtain support is necessary and salutogenic. Thus, the presence of supportive people who can be confided in and whose caring is deemed important must be one of the most salient types of support.
In summary, this large international study demonstrates that several indicators of social support and other psychosocial factors are associated with mortality, adherence, nutrition, and physical QoL. In addition, the results corroborate the hypothesis that social-support interventions are needed to enhance patient global care (28). Feeling like a burden to family was consistently linked to the different adverse outcomes of our study, and interventions should probably focus on this feeling. The guilt patients often feel about being ill might prevent them from asking for support. Further studies are needed on the effect of family members' involvement on patients' feelings of being a burden, sense of support, and clinical outcomes.
Disclosures
None.
Acknowledgments
DOPPS is administered by Arbor Research Collaborative for Health and is supported by scientific research grants from Amgen (since 1996), Kyowa Hakko Kirin (since 1999, in Japan), Genzyme (since 2009), and Abbott (since 2009) without restrictions on publications. Shauna Leighton of the Arbor Research Collaborative for Health provided editorial assistance.
Footnotes
Published online ahead of print. Publication date available at www.cjasn.org.
References
- 1. Kimmel PL: Psychosocial factors in dialysis patients. Kidney Int 59: 1599–1613, 2001 [DOI] [PubMed] [Google Scholar]
- 2. House JS, Landis KR, Umberson D: Social relationships and health. Science 241: 540–545, 1988 [DOI] [PubMed] [Google Scholar]
- 3. Cohen S, Underwood L, Gottlieb B: Measuring and intervening in social support. New York, Oxford University, Press, 2000 [Google Scholar]
- 4. Christensen AJ, Wiebe JS, Smith TW, Turner CW: Predictors of survival among hemodialysis patients: Effect of perceived family support. J Health Psychol 13: 521–525, 1994 [DOI] [PubMed] [Google Scholar]
- 5. Kimmel PL, Peterson RA, Weihs KL, Simmens SJ, Alleyne S, Cruz I, Veis JH: Psychosocial factors, behavioral compliance and survival in urban hemodialysis patients. Kidney Int 54: 245–254, 1998 [DOI] [PubMed] [Google Scholar]
- 6. Thong MSY, Kaptein AA, Krediet RT, Boeschoten EW, Dekker FW: Social support predicts survival in dialysis patients. Nephrol Dial Transplant 22: 845–850, 2007 [DOI] [PubMed] [Google Scholar]
- 7. DeOreo PB: Hemodialysis patient-assessed functional health status predicts continued survival, hospitalization, and dialysis-attendance compliance. Am J Kidney Dis 30: 204–212, 1997 [DOI] [PubMed] [Google Scholar]
- 8. Leggat JE, Hulbert-Shearon TE, Golper TA, Jones CA, Held PJ, Port FK: Noncompliance in hemodialysis: Predictors and survival analysis. Am J Kidney Dis 32: 139–145, 1998 [DOI] [PubMed] [Google Scholar]
- 9. Lopes AA, Bragg-Gresham JL, Satayathum S, McCullough K, Pifer T, Goodkin DA, Mapes DL, Young EW, Wolfe RA, Held PJ, Port FK; Worldwide Dialysis Outcomes and Practice Patterns Study Committee: Health-related quality of life and associated outcomes among hemodialysis patients of different ethnicities in the United States: The Dialysis Outcomes and Practice Patterns Study (DOPPS). Am J Kidney Dis 41: 605–615, 2003 [DOI] [PubMed] [Google Scholar]
- 10. Mapes DL, Lopes AA, Satayathum S, McCullough KP, Goodkin DA, Locatelli F, Fukuhara S, Young EW, Kurokawa K, Saito A, Bommer J, Wolfe RA, Held PJ, Port FK: Health-related quality of life as a predictor of mortality and hospitalization: The Dialysis Outcomes and Practice Patterns Study (DOPPS). Kidney Int 64: 339–349, 2003 [DOI] [PubMed] [Google Scholar]
- 11. Saran R, Bragg-Gresham JL, Rayner HC, Goodkin DA, Keen ML, Van Dijk PC, Kurokawa K, Piera L, Saito A, Fukuhara S, Young EW, Held PJ, Port FK: Nonadherence in hemodialysis: Associations with mortality, hospitalization, and practice patterns in the DOPPS. Kidney Int 64: 254–262, 2003 [DOI] [PubMed] [Google Scholar]
- 12. Christensen AJ, Smith TW, Turner CW, Holman JM, Jr, Gregory MC, Rich MA: Family support, physical impairment, and adherence in hemodialysis: An investigation of main and buffering effects. J Behav Med 15: 313–325, 1992 [DOI] [PubMed] [Google Scholar]
- 13. Kara B, Caglar K, Kilic S: Nonadherence with diet and fluid restrictions and perceived social support in patients receiving hemodialysis. J Nurs Scholarsh 39: 243–248, 2007 [DOI] [PubMed] [Google Scholar]
- 14. Kutner NG, Zhang R, McClellan WM, Cole SA: Psychosocial predictors of non-compliance in haemodialysis and peritoneal dialysis patients. Nephrol Dial Transplant 17: 93–99, 2002 [DOI] [PubMed] [Google Scholar]
- 15. Leggat JE: Adherence with dialysis: A focus on mortality risk. Semin Dial 18: 137–141, 2005 [DOI] [PubMed] [Google Scholar]
- 16. Helgeson VS: Social support and quality of life. Qual Life Res 12: 25–31, 2003 [DOI] [PubMed] [Google Scholar]
- 17. Tell GS, Mittelmark MB, Hylander B, Shumaker SA, Russell G, Burkart JM: Social support and health-related quality of life in black and white dialysis patients. ANNA J 22: 301–309, 1995 [PubMed] [Google Scholar]
- 18. Vazquez I, Valderrabano F, Fort J, Jofré R, López-Gómez JM, Moreno F, Sanz-Guajardo D; Spanish Cooperative Renal Patients Quality of Life Study Group: Psychosocial factors and health-related quality of life in hemodialysis patients. Qual Life Res 14: 179–190, 2005 [DOI] [PubMed] [Google Scholar]
- 19. Young EW, Goodkin DA, Mapes DL, Port FK, Keen ML, Chen K, Maroni BL, Wolfe RA, Held PJ: The Dialysis Outcomes and Practice Patterns Study (DOPPS): An international hemodialysis study. Kidney Int 57: 74–81, 2002 [Google Scholar]
- 20. Lopes AA, Bragg-Gresham JL, Goodkin DA, Fukuhara S, Mapes DL, Young EW, Gillespie BW, Akizawa T, Greenwood RN, Andreucci VE, Akiba T, Held PJ, Port FK: Factors associated with health-related quality of life among hemodialysis patients in the DOPPS. Qual Life Res 16: 545–557, 2007 [DOI] [PubMed] [Google Scholar]
- 21. Pisoni RL, Gillespie BW, Dickinson DM, Chen K, Kutner MH, Wolfe RA: The Dialysis Outcomes and Practice Patterns Study (DOPPS): Design, data elements, and methodology. Am J Kidney Dis 44: S7–S15, 2004 [DOI] [PubMed] [Google Scholar]
- 22. Diaz-Buxo JA, Lowrie EG, Lew NL, Zhang H, Lazarus JM: Quality-of-life evaluation using Short Form 36: Comparison in hemodialysis and peritoneal dialysis patients. Am J Kidney Dis 35: 293–300, 2000 [DOI] [PubMed] [Google Scholar]
- 23. Raghunathan TE, Lepkowski JM, van Hoewyk J, Solenberger PW: A multivariate technique for multiply imputing missing values using a sequence of regression models. Surv Methodol 27: 85–95, 2001 [Google Scholar]
- 24. Little RJA, Rubin DB: Statistical Analysis with Missing Data, Hoboken, NJ, John Wiley & Sons, 2002 [Google Scholar]
- 25. Kalantar-Zadeh K, Block G, McAllister CJ, Humphreys MH, Kopple JD: Appetite and inflammation, nutrition, anemia, and clinical outcome in hemodialysis patients. Am J Clin Nutr 80: 299–307, 2004 [DOI] [PubMed] [Google Scholar]
- 26. Seligman MEP, Csikszentmihalyi M: Positive psychology. Am Psychol 55: 5–14, 2000 [DOI] [PubMed] [Google Scholar]
- 27. Hobfoll SE: Conservation of resources: A new attempt at conceptualizing stress. Am Psychol 44: 513–524, 1989 [DOI] [PubMed] [Google Scholar]
- 28. Patel SS, Peterson RA, Kimmel PL: The impact of social support on end-stage renal disease. Semin Dial 18: 98–102, 2005 [DOI] [PubMed] [Google Scholar]