STRENGTHS AND LIMITATIONS OF THIS STUDY

A nationally representative sample was employed to probe into the association between the Controlling Nutritional Status (CONUT) score and the risk of all-cause and cause-specific mortality in patients with diabetic kidney disease (DKD).

The association between the CONUT score and the risk of mortality was further assessed in different sex, cardiovascular disease and diabetic retinopathy subpopulations.

The diagnosis of DKD was based on a single measurement of estimated glomerular filtration rate and urinary albumin to creatinine ratio, rather than continuous observation for 3 months.

Some possible confounding factors, such as treatment during follow-up, have not been adjusted for in this analysis.

This research was conducted using the data from the American population, which may affect the applicability of the findings to other populations.

Introduction

Diabetes mellitus is a common metabolic disease, affecting about 537 million people worldwide,1 and type 2 diabetes mellitus (T2DM) accounts for more than 90% of diabetic cases, which can lead to microvascular and macrovascular complications.2 Diabetic kidney disease (DKD) is a main microvascular complication of diabetes,3 which occurs in 30%–40% of patients with diabetes. DKD is the major cause of end-stage renal disease and is associated with a high risk of death, resulting in a serious disease burden.4 5 Thus, investigating effective prognostic markers has important clinical significance for stratified risk management and mortality reduction in DKD.

Recent evidence suggests that DKD is a metabolic-driven immunological disease, with pathological mechanisms involving multiple aspects such as metabolism and inflammation.6 The Controlling Nutritional Status (CONUT) score is a commonly used immuno-nutritional marker that reflects chronic inflammation, immune status and nutritional status in individuals, evaluated jointly by total cholesterol (TC), serum albumin and lymphocyte count.7 8 Mineoka et al 9 reported that a high CONUT score was associated with a greater risk of diabetic foot. In addition, an association was found between malnutrition assessed by the CONUT score and an elevated risk of all-cause mortality among diabetic patients.10 According to a previous study, the CONUT score was significantly correlated with the risk of all-cause death in individuals with chronic kidney disease (CKD) who initiated dialysis.11 Nevertheless, the relationship between the CONUT score and the risk of mortality in patients with DKD is still unknown.

The objective of this study was to probe into the association between the CONUT score and all-cause and cause-specific mortality among patients with DKD, based on the data from the National Health and Nutrition Examination Survey (NHANES) 2009–2018. Subgroup analysis was conducted in terms of sex, cardiovascular disease (CVD) and diabetic retinopathy (DR) to assess whether the association varied among different subpopulations.

MethodsPatient and public involvement

No patients were involved.

Study population

This retrospective cohort study extracted data on patients with DKD from five cycles (2009–2010, 2011–2012, 2013–2014, 2015–2016 and 2017–2018) of the NHANES. The NHANES is a series of studies designed to evaluate the health and nutritional status of the nationally representative, non-institutionalised population in the USA. The survey combines interviews and physical examinations and is approved by the National Center for Health Statistics Research.12 The study involved individuals (1) aged ≥18 years, (2) diagnosed with DKD, (3) with the assessment of serum albumin, TC and total lymphocyte count and (4) with complete survival data. Individuals without data on (1) follow-up time or (2) the cause of death were excluded. Patients were followed up from the data of survey participation to 31 December 2019.

Assessment of DKD

Diabetes was defined as a self-reported diabetes diagnosis, use of diabetes medication or insulin, haemoglobin A1c ≥6.5%, or a fasting glucose level ≥7.0 mmol/L. CKD was defined as urinary albumin to creatinine ratio (UACR)>30 mg/g and/or estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m2 according to the ‘KDIGO 2021 Guidelines’.13 The CKD_EPI_Scr equation was applied to calculate eGFR.13 14 DKD was defined as CKD combined with diabetes.15 16

Calculation of the CONUT score

The CONUT score was calculated with the data on serum albumin, TC and total lymphocyte count (obtained from a blood examination), ranging from 0 to 12.7 A total score of 0–1 was regarded as normal nutrition, and a score of 2–4, 5–8 and 9–12 in total was defined as mild, moderate and severe malnutrition, respectively.

Assessment of mortality

Mortality evaluated in this study included all-cause mortality, CVD-related mortality, diabetes-related mortality and nephropathy-related mortality. All-cause and cause-specific mortality was determined via linkage to the National Death Index until 31 December 2019. The 10th revision of the International Classification of Disease was used to determine the cause of death. All-cause mortality was defined as death from any cause. CVD-related mortality was defined as death from diseases of heart (I00–I09, I11, I13, I20–I51) and cerebrovascular diseases (I60–I69). Diabetes-related mortality was defined as death from diabetes mellitus (E10–E14). Nephropathy-related mortality was defined as death from nephritis, nephrotic syndrome and nephrosis (N00–N07, N17–N19, N25–N27).

Other variables

We collected information on age (years), gender, race (Mexican American, other Hispanic, non-Hispanic white, non-Hispanic black and other race), body mass index (BMI, kg/m2), education level (less than 9th grade, 9th–11th grade, high school graduate/general education development or equivalent, some college or college or associate (AA) degree, college graduate or above), marital status (married, widowed, divorced, separated, never married, living with partner), smoking, alcohol drinking, hypertension, hyperlipidaemia, CVD, DR, poverty income ratio (PIR; <1, ≥1, unknown), physical activity (MET×min/week), fasting glucose (mmol/L), antidiabetics (no, only hypoglycaemic drugs, hypoglycaemic drugs and insulin), diuretics, ACE inhibitor (ACEI), UACR, uric acid (μmol/L), energy (kcal), protein (g), carbohydrate (g), total fat (g), sodium (mg), potassium (mg), eGFR (mL/min/1.73 m2) and follow-up time (months). The presence of hypertension and dyslipidaemia was determined according to laboratory examination, self-reported medical history and medication history. Hypertension refers to systolic blood pressure ≥140 mm Hg, or diastolic blood pressure ≥90 mm Hg, or self-reported hypertension, or use of antihypertensive drugs. Dyslipidaemia refers to TC≥200 mg/dL (5.2 mmol/L), or triglyceride ≥150 mg/dL (1.7 mmol/L), or low-density lipoprotein cholesterol ≥130 mg/dL (3.4 mmol/L), or how-density lipoprotein cholesterol ≤40 mg/dL (1.0 mmol/L), or self-reported hypercholesterolaemia, or receiving lipid-lowering treatment. DR was defined by the question, ‘has a doctor ever told that diabetes has affected eyes or that had retinopathy (ret-in-op-ath-ee)?’ Physical activity was converted into energy consumption. Energy consumption (MET×min)=recommended metabolic equivalent (MET)×exercise time of the corresponding activity (min).

Statistical analysis

Continuous data were illustrated as mean (SE), and the weighted t-test was used for intergroup comparisons; categorical data were reported as the number of cases and the constituent ratio (n (%)), and comparisons between groups were conducted using the χ2 test. Continuous data were standardised. Missing data were filled with multiple imputation and sensitivity analysis was performed for data before and after the imputation (online supplemental table 1).

The included patients were divided into three groups according to the CONUT score: normal nutrition group (a score of 0–1), mild malnutrition group (a score of 2–4), and moderate and severe malnutrition group (a score of 5–12). Univariate weighted Cox regression was employed to screen covariates and examine the association between the CONUT score and mortality. Multivariate weighted Cox regression was used to further explore the association between the CONUT score and mortality, with adjustment for age, race, marital status, smoking, hypertension, CVD, DR, PIR, antidiabetics, diuretics, UACR, uric acid, energy, protein, total fat, sodium and eGFR. Subgroup analysis was carried out based on sex, CVD and DR to assess whether the association between the CONUT score and mortality was different in subpopulations, and the controlled covariates were selected through new univariate weighted Cox regression models. HRs and 95% CIs were calculated.

Data extraction and cleaning were completed by SAS V.9.4 (SAS Institute). R V.4.2.0 (R Foundation for Statistical Computing, Vienna, Austria) was adopted for statistical analysis. A difference was regarded as significantly different when a p<0.05.

ResultsCharacteristics of the study population

From the NHANES 2009–2018, 1723 patients with DKD were enrolled. After excluding patients less than 18 years (n=3), and without information on the CONUT score (n=3) and follow-up time (n=3), a total of 1714 patients were eligible for this study. Figure 1 demonstrates the selection process of eligible patients. According to the CONUT score, the number of patients in normal nutrition group, mild malnutrition group, and moderate and severe malnutrition group was 1119 (65.29%), 553 (32.26%) and 42 (2.45%), respectively. The mean age of these patients was 64.08 years. Most of the patients were non-Hispanic white people (59.62%) and were married (54.30%). The average follow-up time was 58.23 months. The characteristics of the included patients with DKD are presented in table 1. Significant differences were found among the three groups in age, gender, race, BMI, marital status, hypertension, CVD, PIR, physical activity, antidiabetics, diuretics, UACR, energy, protein, carbohydrate, total fat, eGFR, follow-up time and vital status (all p<0.05).

Figure 1

Flow chart of participant selection. CONUT, Controlling Nutritional Status; DKD, diabetic kidney disease; NHANES, National Health and Nutrition Examination Survey.

Table 1

Characteristics of the included patients with DKD

Association between the CONUT score and mortality

The univariate analysis showed that patients with an increased CONUT score had a significantly higher risk of all-cause death (HR 1.53, 95% CI 1.37 to 1.71, p<0.001) and other cause-related death (HR 1.56, 95% CI 1.35 to 1.80, p<0.001). Compared with a CONUT score of 0–1, a CONUT score of 2–4 was associated with a significantly greater risk of all-cause death (HR 1.68, 95% CI 1.31 to 2.16, p<0.001) and other cause-related death (HR 1.60, 95% CI 1.15 to 2.22, p=0.005) and a CONUT score of 5–12 was associated with a significantly elevated risk of all-cause death (HR 4.48, 95% CI 2.41 to 8.36, p<0.001), diabetes-related death (HR 1.90, 95% CI 1.10 to 3.28, p=0.022), nephropathy-related death (HR 1.84, 95% CI 1.07 to 3.17, p=0.026) and other cause-related death (HR 5.54, 95% CI 2.45 to 12.53, p<0.001). After controlling for age, race, marital status, smoking, hypertension, CVD, DR, PIR, antidiabetics, diuretics, UACR, uric acid, energy, protein, total fat, sodium and eGFR, a higher CONUT score was associated with a significantly greater risk of all-cause death (HR 1.30, 95% CI 1.15 to 1.46, p<0.001) and death from other causes (HR 1.54, 95% CI 1.31 to 1.82, p<0.001). In contrast to patients with a CONUT score of 0–1, those who scored 5–12 had a significantly increased risk of all-cause death (HR 2.80, 95% CI 1.42 to 5.51, p=0.003), diabetes-related death (HR 1.78, 95% CI 1.02 to 3.11, p=0.041), nephropathy-related death (HR 1.84, 95% CI 1.04 to 3.24, p=0.036) and other cause-related death (HR 6.54, 95% CI 3.18 to 13.45, p<0.001) (table 2).

Table 2

Association between the CONUT score and mortality in patients with DKD

Association between the CONUT score and mortality in subpopulationsSex

For males, compared with a CONUT score of 0–1, a score of 2–4 was associated with a significantly higher risk of all-cause death (HR 1.34, 95% CI 1.02 to 1.77, p=0.033), and a score of 5–12 was associated with significantly elevated risks of all-cause death (HR 4.40, 95% CI 2.52 to 7.69, p<0.001), diabetes-related death (HR 2.53, 95% CI 1.42 to 4.49, p=0.002), nephropathy-related death (HR 2.50, 95% CI 1.42 to 4.39, p=0.001) and other cause-related death (HR 5.65, 95% CI 2.65 to 12.03, p<0.001). For females, no significant differences were found in the risk of death between women with the CONUT scores of 0–1 and 2–4, and between women with the scores of 0–1 and 5–12 (all p>0.05) (online supplemental table 2, figure 2).

Figure 2

Association between the CONUT score and mortality in subpopulations. For the male subgroup, age, race, marital status, smoking, hypertension, CVD, physical activity, fasting glucose, diuretics, uric acid, energy, protein, sodium and eGFR were adjusted for; for the female subgroup, age, race, BMI, education level, marital status, smoking, CVD, PIR, physical activity, diuretics, uric acid, protein, carbohydrate and eGFR were adjusted for; for the subgroup with CVD, age, race, BMI, marital status, smoking, CVD, diuretics, UACR, uric acid and eGFR were adjusted for; for the subgroup without CVD, age, race, CVD, physical activity, diuretics, uric acid and eGFR were adjusted for; for the subgroup with DR, age, gender, race, BMI, marital status, smoking, hyperlipidaemia, physical activity, antidiabetics, diuretics, UACR, uric acid, carbohydrate, sodium and eGFR were adjusted for; for the subgroup without DR, age, race, marital status, diuretics, uric acid and eGFR were adjusted for. BMI, body mass index; CONUT, Controlling Nutritional Status; CVD, cardiovascular disease; DR, diabetic retinopathy; eGFR, estimated glomerular filtration rate; PIR, poverty income ratio; Ref, reference; UACR, urinary albumin to creatinine ratio.

Cardiovascular disease

Patients with CVD who had a CONUT score of 5–12 exhibited significantly greater risks of all-cause death (HR 2.60, 95% CI 1.41 to 4.79, p=0.002), CVD-related death (HR 3.09, 95% CI 1.27 to 7.52, p=0.013), diabetes-related death (HR 3.85, 95% CI 2.07 to 7.18, p<0.001), nephropathy-related death (HR 4.07, 95% CI 2.18 to 7.57, p<0.001) and other cause-related death (HR 3.76, 95% CI 1.52 to 9.30, p=0.004) than those who had a score of 0–1. Among patients without CVD, a CONUT score of 5–12 was associated with significantly increased risks of all-cause death (HR 5.29, 95% CI 2.59 to 10.81, p<0.001) and other cause-related death (HR 6.03, 95% CI 2.43 to 14.97, p<0.001) than that of 0–1 (online supplemental table 2, figure 2).

Diabetic retinopathy

In patients with DR, a CONUT score of 5–12 was associated with significantly elevated risks of all-cause death (HR 3.74, 95% CI 1.97 to 7.08, p<0.001), CVD-related death (HR 2.55, 95% CI 1.22 to 5.34, p=0.013), diabetes-related death (HR 2.58, 95% CI 1.37 to 4.86, p=0.003), nephropathy-related death (HR 2.61, 95% CI 1.39 to 4.92, p=0.003) and other cause-related death (HR 4.76, 95% CI 2.02 to 11.21, p<0.001), as compared with a score of 0–1. For patients without DR, the risks of all-cause death (HR 3.43, 95% CI 1.80 to 6.54, p<0.001), nephropathy-related death (HR 2.47, 95% CI 1.23 to 4.95, p=0.011) and other cause-related death (HR 3.43, 95% CI 1.34 to 8.73, p=0.010) were significantly higher in those with a CONUT score of 5–12 than those with a score of 0–1 (online supplemental table 2, figure 2).

Discussion

The current study investigated the association between the CONUT score and all-cause and cause-specific mortality among patients with DKD, and further assessed the association in different sex, CVD and DR subgroups. It was illustrated that for patients with DKD, moderate and severe malnutrition was associated with higher risks of all-cause death, diabetes-related death and nephropathy-related death than normal nutritional status. The association between the CONUT score and mortality varied across different sex, CVD and DR subgroups. These findings may act as evidence for risk stratification management and prognosis improvement in patients with DKD.

Malnutrition is a primary issue for patients with chronic diseases. In patients with DKD, malnutrition can exacerbate inflammatory activity and further impair nutrition intake, lowering the quality of life and elevating mortality.17 Medical nutritional therapy has been proposed for DKD, which is beneficial for health and survival.18 The CONUT score assesses nutritional status with three objective indicators: serum albumin (protein metabolism), TC (lipid metabolism) and total lymphocyte count (immune function).7 Serum albumin plays an essential role in nutrition maintenance, metabolic transport and plasma colloid osmotic pressure.19 Hypoalbuminaemia was identified as a prognostic factor for death in elderly individuals.20 Sun et al showed that a reduced level of serum albumin was associated with a greater risk of all-cause mortality in CKD, with the optimum threshold of 4 g/dL.21 High cholesterol levels, a low-risk factor for undernutrition in the CONUT, were related to all-cause mortality in the general population.22 As a marker of immunological status, a decline in the total lymphocyte count can reflect susceptibility to infectious diseases, and malnutrition may lead to decreased lymphocyte maturation and circulating lymphocyte counts.11 In a study by Tojek et al,23 an association was found between the total lymphocyte count of less than 800 mg/L and the highest risk of in-hospital mortality.

With this CONUT score, this study found that compared with normal nutritional status, moderate and severe malnutrition (a score of 2–12) was associated with an increased risk of all-cause mortality. As a dimension of nutritional status assessment, inflammation facilitates the development of DKD,24 possibly through releasing interleukin-1 (IL-1) from monocytes, which might initiate major complications and elevated mortality.25 Additionally, inflammation is correlated with malnutrition and protein-energy wasting, potentially contributing to mortality in DKD.26 As regards immune status, infectious complications can be caused by an immunosuppressive state, which may be associated with morbidity and mortality of patients with DKD.27 In terms of nutritional status, a prior review has indicated that improvement in nutrition plays an important role in mortality among people with CKD.28 Moderate and severe malnutrition was also identified to be associated with diabetes-related mortality and nephropathy-related mortality. Inflammatory response may facilitate the occurrence of T2DM via inducing insulin resistance, and it can be aggravated in the case of hyperglycaemia, promoting long-term complications of diabetes,29 which may contribute to the risk of death. This is a potential explanation for diabetes-related mortality in DKD. Increases in inflammation may also account for mortality due to nephropathy.30 Additionally, we did not find an association between the nutritional status measured by the CONUT score and CVD-related death. This may be attributed to the relatively small number of patients involved herein. Further, we found that undernourished males had higher risks of all-cause, diabetes-related and nephropathy-related death, while in females, no significant associations were observed between nutritional status and mortality. For patients with CVD or DR, moderate and severe malnutrition was associated with a greater risk of all-cause, CVD-related, diabetes-related and nephropathy-related death. Among patients without CVD, moderate and severe malnutrition was related to an elevated risk of all-cause death. For patients without DR, those with moderate and severe malnutrition had a higher risk of all-cause and nephropathy-related death. The significantly discrepant association of the CONUT score and mortality between males and females may be attributed to the limited ability of the CONUT score to distinguish between nutritional status in different genders since the score has no threshold differences in TC, serum albumin and lymphocyte count between men and women.7 Besides, biological differences between men and women with DKD may also be contributors, such as sex hormones, kidney haemodynamic function, adiponectin and oxidative stress.31 For example, higher levels of adiponectin in women might exert a compensatory action against further progression of DKD,32 33 and greater degrees of oxidative stress in men than women may be associated with a worse prognosis in DKD.34 Clinicians may provide personalised advice for different subpopulations at a high risk of death. Large-scale studies are warranted to corroborate our findings.

In the current study, a nationally representative sample was employed to probe into the association between the CONUT score and the risk of mortality in patients with DKD for the first time. Close monitoring of immuno-nutritional status and appropriate nutritional care (eg, dietary regulation) for patients with DKD may help improve prognosis. Of note, the indicators required for CONUT score calculation are common and easily obtainable in clinical practice, with high practicality. Several limitations should be mentioned when interpreting our results. First, the diagnosis of DKD was based on a single measurement of eGFR and UACR, rather than continuous observation for 3 months. Second, some possible confounding factors, such as treatment during follow-up, have not been adjusted for in this analysis. Third, this research was conducted using the data from the American population, which may affect the applicability of the findings to other populations.

Conclusion

Compared with normal nutritional status, moderate and severe malnutrition was associated with higher risks of all-cause mortality, diabetes-related mortality and nephropathy-related mortality in DKD. Close attention should be paid to the immuno-nutritional status of patients with DKD to promote prognosis management and improvement. These findings need to be confirmed in the future studies.

Data availability statement

Data are available in a public, open access repository. Data are available in a public, open access repository. Open access data are available on the NHANES website (www.cdc.gov/nchs/nhanes/).

Ethics statementsPatient consent for publication

Consent obtained directly from patient(s).

Ethics approval

This study involves human participants but the requirement of ethical approval for this was waived by the Institutional Review Board of Shanxi Bethune Hospital, Shanxi Academy of Medical Sciences, Tongji Shanxi Hospital, Third Hospital of Shanxi Medical University, because the data were accessed from NHANES (a publicly available database). The need for written informed consent was waived by the Institutional Review Board of Shanxi Bethune Hospital, Shanxi Academy of Medical Sciences, Tongji Shanxi Hospital, Third Hospital of Shanxi Medical University due to retrospective nature of the study.