Patient-reported mental health problems and clinical outcomes in adults with non-dialysis chronic kidney disease

Ethics statement

This study was conducted by the principles of the Declaration of Helsinki and the study protocol was approved by the Institutional Review Board of the Kangbuk Samsung Hospital in Seoul, Korea (No. 2011-01-076-001). Written informed consents were acquired from all participants.

Study cohort

The KNOW-CKD is a prospective cohort study investigating the clinical outcomes of Koreans with non-dialysis CKD. The study design, methods, and protocol summary are detailed elsewhere [23,24]. This cohort enrolled 2,238 adults between the ages of 20 and 75 years with non-dialysis CKD stages G1 to G5 between 2011 and 2016. We excluded patients with any of the following: unwillingness to provide written consent, inability to fill in self-reported questionnaires, previous dialysis, heart failure (New York Heart Association class 3 or 4), liver cirrhosis, pregnancy, history of malignancy, or severe mental illness including schizophrenia, bipolar disorder, and major depressive disorder. We analyzed 1,879 participants from this cohort who underwent extensive laboratory tests, completed health questionnaires, and had available follow-up clinical outcomes data.

Clinical data and health-related quality of life

Baseline demographics and laboratory data were retrieved from the electronic data management system (Phacta X). Demographic characteristics, medical history, and lifestyle factors were collected from patient-reported questionnaires and a review of the medical records. Serum creatinine was measured at a central lab using an isotope dilution mass spectrometry-calibrated method. The estimated glomerular filtration rate (eGFR) was calculated using the CKD-EPI (CKD Epidemiology Collaboration) creatinine equation. Urinary albumin excretion was determined using the spot urine albumin-to-creatinine ratio (ACR). We collected the following characteristics through patient-reported questionnaires: educational level, marital status, current smoking, alcohol drinking, mental health, and physical activities using the International Physical Activity Questionnaire. HRQOL was measured using the KDQOL-SF, which is composed of 43 multidomain kidney disease component summary (KDCS) and generic 36-item health survey as physical component summary (PCS), and mental component summary (MCS) [25].

Assessment of exposure: patient-reported mental health problems

Data about mental health problems were obtained using a patient-reported mental health questionnaire from the KNHANES. We evaluated four indicators of mental health: sleep duration, perceived stress, depressive symptoms, and suicidal ideation [22]. Sleep duration was identified using the question, “How many hours do you sleep a day?” Sleep duration from 6 to 9 hours was deemed adequate, and short (5 hours) or long (≥9 hours) was deemed inadequate [26]. Perceived stress was assessed using the question, “How much stress do you feel in your daily life?” The subjective distress was those who answered “very much” or “much”; the unstressed was those who answered “a little” or “none.” Depressive symptoms were identified by a positive answer to the question, “Have you experienced sadness or despair severe enough to interfere with daily life for 2 consecutive weeks or longer in the last 12 months?” Suicidal ideation was assessed by a positive answer to the question, “Have you thought about wanting to die in the past 12 months?” We categorized the study participants into three groups by the numbers of mental health problems; adults of good, fair, and poor mental health had 0, 1, and 2 to 4 mental health problems of inadequate sleep duration, subjective distress, depressive symptom, or suicidal ideation.

Assessment of clinical outcomes: end-stage kidney disease and all-cause death

The primary outcomes of interest were incident ESKD and all-cause death during follow-up. ESKD was defined as CKD G5 requiring dialysis or kidney transplantation. The time of kidney replacement therapy and death were collected for a median follow-up of 6.4 years. Patients who reached the endpoints were reported by each center.

Statistical analysis

We described the distribution of each mental health problem by sex, age, and eGFR. KDQOL-SF is a self-reported measure developed for individuals with CKD and those on dialysis [25]. In addition, KDQOL-SF differences between groups of patients with different mental health problems were analyzed using the scaled effect size (ES) of Cohen’s d. The ES of KDQOL-SF domain differences according to the mental health groups was calculated using the following equation.

ES = (Mean_{KDQOL-SFgroup1} - Mean_{KDQOL-SFgroup2})/SD_{KDQOL-SFgroup1}

We interpreted the ES as follows: 0.2 to 0.5 small, 0.5 to 0.8 moderate, and greater than 0.8 was considered large [27].

The cumulative event probabilities were estimated using a Kaplan-Meier analysis and log-rank tests. Cox proportional hazard models were used to investigate the association between mental health problems and clinical outcomes. The data were expressed as hazard ratios (HRs) with 95% confidence intervals (CIs). Model 1 considered baseline age, sex, eGFR, and ACR. Model 2 added body mass index, serum albumin, a history of diabetes mellitus, hypertension, and cardiovascular disease. Model 3 added marital status, education status, employment, current smoking, current alcohol drinking, and health-enhancing physical activity. Because patient-reported mental health problems differed by age, sex, and eGFR, we evaluated the association between mental health, suicidal ideation, and clinical outcomes stratified by age, sex, and eGFR group. All analyses were performed using Stata version 17 (StataCorp LLC).

Baseline characteristics of study participants

The baseline characteristics of the study participants are shown in Table 1 and Supplementary Table 1 (available online). All study participants are Koreans, the mean and standard deviation (SD) of age was 53 ± 12 years, 1,157 participants were males (61.6%), and the mean and SD of eGFR was 54 ± 31 mL/min/1.73 m². The mental health problems of inadequate sleep duration (17.4%), subjective distress (27.3%), depression symptoms (13.2%), and suicidal ideation (16.8%) were common among non-dialysis CKD adults but were not well managed by psychiatric counseling (1.9%) and antidepressant therapy (2.8%). The prevalence of patient-reported mental health problems differed by sex, age, and eGFR categories (Fig. 1). Females had more mental health problems than males. Older adults (≥65 years) had fewer mental health problems than younger adults except for sleep duration. The J-shaped pattern of mental health problems was shown according to eGFR categories.

Table 1.

Characteristics of participants with non-dialysis chronic kidney disease according to patient-reported mental health

Figure 1.

Prevalence of patient-reported mental health.

Inadequate sleep duration, subjective distress, depressive symptoms, and suicidal ideation according to (A) sex, (B) age, and (C) estimated glomerular filtration rate (eGFR) stage.

Characteristics of patient-reported mental health questionnaire

The meaningful clinical interpretation of measuring patient-reported mental health questionnaire differences between groups is to anchor those data to a relevant tool. The KDQOL-SF is a reliable and valid method for collecting individual well-being of health status in patients with CKD. We analyzed KDQOL-SF score differences between the mental health groups and the groups divided by each item using a scaled ES of Cohen’s d. Differences between good and poor mental health groups showed large ES (KDCS = 1.27, PCS = 1.13, and MCS = 1.64) (Supplementary Table 2, available online). Among the mental health items, suicidal ideation showed the largest ES (KDCS = 1.18, PCS = 1.00, and MCS = 1.33) (Supplementary Table 3, available online). Among the KDQOL-SF components, MCS showed the largest ES (Supplementary Table 4, available online). The box plot of the KDQOL-SF subscales scores shows the distributions of numeric data, which gradually decrease according to the mental health groups (Fig. 2). This suggested that the mental health questionnaire of this study was reflected in the clinical effect of mental well-being in non-dialysis CKD.

Figure 2.

Box plot of KDQOL subscales (physical, mental, and kidney disease summary score) according to mental health groups.

Within each box, horizontal black lines denote median values, box extends from the 25th to 75th percentile of each group’s distribution values. Dots denote observations outside the range of 1.5 interquartile range of the 25th and 75th percentile of each group.

KDCS, kidney disease component summary; KDQOL, Kidney Disease Quality of Life; MCS, mental component summary; PCS, physical component summary.

Age-specific mortality according to mental health and suicidal ideation

Crude mortality rates were 10.2 and 15.2 per 1,000 person/yr in adults with good and poor mental health, respectively, and 9.9 and 17.7 per 1,000 person/yr in adults without suicidal ideation and with suicidal ideation, respectively (Supplementary Table 5, available online). Among categories of cause of death, there was no suicide death in the study participants. The age-specific mortality rate according to mental health categories and the presence of suicidal ideation differed more in adults with younger ages than those with older ages.

Association between mental health and risk of end-stage kidney disease and death

The Kaplan-Meier curves revealed that the cumulative probabilities of ESKD and all-cause death during follow-up were higher in patients with poor mental health compared to those with good mental health (log-rank p < 0.05) (Fig. 3). Compared with the good mental health group, the adjusted HRs for ESKD were 1.03 (95% CI, 0.84–1.26) and 1.46 (95% CI, 1.18–1.80) respectively in fair and poor mental health group (model 3). The adjusted HRs for all-cause death were 0.94 (95% CI, 0.62–1.44) and 1.55 (95% CI, 1.04–2.32), respectively in fair and poor mental health group, compared with the good mental health group (model 3) (Table 2).

Figure 3.

Kaplan-Meier curve of time to end-stage kidney disease and all-cause death.

Patient-reported mental health categories (A, B) and suicidal ideation (C, D).

Table 2.

HRs for progress to ESKD and all-cause death according to the patient-reported mental health

Association between each mental health indicator and the risk of end-stage kidney disease and death

Among the four indicators of mental health, suicidal ideation was significantly associated with adverse clinical outcomes. The Kaplan-Meier curves revealed that the cumulative probabilities of the ESKD and all-cause death were higher in adults with suicidal ideation than in adults without suicidal ideation (log-rank p < 0.05) (Fig. 3). The adjusted HRs for ESKD was 1.37 (95% CI, 1.11–1.69) in adults with suicidal ideation compared with adults without suicidal ideation (model 3). The adjusted HRs for all-cause death was 1.98 (95% CI, 1.34–2.92) in adults with suicidal ideation compared with adults without suicidal ideation (model 3) (Table 3). Investigating data linked with the national death registers to ascertain the direct causes of death in the death certification, there was no intentional self-harm in the study participants, but the uncertain cause of death showed a higher rate in adults with poor mental health than those with good mental health (Supplementary Table 6, available online).

Table 3.

HRs for progress to ESKD and all-cause death according to each indicator of the patient-reported mental health

Subgroup analysis by age, sex, and estimated glomerular filtration rate

Because mental health problems differed by age, sex, and eGFR groups, we further examined how modifying those factors affected the risk of ESKD and all-cause death. The association between poor mental health and risk of ESKD and all-cause death was higher in females and younger adults than in males and older adults (p for interaction < 0.001). The association between suicidal ideation and all-cause death was also higher in female and younger adults (p for interaction < 0.001). In the adults with advanced CKD (eGFR <45 mL/min/1.73 m²), the relationship between poor mental health or suicidal ideation and the risk of ESKD and all-cause death was consistent with that in the overall analysis (p < 0.001) (Fig. 4).

Figure 4.

Subgroup analysis.

There were different risks of (A) end-stage kidney disease and (B) all-cause death according to sex, age, and estimated glomerular filtration rate (eGFR) in the multivariable-adjusted Cox proportional hazards model (model 3).

CI, confidence interval; HR, hazard ratio.