Predictive Effect of Salt Intake on Patient and Kidney Survival in Non-Dialysis CKD

Competing Risk Analysis in Older Versus Younger Patients Under Nephrology Care

Carlo Garofalo; Michele Provenzano; Michele Andreucci; Antonio Pisani; Luca De Nicola; Giuseppe Conte; Silvio Borrelli

Nephrol Dial Transplant. 2021;36(12):2232-2240.

Abstract and Introduction

Abstract

Background: The optimal level of salt intake remains ill-defined in non-dialysis chronic kidney disease (CKD) patients under regular nephrology care. This unanswered question becomes critical in older patients who are exposed to higher risk of worsening of cardiorenal disease due to volemic changes.

Methods: In this pooled analysis of four prospective studies in CKD, we compared the risk of all-cause mortality and end-stage kidney disease (ESKD) between patients ≤65 and >65 years of age stratified by salt intake level (<6, 6–8 and >8 g/day) estimated from two measurements of 24-h urinary sodium.

Results: The cohort included 1785 patients. The estimated glomerular filtration rate was 37 ± 21 mL/min/1.73 m² overall, 41 ± 25 in younger patients and 34 ± 16 in older patients (P < 0.001). The median 24-h urinary sodium excretion was 143 mEq [interquartile range (IQR) 109–182] in all, 147 (112–185) in younger patients and 140 (106–179) in older patients (P = 0.012). Salt intake was ≤6, 6–8 and >8 g sodium chloride/day in 21.9, 26.2 and 52.0% of older patients and 18.6, 25.2 and 56.2% in younger patients, respectively (P = 0.145). During a median follow-up of 4.07 years we registered 383 ESKD and 260 all-cause deaths. In the whole cohort, the risks of ESKD and all-cause death did not differ by salt intake level. In older patients, ESKD risk [multi-adjusted hazard ratio (HR) and 95% confidence interval (CI)] was significantly lower at salt intakes of 6–8 g/day [HR 0.577 (95% CI 0.361–0.924)] and >8 g/day [HR 0.564 (95% CI 0.382–0.833)] versus the reference group (<6 g/day). Mortality risk was higher in older versus younger patients, with no difference across salt intake categories. No effect of salt intake on ESKD and mortality was observed in younger patients.

Conclusions: CKD patients under nephrology care show a moderate salt intake (8.4 g/day) that is lower in older versus younger patients. In this context, older patients are not exposed to higher mortality across different levels of salt intake, while salt intake <6 g/day poses a greater risk of ESKD.

Introduction

Ageing of the general population is the major determinant of the observed increase of non-communicable chronic disease with higher rates of disabilities and frailty in elderly subjects. In this context, non-dialysis chronic kidney disease (CKD) plays a key role because its prevalence is more than double in individuals >65 years of age.^[1,2] Since the epidemiological burden of CKD^[3,4] leads to dramatic consequences in terms of social and economic costs, public health interventions aimed at slowing CKD progression and preventing the need for renal replacement therapy become crucial in elderly CKD patients.

In CKD, the mainstay of multifactorial therapy is the prescription of a low salt diet. The World Health Organization recommends a salt intake of <5 g/day in the general population to reduce blood pressure (BP) and cardiovascular (CV) mortality.^[5] In a salt-sensitivity condition, such as CKD, a low salt intake is recommended to decrease BP and albuminuria; the latter effect being related to the synergism with renin–angiotensin–aldosterone system (RAAS) inhibitors.^[6–9] Indeed, BP and albuminuria are recognized as major determinants of CKD progression.^[8–11] However, few data are available on the effect of reducing sodium intake on renal outcome. A recent meta-analysis of randomized controlled trials has shown that a low salt diet is efficacious in reducing BP and albuminuria in CKD patients;^[12] however, all included trials were characterized by short-term follow-ups. Long-term effects of low salt intake have been investigated in the large cohort of the Chronic Renal Insufficiency Cohort (CRIC) study and results showed a significant association between the highest quartile of sodium intake and the risk of CKD progression and CV mortality; the outcomes of the lowest quartile did not differ from those of the second and third quartiles.^[13,14] However, the CRIC study was performed in middle-aged patients and did not provide evidence in elderly patients.

The gap of knowledge in elderly patients is not trivial because this large subgroup of the CKD population is characterized by impaired renal autoregulation due to atherosclerosis that can be enhanced by the effects of low salt intake on systemic and renal haemodynamics.^[15] Indeed, elderly subjects are more susceptible to acute kidney dysfunction due to renal hypoperfusion even when BP levels are within the normal range.^[16,17] In this regard, relevant are the signals of the potential risk of all-cause and CV mortality associated with lower salt intake in the general as well as the CKD population.^[18–20] In contrast, older patients may be exposed to higher CV risk in the presence of higher salt intake because of the well-known sodium sensitivity of BP and the risk of heart failure.^[6,21,22]

Therefore, to gain insights into the association between salt intake level and risks of end-stage kidney disease (ESKD) and mortality by age, we performed an observational study in a cohort of patients with CKD Stages 1–5 under regular nephrology care. The risk of either hard endpoint was compared between patients ≤65 or >65 years of age stratified by salt intake category (<6, 6–8 and >8 g/day).

1 2 3 4

Next Section

Table 1. Characteristics of patients stratified according to age groups and salt intake categories
Salt intake, (g/day)
Characteristics	<6			6–8			>8
Characteristics	Age ≤65 years	Age >65 years	P-value	Age ≤65 years	Age >65 years	P-value	Age ≤65 years	Age > 65 years	P-value
Patients, n	143	222		194	266		432	528
Age (years)	51.7 ± 10.8	75.9 ± 6.1		50.7 ± 12.3	75.3 ± 6.1		49.7 ± 11.9	74.2 ± 5.8
Male gender, %	46.9	52.7	0.275	46.9	52.3	0.257	63.0	67.6	0.131
BMI (kg/m²)	26 ± 6	27 ± 5	0.013	27 ± 5	28 ± 5	0.027	28 ± 5	28 ± 5	0.149
Smoking, %	25.0	9.2	<0.001	16.9	9.2	0.028	21.5	9.9	<0.001
Diabetes, %	17.5	31.1	0.004	20.1	30.1	0.016	16.4	37.5	<0.001
Previous CV events, %	23.8	48.6	<0.001	22.2	37.2	0.001	23.1	45.5	<0.001
Chronic heart failure, %	1.9	9.8	0.010	2.1	5.7	0.099	3.1	6.3	0.053
LVH, %	34.3	58.3	<0.001	35.8	62.4	<0.001	42.7	67.2	<0.001
CKD stage, %			0.046			0.004			<0.001
1–2	11.9	4.5		14.9	6.4		21.8	6.8
3A	17.5	14.9		14.9	20.7		18.8	18.2
3B	25.2	26.1		32.0	32.7		25.9	35.6
4	28.7	39.2		23.7	31.2		24.3	29.5
5	16.8	15.3		14.4	9.0		9.3	9.8
Primary renal disease, %			<0.001			<0.001			<0.001
Hypertension	16.8	43.2		16.5	39.5		16.0	39.2
Diabetes	8.4	11.7		11.3	15.8		7.9	14.4
GN	32.9	9.9		28.9	9.4		29.4	9.7
PKD	7.7	1.4		10.3	2.3		9.3	3.6
Tubulointerstitial	7.0	8.6		12.4	7.9		15.0	8.0
Other/unknown	27.3	25.2		20.6	25.2		22.5	25.2
Systolic BP (mmHg)	132 ± 17	137 ± 20	0.011	127 ± 19	137 ± 18	<0.001	131 ± 19	139 ± 18	<0.001
Diastolic BP (mmHg)	78 ± 10	75 ± 11	0.051	79 ± 11	76 ± 10	0.001	80 ± 11	77 ± 10	<0.001
Phosphorous (mg/dL)	3.9 ± 0.8	3.8 ± 0.8	0.495	4.0 ± 1.1	3.7 ± 0.7	0.001	3.8 ± 0.8	3.6 ± 0.8	0.018
Serum albumin (g/dL)	4.0 ± 0.5	4.0 ± 0.4	0.798	4.2 ± 0.5	4.1 ± 0.5	0.053	4.2 ± 0.4	4.1 ± 0.4	<0.001
Total cholesterol (mg/dL)	204 ± 57	187 ± 43	0.002	200 ± 49	193 ± 47	0.114	199 ± 42	191 ± 44	0.007
Haemoglobin (g/dL)	12.6 ± 1.8	12.4 ± 1.6	0.261	12.8 ± 1.8	12.6 ± 1.6	0.117	13.3 ± 1.7	12.7 ± 1.7	<0.001
eGFR (mL/min/1.73 m²)	36 ± 23	31 ± 15	0.011	38 ± 22	35 ± 16	0.110	44 ± 27	35 ± 16	<0.001
Proteinuria (g/day), median (IQR)	0.35 (0.15–1.00)	0.23 (0.07–0.72)	<0.001	0.33 (0.11–0.99)	0.20 (0.09–0.73)	<0.001	0.46 (0.151.22)–	0.24 (0.10–0.90)	<0.001
Urinary Na (mEq/day)	80 ± 16	79 ± 17	0.733	120 ± 10	118 ± 10	0.793	190 ± 45	185 ± 38	0.220
Urinary K (mEq/day)	38 (28–51)	40 (30–57)	0.843	49 (36–65)	48 (34–59)	0.066	53 (42–67)	54 (42–70)	0.722

Data are presented as mean ± SD unless stated otherwise. LVH: left ventricular hyperthrophy; GN: glomerulonephritis; PKD: polycystic kidney disease; K: potassium.

Table 2. Main treatment features in CKD patients stratified according to age ≤65 or >65 years and salt intake category
Salt intake, (g/day)
Treatments	<6			6–8			>8
Treatments	Age ≤65 years	Age >65 years	P-value	Age ≤65 years	Age >65 years	P-value	Age ≤65 years	Age > 65 years	P-value
Antihypertensive drugs, median n (IQR)	2 (2–3)	2 (2–3)	0.183	2 (1–3)	3 (2–3)	<0.001	2 (1–3)	3 (2–4)	<0.001
Anti-RAS, %	67.8	70.7	0.558	69.9	73.7	0.094	66.5	71.4	0.613
Diuretics, %	18.9	36.5	<0.001	18.8	37.6	<0.001	21.6	36.7	<0.001
Beta-blockers, %	29.6	27.3	0.673	33.1	32.5	0.311	27.5	32.4	0.834
CCB, %	38.0	42.6	0.435	35.9	43.4	0.066	33.8	50.3	<0.001
Statins, %	35.9	43.2	0.165	34.0	36.8	0.830	37.8	44.4	0.001
Erythropoietin, %	15.5	27.0	0.010	10.5	19.9	0.023	11.9	16.4	0.008
Vitamin D, %	22.5	28.4	0.216	17.8	18.4	0.511	16.1	22.1	0.103

RAS: renin–angiotensin system; CCB: calcium channel blockers.

Table 3. Adjusted cause-specific HRs for ESKD and all-cause death
Salt intake, (g/day)
Variables	<6		6–8		>8
Variables	Events/patients (n/N)	HR (95% CI)	Events/patients (n/N)	HR (95% CI)	Events/patients (n/N)	HR (95% CI)
ESKD
Overall	90/365	Reference	93/460	0.897 (0.658–1.222)	200/960	0.829 (0.634–1.084)
Age ≤65 years	48/143	Reference	62/194	1.189 (0.788–1.795)	120/432	1.096 (0.763–1.574)
Age >65 years	42/222	1.024 (0.656–1.599)	31/266	0.577 (0.361–0.924)^a	80/528	0.564 (0.382–0.833)^a
Death
Overall	59/365	Reference	74/460	0.982 (0.692–1.394)	127/960	0.857 (0.619–1.185)
Age ≤65 years	9/143	Reference	10/194	0.791 (0.318–1.966)	18/432	0.620 (0.275–1.397)
Age >65 years	50/222	3.810 (1.829–7.936)	64/266	3.695 (1.808–7.550)^a	109/528	2.938 (1.457–5.926)^a

Adjusted for gender, eGFR, urinary protein excretion, primary renal disease, history of CVD, BMI, systolic BP, haemoglobin, phosphorous, albumin, use of RAAS inhibitors and use of diuretics. HR, hazard ratio.
^aRefers to a significant (P < 0.05) difference between the same salt intake category.

Table 4. Main features of patients with CKD from previous observational cohort studies (light grey) and our study (dark grey) testing the effect on CKD progression of salt intake measured on two 24-h urine collections
Main features	He et al. [13]	Fan et al. [33]	Lambers Heerspink et al. [34]	Vegter et al. [35]	Garofalo et al. (current)
Patients, n	3939	840	1177	500	1785
Male, %	56	60.5	65.3	76.0	58.4
Age (years), mean	57.7	51.7	59.0	51.9	64.3
Diabetes mellitus, %	47.8	5.1	100	0	27.0
CVD, %	33.0	13.1	NR	NR	35.0
eGFR (mL/min), mean	44.5	32.5	44.0	43.3	37.0
Proteinuria (g/day)	0.2	0.32	NR	2.9	0.4
Urinary sodium (mEq/day)/salt intake (g/day)	161/9.4	151/8.9	181/10.6	177/10.4	146/8.6
Renal outcome	ESKD or halving eGFR	ESKD	ESKD or doubling of serum creatinine	ESKD or doubling of serum creatinine	ESKD
Main results	HR for the highest quartile of urinary sodium excretion (≥194.6 mmol/24 h) was 1.54 (95% CI 1.23–1.92)	No association was found between urinary sodium excretion and ESKD	HR for the highest tertile of urinary sodium creatinine ratio (≥153 mmol/g) was 1.27 (95% CI 0.86–1.88).	High salt group (>200 mEq/g of sodium:creatinine ratio) had a 3.3-fold (95% CI 1.7–6.4) excess risk compared with low salt group	See text