Effects of Hepatorenal Function on Urinary Alzheimer-Associated Neuronal Thread Protein: A Laboratory-Based Cross-Sectional Study Among the Older Chinese Population

Abstract

Background:

Urinary Alzheimer-associated neuronal thread protein (AD7c-NTP) is a biomarker for the early diagnosis of Alzheimer’s disease (AD). It remains unclear whether hepatorenal function affects the urinary AD7c-NTP level.

Objective:

To evaluate the effects of hepatorenal function on urinary AD7c-NTP level.

Methods:

We enrolled 453 participants aged 60–100 years. An automated chemistry analyzer was used to determine the indicators of serum hepatorenal function. Enzyme-linked immunosorbent assay was used to measure the urinary AD7c-NTP level.

Results:

Spearman’s correlation analysis showed a negative correlation between urinary AD7c-NTP levels and indicators of hepatorenal function, including albumin (r = –0.181, p < 0.001), albumin/globulin ratio (r = –0.224, p < 0.001), cholinesterase (r = –0.094, p = 0.046), total carbon dioxide (r = –0.102, p = 0.030), and glomerular filtration rate (r = –0.260, p < 0.001), as well as a positive correlation with globulin (r = 0.141, p = 0.003), aspartate transaminase (r = 0.186, p < 0.001), blood urine nitrogen (r = 0.210, p < 0.001), creatinine (r = 0.202, p < 0.001), uric acid (r = 0.229, p < 0.001), and cystatin C (r = 0.265, p < 0.001). The least absolute shrinkage and selection operator (LASSO) regression analysis and multiple linear regression model analyses showed that the statistically significant hepatorenal indicators for predicting AD7c-NTP were A/G (p = 0.007), AST (p = 0.002), BUN (p = 0.019), and UA (p = 0.003).

Conclusions:

The effects of hepatorenal indicators should be considered when using urinary AD7c-NTP levels in clinical settings.

Keywords

Alzheimer’s disease Alzheimer-associated neuronal thread protein kidney liver hepatorenal indicator urine

INTRODUCTION

Alzheimer’s disease (AD) is a neurodegenerative disorder characterized by an insidious onset and progressive decline in behavioral and cognitive functions, including memory, comprehension, language, attention, reasoning, and judgment, ultimately leading to shortened life expectancy [1]. Despite the absence of a cure for AD, once common clinical symptoms manifest, some treatments exist that may offer limited improvement [2]. However, a significant portion of early-onset AD cases remain undiagnosed or unrecognized [3, 4]. In 2018, the National Institute on Aging and Alzheimer’s Association jointly defined AD by its underlying pathological processes, which can be evidenced through postmortem examination or in vivo via biomarkers [5]. Biomarkers play a crucial role in the accurate and early detection of AD and are essential for disease management [6]. Currently recognized core AD biomarkers associated with the pathology include amyloid-β (Aβ), pathologic tau, and markers indicating neurodegeneration, such as total tau (t-tau) and neurofilament light chain (NFL) [7 –9]. As understanding of the pathophysiological, molecular, and structural changes in AD has advanced, biomarkers associated with synaptic damage, neuroinflammation, neuroimmunity, microglia and astrocyte activation, systemic immunity, systemic inflammation, nutrition and metabolism, apoptosis, mitochondrial dysfunction, and oxidative stress have also emerged as diagnostic tools [6 , 10–12]. Alzheimer-associated neuronal thread protein (AD7c-NTP) was identified from a novel 1,442-nucleotide, Alu sequence-containing cDNA, first isolated by de la Monte et al. from an AD brain expression library in 1996 [13]. AD7c-NTP is an approximately 41 kD membrane-spanning phosphoprotein. AD7c-NTP has emerged as a candidate biomarker for AD due to its elevated levels specifically observed in the putative brain domains, cerebrospinal fluid (CSF), and urine of AD patients [13 –15]. Mechanism studies have revealed that AD7c-NTP co-localizes with phosphorylated tau (p-tau) and phosphorylated neurofilament in neuronal perikarya, abnormal neurites, and swollen axons of frontal and temporal tissues from AD brains. Moreover, increased AD7c-NTP expression is most pronounced in neurons that are either cytologically intact or slightly degenerated, rather than in apoptosis cells, providing evidence that AD7c-NTP may offer early diagnostic clues before irreversible neurodegeneration occurs [16]. Given that urinary testing is noninvasive, easy to perform, and more readily accepted by patients than other methods, urinary AD7c-NTP concentration has been recommended as a practical tool for early adjunctive diagnosis of AD and mild cognitive impairment in multicenter studies [17 –20].

However, since urine is anatomically distant from the brain and subject to numerous physiological and pathological influences, it is essential to identify factors affecting the detection of urine AD7c-NTP during both pre-analytical and analytical processes. Regarding pre-analytical factors, our research has demonstrated that AD7c-NTP can be reliably detected in random urine samples, rather than exclusively in first-morning urine. Additionally, when immediate testing is not feasible, boric acid can serve as a preservative if samples are stored at 4°C, provided testing occurs within 5 days [21]. Furthermore, our investigations have shown that urinary AD7c-NTP levels are unaffected by factors such as education, occupation, body mass index, place of residence, family history of dementia, hypertension, stroke, anemia, diabetes, dyslipidemia, history of kidney or liver disease, cancer, chronic lung disease, thyroid dysfunction, or depressive symptoms, as demonstrated in a population-based epidemiological study [22]. However, previous studies have not explored potential factors influencing the excretion of AD7c-NTP from the CSF to blood to urine. Notably, numerous substances associated with liver metabolic function and kidney excretion function may act as interference factors in the detection of urine AD7c-NTP (as illustrated in Fig. 1). Therefore, this study aims to investigate whether urinary AD7c-NTP levels are influenced by hepatorenal indicators among the older Chinese population. By doing so, we aim to provide a more comprehensive understanding of the clinical utility of AD7c-NTP as an adjunct diagnostic or screening tool.

Fig. 1

Schematic diagram of liver and kidney metabolism of urinary Alzheimer-associated neuronal thread protein (AD7c-NTP) in vivo. BBB, blood-brain barrier; GBM, glomerular basement membrane; RT, renal tubule.

MATERIALS AND METHODS

Participants

A total of 453 geriatrics subjects from the Affiliated Hospital of Guilin Medical University were enrolled as study participants between July 2022 and April 2023. The inclusion criteria were as follows: 1) age ≥60 years; 2) completion of comprehensive medical examinations; 3) presence of both normal and varying degrees of abnormal liver and kidney function indicators. The exclusion criteria were as follows: 1) cognitive dysfunction, assessed using the Mini-Mental State Examination (MMSE), with scores≤24 considered abnormal for subjects with over 6 years of education, ≤20 for subjects with 1–6 years of education, and ≤17 for illiterate subjects; 2) detection, through physical examination, of positive indicators of pathological manifestations in the neurological system alongside the presence of focal brain lesions on head and skull imaging; 3) pre-existing conditions such as dyslipidemia, hypertension, and diabetes; 4) recent acute infections, cancer, Parkinson’s disease, depression, anxiety, or hypothyroidism; 5) history of brain trauma, alcohol or substance abuse. Participation in the study was voluntary, and subjects or their guardians provided informed consent. This study received approved from the Ethics Committee of Affiliated Hospital of Guilin Medical University (approval no.: 2022QTLL-47).

Blood biochemical laboratory measurements

All biochemical measurements followed a standardized protocol. Blood samples were collected following a 10–12-h fasting period. Approximately 4 ml of venous blood was drawn between 6 : 00 and 9 : 00 a.m. using pro-coagulation tubes. Analytical tests were completed within 4 h post-collection. The Cobas 8000 modular analyzer series (Roche Group, Basel, Switzerland) was utilized in accordance with the manufacturer’s guidelines for clinical laboratory investigations. Liver function indicators included serum total bilirubin (T-BIL), direct bilirubin (D-BIL), indirect bilirubin (I-BIL), total protein (TP), albumin (ALB), globulin (Glo), albumin/globulin ratio (A/G), prealbumin (PA), alkaline phosphatase (ALP), cholinesterase (CHE), glutamic-pyruvic transaminase (GPT/ALT), and glutamic oxalacetic transaminase (GOT/AST). Kidney function indicators included blood urea nitrogen (BUN), creatinine (Cr), uric acid (UA), total carbon dioxide (TCO₂), glomerular filtration rate (GFR) and cystatin C (cysC). Table 1 presents the normal reference intervals, skewness, and kurtosis for liver and kidney function indicators.

Table 1

Distribution and cut-off values of liver and kidney function indexes of study participants

Variables	Median (interquartile range)	Mean±SD	Normal cut-off value	Skewness	Kurtosis
T-BIL (μmol/L)	9.75 (6.20–19.57)	19.98±40.41	0.001–21.00	7.435±0.115	70.21±0.229
D-BIL (μmol/L)	4.70 (3.00–9.35)	12.55±35.13	0.001–6.00	7.538±0.115	69.230±0.229
I-BIL (μmol/L)	4.70 (3.00–9.09)	7.43±7.89	1.71–15.00	3.875±0.115	22.831±0.229
TP (g/L)	67.80 (63.03–73.25)	68.21±8.14	60.00–83.00	0.361±0.115	1.679±0.229
ALB (g/L)	37.70 (33.45–40.70)	36.84±5.53	35.00–55.00	–0.358±0.115	0.141±0.229
GLo (g/L)	30.60 (27.00–34.26)	31.37±7.09	15.00–33.00	1.594±0.115	6.116±0.229
A/G	1.24 (1.03–1.43)	1.23±0.32	1.30–2.70	0.157±0.115	0.407±0.229
PA (mg/L)	185.91 (138.83–225.81)	184.52±65.25	180.00–390.00	0.006±0.115	–0.021±0.229
ALP (U/L)	79.00 (66.48–104.50)	108.34±108.19	25.00–90.00	4.636±0.115	24.907±0.229
CHE (U/L)	6475.00 (4740.00–7769.71)	6410.78±2266.26	4000.00–13200.00	0.433±0.115	0.897±0.229
ALT (U/L)	15.60 (10.35–23.93)	30.51±71.99	0.001–38.00	10.186±0.115	132.700±0.229
AST (U/L)	19.80 (16.00–28.55)	38.50±126.16	0.001–40.00	17.240±0.115	334.931±0.229
BUN (mmol/L)	5.80 (4.50–8.20)	7.83±10.16	3.20–7.50	11.92±0.115	193.150±0.229
Cr (μmol/L)	77.00 (61.50–105.00)	109.78±125.80	44.00–115.00	4.939±0.115	28.301±0.229
UA (μmol/L)	342.02 (259.50–423.50)	353.04±137.97	110.00–420.00	0.845±0.115	1.353±0.229
TCO₂ (mmol/L)	25.20 (22.70–27.82)	25.23±4.19	23.00–31.00	–0.052±0.115	0.311±0.229
GFR (mL/min)	50.53 (35.19–73.84)	55.52±27.68	47.00–150.00	0.635±0.115	0.120±0.229
cysC (mg/L)	1.42 (1.03–1.92)	1.69±1.04	0.57–1.50	2.706±0.115	10.248±0.229

T-BIL, total bilirubin; D-BIL, direct bilirubin; I-BIL, indirect bilirubin; TP, total protein; ALB, albumin; Glo, globulin; A/G, albumin/globulin; PA, prealbumin; ALP, alkaline phosphatase; CHE, cholinesterase; ALT, glutamic-pyruvic transaminase/GPT; AST, glutamic oxalacetic transaminase/GOT; BUN, blood urea nitrogen; Cr, creatinine; UA, uric acid; TCO₂, total carbon dioxide; GFR, glomerular filtration rate; cysC, cystatin C.

Urinary AD7c-NTP laboratory measurements

Clean midstream urine samples were collected from all participants in the morning. Urine specimens were transferred to Eppendorf tubes containing boric acid (2 g/L) as a preservative and stored at 4°C [21]. Urinary AD7c-NTP levels were quantified using an ELISA kit (Anqun Biological Technology, Shenzhen, China). Following the manufacturer’s instructions, 100μL of the urine sample was added to the corresponding well and incubated at 37°C for 1 h. The wells were then washed five times with phosphate-buffered saline (PBS), ensuring the avoidance of bubble formation during the washing process. Subsequently, a biotinylated rabbit anti-AD7c-NTP antibody was added and incubated at 37°C for 30 min. After another five washes with PBS, the samples were incubated with horseradish peroxidase-labeled avidin at 37°C for an additional 30 min. Following another five washes with PBS, chromogenic reagents A and B were added sequentially and incubated at 37°C for 15 min. The reaction was terminated by adding 50μL of sulfuric acid. The optical density of each sample was measured at a wavelength of 450 nm using a Multiskan Spectrum Microplate Spectrophotometer (Thermo Fisher Scientific, Waltham, USA). Urinary AD7c-NTP levels were calculated based on the standard curve of recombinant human AD7c-NTP peptides.

Statistical analysis

The Statistical Package for Social Sciences (SPSS) version 27.0, R version 4.2, and SAS version 9.4 were employed for data analysis. Continuous variables were presented as mean±standard deviation (SD) if they followed a normal distribution or as median (interquartile range) if they did not. The two-independent-sample t-test was utilized to compare data between two groups, such as sex. Analysis of variance (ANOVA) was employed to compare data among three groups, such as age. The Mann-Whitney U test was applied to compare data for variables with non-normal distributions, such as urinary AD7c-NTP levels. After adjusting for age and sex, general linear models were used to analyze differences between urinary AD7c-NTP levels and hepatorenal function indicators. Spearman’s correlation analysis was conducted to further explore the correlations between urinary AD7c-NTP levels and hepatorenal function indicators. Least absolute shrinkage and selection operator (LASSO) regression analysis was performed using the glmnet package (version 4.1-8) in R software. Variance inflation was used to assess collinearity among different variables. After conducting LASSO regression analysis to detect serious multicollinearity among several variables, the selected predictive variables were further analyzed. The most significant variables identified through LASSO regression were included in multiple linear regression models to predict AD7c-NTP levels. Statistical significance was defined as p < 0.05.

RESULTS

Urinary AD7c-NTP levels by gender and age

In this study, 453 participants aged 60–100 years were enrolled, with an average age of 69.97±7.63 years, and 50.55% of the participants (n = 229) were male. We analyzed the distribution of urinary AD7c-NTP across different age groups based on previous published articles [22, 23]. As presented in Table 2, urinary AD7c-NTP levels were slightly higher in female subjects [0.66 (0.43, 1.41) ng/mL] than in male subjects [0.62 (0.36, 1.35) ng/mL], although the difference was not statistically significant (p = 0.184). Furthermore, Table 2 shows significant differences in urinary AD7c-NTP levels among the age groups 60–69 [0.57 (0.37, 1.22) ng/mL], 70–79 [0.79 (0.42, 1.47) ng/mL], and 80–89 [0.99 (0.47, 1.72) ng/mL] (p = 0.013), indicating a tendency for urinary AD7c-NTP levels to increase with age.

Table 2

Urinary Alzheimer-associated neuronal thread protein (AD7c-NTP) levels according to gender and age

Variables	Group	Age (years)	Male:female	No. of cases	Percent (%)	AD7c-NTP (ng/mL)	Z	p
	All	69.97±7.63	229 : 224	453	100	0.65 (0.38–1.36)
Gender	Male	69.36±7.05		229	50.55	0.62 (0.36–1.35)	–1.329	0.184
	Female	70.60±8.16		49.45	0.66 (0.43–1.41)
Age	60–69	64.67±2.68	135 : 126	261	57.62	0.57 (0.37–1.22)	8.756	0.013 ^*
	70–79	73.73±2.80	68 : 63	131	28.92	0.79 (0.42–1.47)
	≥80	86.62±4.43	26 : 35	61	13.47	0.99 (0.47–1.72)

Age is expressed as mean±standard deviation, AD7c-NTP is expressed as median (interquartile range); asterisk indicates p value < 0.05; Z refers to the statistical value of comparison between groups using Mann-Whitney U test.

Association between urinary AD7c-NTP levels and liver functions

After adjusting for age and sex, no significant differences were observed in urinary AD7c-NTP levels among subjects with different D-BIL, TP, ALP, CHE, ALT, or AST levels. However, significant differences were noted in urinary AD7c-NTP levels among subjects with different T-BIL (p = 0.046), I-BIL (p = 0.012), ALB (p = 0.011), Glo (p < 0.001), A/G (p < 0.001), and PA (p = 0.043) levels (Table 3). Specifically, urinary AD7c-NTP levels were significantly lower in the T-BIL elevated [0.43 (0.32, 1.02) ng/mL] and I-BIL elevated [0.40 (0.31, 0.94) ng/mL] groups compared to those in the T-BIL normal [0.65 (0.38, 1.50) ng/mL] and I-BIL normal [0.65 (0.39, 1.46) ng/mL] groups, respectively. Conversely, urinary AD7c-NTP levels were significantly higher in the ALB reduced [0.65 (0.43, 1.41) ng/mL], Glo elevated [0.91 (0.52, 2.47) ng/mL], A/G reduced [0.79 (0.44, 1.72) ng/mL], and PA reduced [0.65 (0.59, 1.44) ng/mL] groups compared to those in the ALB normal [0.62 (0.37, 1.32) ng/mL], Glo normal [0.52 (0.35, 1.17) ng/mL], A/G normal [0.49 (0.35, 1.02) ng/mL], and PA normal [0.62 (0.22,1.29) ng/mL] groups, respectively.

Table 3

Urinary Alzheimer-associated neuronal thread protein (AD7c-NTP) levels according to liver biochemical indicators, and correlation of urinary AD7c-NTP levels with liver biochemical indicators

Variables	Group	Age (years)	Male:female	No. of cases	AD7c-NTP (ng/ml)	F	p ^(a)	r	p ^(b)
T-BIL (mu mol/L)	normal	70.23±7.61	162 : 184	346	0.65 (0.38–1.50)	2.683	0.046 ^*	–0.023	0.619
T-BIL (mu mol/L)	elevated	69.13±7.68	67 : 40	107	0.43 (0.32–1.02)	2.683	0.046 ^*	–0.023	0.619
D-BIL (mu mol/L)	normal	69.91±7.39	125 : 153	278	0.68 (0.42–1.61)	2.065	0.104	–0.016	0.732
D-BIL (mu mol/L)	elevated	70.07±8.04	104 : 71	175	0.59 (0.40–1.14)	2.065	0.104	–0.016	0.732
I-BIL (mu mol/L)	normal	70.25±7.73	197 : 207	404	0.65 (0.39–1.46)	3.661	0.012^*	–0.048	0.305
I-BIL (mu mol/L)	elevated	67.67±6.43	32 : 19	49	0.40 (0.31–0.94)	3.661	0.012^*	–0.048	0.305
TP (g/L)	reduced	70.10±7.95	29 : 32	61	0.65 (0.43–1.34)	1.885	0.112	0.002	0.973
TP (g/L)	normal	70.14±7.63	192 : 184	376	0.63 (0.38–1.50)	1.885	0.112	0.002	0.973
	elevated	65.50±5.24	8 : 8	16	0.43 (0.36–0.80)
ALB (g/L)	reduced	71.42±8.75	69 : 75	144	0.65 (0.43–1.41)	3.762	0.011^*	–0.181	<0.001^*
ALB (g/L)	normal	69.30±6.97	160 : 149	309	0.62 (0.37–1.32)	3.762	0.011^*	–0.181	<0.001^*
Glo (g/L)	normal	69.55±7.34	162 : 147	309	0.52 (0.35–1.17)	5.759	<0.001^*	0.141	0.003^*
Glo (g/L)	elevated	70.89±8.19	67 : 77	144	0.91 (0.52–2.47)	5.759	<0.001^*	0.141	0.003^*
A/G	reduced	70.63±8.05	126 : 138	264	0.79 (0.44–1.72)	7.098	<0.001^*	–0.224	<0.001^*
A/G	normal	69.06±6.94	103 : 86	189	0.49 (0.35–1.02)	7.098	<0.001^*	–0.224	<0.001^*
PA (mg/L)	reduced	71.03±8.40	97 : 108	205	0.65 (0.59–1.44)	2.747	0.043^*	–0.066	0.163
PA (mg/L)	normal	69.10±6.83	132 : 116	248	0.62 (0.22–1.29)	2.747	0.043^*	–0.066	0.163
ALP (U/L)^c	normal	69.71±7.63	135 : 143	278	0.65 (0.38–1.30)	1.625	0.167	–0.026	0.583
ALP (U/L)^c	elevated	70.32±7.62	93 : 81	174	0.62 (0.36–1.50)	1.625	0.167	–0.026	0.583
CHE (U/L)^d	reduced	71.78±8.81	36 : 33	69	0.67 (0.42–1.15)	1.669	0.156	–0.094	0.046
CHE (U/L)^d	normal	69.63±7.38	193 : 189	382	0.62 (0.26–1.35)	1.669	0.156	–0.094	0.046
ALT (U/L)	normal	70.01±7.65	193 : 200	393	0.65 (0.39–1.47)	2.043	0.107	0.092	0.050
ALT (U/L)	elevated	69.77±7.56	36 : 24	60	0.43 (0.35–1.10)	2.043	0.107	0.092	0.050
AST (U/L)	normal	69.93±7.55	192 : 192	384	0.65 (0.39–1.47)	1.833	0.140	0.186	<0.001
AST (U/L)	elevated	70.19±8.16	37 : 32	69	0.42 (0.32–1.08)	1.833	0.140	0.186	<0.001

Notes: T-BIL, total bilirubin; D-BIL, direct bilirubin; I-BIL, indirect bilirubin; TP, total protein; ALB, albumin; Glo, globulin; A/G, albumin/globulin; PA, prealbumin; ALP, alkaline phosphatase; CHE, cholinesterase; ALT, glutamic-pyruvic transaminase/GPT; AST, glutamic oxalacetic transaminase/GOT. Abbreviations: Age is expressed as mean±standard deviation, AD7c-NTP is expressed as median (interquartile range); asterisk indicates p value < 0.05; p^(a) represents the value of general linear model analysis; p^(b) represents the value of Spearman’s correlation analysis; ^crepresents one case in the reduced ALP group that was not included in the analysis; ^drepresents two cases in the elevated CHE groups that were not included in the analysis.

Correlation analyses between urinary AD7c-NTP levels and liver indicator values

Further analysis presented in Table 3 indicates that there was no significant correlation between urinary AD7c-NTP levels and T-BIL, D-BIL, I-BIL, TP, PA, ALP, or ALT levels. However, a negative correlation was observed between urinary AD7c-NTP levels and ALB (r = –0.181, p < 0.001), A/G (r = –0.224, p < 0.001), and CHE (r = –0.094, p = 0.046). Conversely, a positive correlation was found between urinary AD7c-NTP levels and Glo (r = 0.141, p = 0.003) and AST (r = 0.186, p < 0.001).

Association between urinary AD7c-NTP levels and kidney functions

After adjusting for age and sex, significant differences were observed in urinary AD7c-NTP levels among subjects with different BUN (p < 0.001), Cr (p < 0.001), UA (p < 0.001), TCO₂ (p = 0.026), GFR (p < 0.001), and cysC (p < 0.001) levels (Table 4). Specifically, urinary AD7c-NTP levels were significantly lower in the BUN reduced [0.52 (0.38, 0.74) ng/mL], Cr reduced [0.51 (0.35, 1.06) ng/mL], UA reduced [0.58 (0.31, 1.12) ng/mL], and TCO₂ elevated [0.52 (0.31, 1.30) ng/mL] groups. Conversely, urinary AD7c-NTP levels were significantly higher in the BUN elevated [1.06 (0.45, 2.15) ng/mL], Cr elevated [1.15 (0.50, 2.56) ng/mL], UA elevated [0.91 (0.50, 1.81) ng/mL], and TCO₂ reduced [0.80 (0.51, 1.72) ng/mL] groups compared to those in the BUN normal [0.60 (0.35, 1.14) ng/mL], Cr normal [0.60 (0.37, 1.19) ng/mL], UA normal [0.60 (0.36, 1.25) ng/mL], and TCO₂ normal [0.61 (0.38, 1.21) ng/mL] groups, respectively. Furthermore, urinary AD7c-NTP levels were significantly higher in the GFR reduced (or cysC elevated) group [0.81 (0.46, 1.91) ng/mL] compared to those in the GFR normal (or cysC normal) group [0.56 (0.34, 1.04) ng/mL].

Table 4

Urinary Alzheimer-associated neuronal thread protein (AD7c-NTP) levels according to kidney biochemical indicators, and correlation of urinary AD7c-NTP levels with kidney biochemical indicators

Variables	Group	Age (years)	Male:female	No. of cases	AD7c-NTP (ng/ml)	F	p^(a)	r	p^(b)
BUN	reduced	67.73±6.17	13 : 20	33	0.52 (0.38–0.74)	8.914	<0.001^*	0.210	<0.001^*
	normal	69.16±7.11	145 : 145	290	0.60 (0.35–1.14)
	elevated	72.37±8.53	71 : 59	130	1.06 (0.45–2.15)
Cr	reduced	70.92±11.45	3 : 9	12	0.51 (0.35–1.06)	8.628	<0.001^*	0.202	<0.001^*
	normal	69.36±7.07	164 : 164	348	0.60 (0.37–1.19)
	elevated	72.16±8.71	62 : 31	93	1.15 (0.50–2.56)
UA	reduced	73.29±8.98	4 : 3	7	0.58 (0.31–1.12)	5.354	<0.001^*	0.229	<0.001^*
	normal	69.54±7.24	150 : 182	332	0.60 (0.36–1.25)
	elevated	71.04±8.55	75 : 39	114	0.91 (0.50–1.81)
TCO2	reduced	69.36±7.00	64 : 56	120	0.80 (0.51–1.72)	2.799	0.026^*	–0.102	0.030^*
	normal	69.50±7.05	142 : 150	292	0.61 (0.38–1.21)
	elevated	75.17±11.01	23 : 18	41	0.52 (0.31–1.30)
GFR	reduced	72.13±8.64	118 : 85	203	0.81 (0.46–1.91)	7.360	<0.001^*	–0.260	<0.001^*
	normal	68.23±6.20	111 : 139	250	0.56 (0.34–1.04)
cysC (mg/L)	normal	68.23±6.20	111 : 139	250	0.56 (0.34–1.04)	9.446	<0.001^*	0.265	<0.001^*
	elevated	72.13±8.64	118 : 85	203	0.81 (0.46 1.91)

BUN, blood urea nitrogen; Cr, creatinine; UA, uric acid; TCO₂, total carbon dioxide; GFR, glomerular filtration rate; cysC, cystatin C. Age is expressed as mean±standard deviation, AD7c-NTP is expressed as median (interquartile range); asterisk indicates p value < 0.05; p^(a) represents the value of general linear model analysis, p^(b) represents the value of Spearman’s correlation analysis.

Correlation analyses between urinary AD7c-NTP levels and kidney indicator values

Further analysis presented in Table 4 indicates that positive correlations were observed between urinary AD7c-NTP levels and BUN (r = 0.210, p < 0.001), Cr (r = 0.202, p < 0.001), UA (r = 0.229, p < 0.001), and cysC (r = 0.265, p < 0.001). Additionally, a negative correlation were found between urinary AD7c-NTP levels and TCO₂ (r = –0.102, p = 0.030) and GFR (r = –0.260, p < 0.001).

Correlation analyses between urinary AD7c-NTP levels and seven most important hepatorenal indicators

Figure 2 illustrates the outcomes of the LASSO logistic regression model. The number of variables associated with predicting complete response was reduced to 7 using the LASSO regression method. The selected variables included ALB, A/G, AST, BUN, UA, GFR, and cysC. As indicated in Table 5, the hepatorenal variables significantly associated with predicting AD7c-NTP were A/G (p = 0.007), AST (p = 0.002), BUN (p = 0.019), and UA (p = 0.003).

Fig. 2

Variable selection using the least absolute shrinkage and selection operator (LASSO) binary logistic regression model. A) Cross-validation results; B) LASSO coefficient profiles were created against the log lambda sequence; C) Most important variables were selected using the LASSO regression model.

Table 5

The seven most important variables in Lasso regression were included in multiple linear regression analysis for predicting urinary Alzheimer-associated neuronal thread protein (AD7c-NTP)

Intercept and variables	Prediction model
	β	Standard error	t	p	CI (2.5%)	CI (97.5%)
Intercept	1.445	0.497	2.9	0.004	0.467	2.423
ALB	–0.010	0.014	–0.74	0.461	–0.037	0.017
A/G	–0.626	0.229	–2.74	0.007^*	–1.076	–0.176
AST	0.001	0.000	3.14	0.002^*	0.001	0.002
BUN	0.014	0.006	2.36	0.019^*	0.002	0.026
UA	0.001	0.000	2.96	0.003^*	0.000	0.002
GFR	0.000	0.003	–0.11	0.912	–0.007	0.006
cysC	0.123	0.087	1.42	0.157	–0.048	0.294

Parameters were estimated using the indicated model; asterisk indicates p value < 0.05.

DISCUSSION

AD7c-NTP has been identified as a biomarker due to its elevated levels in putative brain domains, CSF, and urine. However, the diagnostic efficacy of AD7c-NTP detected in urine is not as robust as that of AD7c-NTP detected in CSF. The metabolism of AD7c-NTP from the CSF to blood to urine via the liver and kidney is likely intricate. A “double-balanced theory” has been proposed to elucidate the in vivo stability of AD7c-NTP [24]. Nonetheless, the impact of liver metabolic function, kidney excretion function, and certain blood components on detected urinary AD7c-NTP during the blood metabolism process remains unclear. The objective of this study was to assess the influence of liver and kidney functions on urinary AD7c-NTP through measurable hepatorenal function indicators.

In this study, urinary AD7c-NTP levels exhibited a tendency to increase with age and were slightly higher in females than in males, consistent with previous research [23, 25]. Age is widely recognized as the most significant risk factor for AD dementia, with the incidence of AD and mild cognitive impairment escalating exponentially with advancing age. Given its association with age, urinary AD7c-NTP emerge as an aging-related protein. Furthermore, several studies have explored potential explanations for the higher prevalence of AD in women than men, including increased susceptibility to the adverse effects of head injuries, fewer educational opportunities, elevated risk of depression, longer life expectancy, APOE gene carriage, and sex steroid hormone levels [26, 27]. This observation aligns with previous findings that AD incidence increases with age and is higher in females than in males [28, 29]. Thus, there remains a need to establish urinary AD7c-NTP thresholds for different age groups and genders.

The liver serves as a significant metabolic and degradation organ in the body, participating in the clearance of various substances, including biomarkers associated with AD [30]. Research has indicated that approximately 13.9% of Aβ₄₂ and 8.9% of Aβ₄₀ in the blood are cleared by the liver [31]. However, the impact of liver function on AD7c-NTP levels has not been investigated previously. In this study, after adjusting for age and sex, no significant difference was observed in urinary AD7c-NTP levels among subjects with different D-BIL, TP, ALP, CHE, ALT, or AST levels, as indicated in Table 3. Conversely, previous studies have shown that participants with liver disease exhibit elevated levels of plasma Aβ₄₂, Aβ₄₀, and NFL [32, 33]. The discrepancy in results may arise from the susceptibility of plasma biomarkers to plasma composition, while urine biomarkers avoid this effect. AD7c-NTP, a 41 kD protein with an isoelectric point of 9.89, exhibits positive polarity in the blood at a pH between 7.35 and 7.45. In contrast, plasma proteins, with an isoelectric point of 4–6, are negatively charged, allowing them to closely combine and remain stable. Furthermore, the ability of substances to pass through the glomerular filtration membrane depends on their size and charge. AD7c-NTP, owing to its low molecular weight and positive polarity, can be filtered through the kidney, facilitated by the negatively charged glomerular basement membrane, which permits passage of low molecular weight substances (Schematic Fig. 1). Some studies even suggest that urine, lacking a homeostatic mechanism, may serve as a superior source of biomarkers [34 –36]. Consequently, we speculate that urinary AD7c-NTP concentrations may to some extent avoid interference from blood components. However, significant differences were observed in urinary AD7c-NTP levels between different T-BIL, I-BIL, ALB, Glo, A/G, and PA levels. Moreover, correlation analyses revealed a negative correlation between urinary AD7c-NTP and A/G levels, consistent with findings by Wang et al., who reported a negative correlation between plasma Aβ_42, Aβ₄₀ levels, and A/G values [37]. This suggests that ALB and Glo may interfere with AD7c-NTP excretion, similar to their effect on Aβ. Additionally, some studies have reported liver enzymes’ involvement in AD pathology [30, 38]. Our study suggests that CHE and AST may be correlated with detected AD7c-NTP levels in urine, warranting further investigation.

Although very limited research exists regarding factors associated with AD-related urine biomarkers, kidney function has been named a research priority in recent recommendations by the Alzheimer’s Association on the appropriate use of biomarkers [39]. Reduced kidney function has been associated with increased levels of NFL, p-tau, t-tau, and Aβ in blood [40 –42]. It is well understood that the decline of kidney function, especially the GFR, leads to an increase in the physical components of the blood that cannot be efficiently filtered out. Moreover, urine levels of these components should theoretically be reduced. However, our results indicated that the urinary AD7c-NTP levels were higher in the groups with elevated BUN, elevated Cr, elevated UA, reduced TCO₂, reduced GFR, and elevated cysC than their corresponding normal groups. The correlation analysis revealed a positive correlation between urinary AD7c-NTP with BUN, Cr, UA, and cysC values and a negative correlation with TCO₂ and GFR values. Two reasons may account for this contradictory phenomenon. First, urinary AD7c-NTP was pseudo elevated in participants with renal dysfunction. Goodman et al. reported that urinary AD7c-NTP was more accurate when urine Cr was between 50 and 225 mg/dL. When Cr was less than 50 mg/dL, excessive urinary dilution might result in false negatives; and when Cr was greater than 225 mg/dL, high concentrations of non-specific excess solutes might lead to false-positive results [17]. Second, it has been published that kidney function itself is associated with the risk of dementia, including AD, concomitantly associated with elevated AD biomarkers [43, 44]. The literature regarding this association, however, has been inconsistent, with several studies reporting that reduced kidney function was solely associated with increased levels of dementia-related blood biomarkers but not with increased dementia risk [45 –48]. Therefore, the relationship between renal function and urine AD7c-NTP needs to be further explored in longitudinal follow-up studies.

Taking into account potential interactions between hepatic and renal function indexes on urinary AD7c-NTP levels, we utilized the LASSO method and multivariable logistic regression analysis to analyze statistically significant hepatorenal indicators. Ultimately, A/G, AST, BUN, and UA were identified as statistically significant hepatorenal variables for predicting AD7c-NTP. Importantly, these indicators were also meaningful based on previous analyses. By comprehensively evaluating the potential effects of hepatorenal function indicators on urinary AD7c-NTP levels, our study suggests that in the clinical application of AD7c-NTP, liver and kidney function indicators, particularly AST, BUN, UA, and cysC, should be simultaneously monitored as auxiliary indices. This study is the first to comprehensively evaluate the effects of hepatorenal function indicators on urinary AD7c-NTP levels, underscoring the importance of considering hepatic and renal functions in the development and application of body-fluid-based biomarkers for assessing neurological disorders.

Limitations

Our study has several limitations. First, the sample size was relatively small, and all participants were Chinese, potentially limiting the generalizability of the findings. Second, being a cross-sectional study, our ability to observe the dynamic relationship between urinary AD7c-NTP levels and blood biochemical indicators was limited, a longitudinal study would be beneficial in this regard. Third, due to the scarcity of previous research on this topic, further validation of our results through multi-center studies is warranted.

Conclusion

This study assessed the impact of hepatorenal function indicators on urinary AD7c-NTP levels and proposed that these indicators should be taken into account in the clinical utilization of urinary AD7c-NTP.

AUTHOR CONTRIBUTIONS

He Jin (Conceptualization; Funding acquisition; Investigation; Project administration; Writing – original draft); Qiu Yang (Data curation; Methodology); Guodong Chen (Methodology); Wei Zhang (Formal analysis; Software); Yanchuan Wu (Methodology); Rong Wang (Conceptualization; Funding acquisition; Writing – review & editing).

Footnotes

ACKNOWLEDGMENTS

We thank all participants of this study. We thank LetPub () for its linguistic assistance during the preparation of this manuscript.

FUNDING

This article was supported by the National Key Research and Development Program of China (grant no. 2022YFC2403504), National Natural Science Foundation of China (grant no. 82160258), Guangxi Natural Science Foundation (grant no. 2023GXNSFBA026173), and Guilin technology application and promotion plan (grant no. 20220139-7-4).

CONFLICT OF INTEREST

The authors have no conflict of interest to report.

DATA AVAILABILITY

The data supporting the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

The supplementary material is available in the electronic version of this article: .

References

Sperling

, Aisen

, Beckett

, Bennett

, Craft

, Fagan

, Iwatsubo

, Jack

, Kaye

, Montine

, Park

, Reiman

, Rowe

, Siemers

, Stern

, Yaffe

, Carrillo

, Thies

, Morrison-Bogorad

, Wagster

, Phelps

(2011) Toward defining the preclinical stages of Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 280–292.

Tatulian

(2022) Challenges and hopes for Alzheimer’s disease. Drug Discov Today 27, 1027–1043.

Rodríguez-Gómez

, Rodrigo

, Iradier

, Santos-Santos

, Hundemer

, Ciudin

, Sannemann

, Zwan

, Glaysher

, Wimo

, Bonn

, Johansson

, Rodriguez

, Alegret

, Gove

, Pinó

, Trigueros

, Kivipelto

, Mathews

, Ciudad

, Ferreira

, Bintener

, Gurruchaga

, Westman

, Belger

, Valero

, Maguire

, Krivec

, Kramberger

, Simó

, Garro

, Visser

, Dumas

, Georges

, Jessen

, Winblad

, Shering

, Stewart

, Campo

, Boada

, Consortium

(2019) The MOPEAD project: Advancing patient engagement for the detection of “hidden” undiagnosed cases of Alzheimer’s disease in the community. Alzheimers Dement 15, 828–839.

Galvin

, Aisen

, Langbaum

, Rodriguez

, Sabbagh

, Stefanacci

, Stern

, Vassey

, Wilde

A de

, West

, Rubino

(2021) Early stages of Alzheimer’s disease: Evolving the care team for optimal patient management. Front Neurol 11, 592302.

Jack

, Bennett

, Blennow

, Carrillo

, Dunn

, Haeberlein

, Holtzman

, Jagust

, Jessen

, Karlawish

, Liu

, Molinuevo

, Montine

, Phelps

, Rankin

, Rowe

, Scheltens

, Siemers

, Snyder

, Sperling

, Contributors, Elliott

, Masliah

, Ryan

, Silverberg

(2018) NIA-AA Research Framework: Toward a biological definition of Alzheimer’s disease. Alzheimers Dement 14, 535–562.

Blennow

, Zetterberg

(2018) The past and the future of Alzheimer’s disease fluid biomarkers. J Alzheimers Dis 62, 1125–1140.

Jack

, Bennett

, Blennow

, Carrillo

, Feldman

, Frisoni

, Hampel

, Jagust

, Johnson

, Knopman

, Petersen

, Scheltens

, Sperling

, Dubois

(2016) A/T/N: An unbiased descriptive classification scheme for Alzheimer disease biomarkers. Neurology 87, 539–547.

Villemagne

, Fodero-Tavoletti

, Masters

, Rowe

(2015) Tau imaging: Early progress and future directions. Lancet Neurol 14, 114–124.

Teunissen

, Verberk

IMW

, Thijssen

, Vermunt

, Hansson

, Zetterberg

, Flier

WM van der

, Mielke

, Campo

M del

(2022) Blood-based biomarkers for Alzheimer’s disease: Towards clinical implementation. Lancet Neurol 21, 66–77.

10.

Huang

, Wang

Y-J

, Guo

(2022) Biofluid biomarkers of Alzheimer’s disease: Progress, problems, and perspectives. Neurosci Bull 38, 677–691.

11.

Khan

, Alkon

(2015) Peripheral biomarkers of Alzheimer’s disease. J Alzheimers Dis 44, 729–744.

12.

Tan

C-C

, Yu

J-T

, Tan

(2014) Biomarkers for preclinical Alzheimer’s disease. J Alzheimers Dis 42, 1051–1069.

13.

Monte

, Ghanbari

, Frey

, Beheshti

, Averback

, Hauser

, Ghanbari

, Wands

(1997) Characterization of the AD7C-NTP cDNA expression in Alzheimer’s disease and measurement of a 41-kD protein in cerebrospinal fluid. J Clin Investig 100, 3093–3104.

14.

Ghanbari

, Ghanbari

(1998) A sandwich enzyme immunoassay for measuring AD7C-NTP as an Alzheimer’s disease marker: AD7C test. J Clin Lab Anal 12, 223–226.

15.

Munzar

, Levy

, Rush

, Averback

(2002) Clinical study of a urinary competitve ELISA for neural thread protein in Alzheimer disease. Neurol Clin Neurophysiol 2002, 2–8.

16.

Monte

SM de la

, Carlson

, Brown

, Wands

(1996) Profiles of neuronal thread protein expression in Alzheimer’s disease. J Neuropathol Exp Neurol 55, 1038–1050.

17.

Goodman

, Golden

, Flitman

, Xie

, McConville

, Levy

, Zimmerman

, Lebedeva

, Richter

, Minagar

, Averback

(2007) A multi-center blinded prospective study of urine neural thread protein measurements in patients with suspected Alzheimer’s disease. J Am Med Dir Assoc 8, 21–30.

18.

Levy

, McConville

, Lazaro

, Averback

(2007) Competitive ELISA studies of neural thread protein in urine in Alzheimer’s disease. J Clin Lab Anal 21, 24–33.

19.

Youn

, Park

K-W

, Han

S-H

, Kim

(2011) Urine neural thread protein measurements in Alzheimer disease. J Am Med Dir Assoc 12, 372–376.

20.

, Chen

, Wang

, Han

, Zhang

, Dong

, Zhang

, Wu

, Zhao

(2015) The level of Alzheimer-associated neuronal thread protein in urine may be an important biomarker of mild cognitive impairment. J Clin Neurosci 22, 649–652.

21.

Jin

, Wang

, Liu

, Jia

, Wu

, Zhao

, Wang

, Zhang

(2018) Some methodological characteristics of Alzheimer-associated urine neuronal thread protein detected by enzyme-linked immunosorbent assay. J Alzheimers Dis 63, 255–262.

22.

Jin

, Guan

, Wang

, Fang

, Liu

, Wu

, Zhang

, Liu

(2018) The distribution of urinary Alzheimer-associated neuronal thread protein and its association with common chronic diseases in the general population. J Alzheimers Dis 65, 433–442.

23.

, Guan

, Jin

, Liu

, Kang

, Wang

, Sheng

, Sun

, Li

, Fang

, Wang

(2020) The relationship between urinary Alzheimer-associated neuronal thread protein and blood biochemical indicators in the general population. Aging (Albany NY) 12, 15260–15280.

24.

Jin

, Wang

(2021) Alzheimer-associated neuronal thread protein: Research course and prospects for the future. J Alzheimers Dis 80, 963–971.

25.

, Chen

, Wang

, Han

, Zhang

, Dong

, Zhao

, Liu

, Chu

(2014) Alzheimer-associated urine neuronal thread protein level increases with age in a healthy Chinese population. J Clin Neurosci 21, 2118–2121.

26.

Mazure

, Swendsen

(2016) Sex differences in Alzheimer’s disease and other dementias. Lancet Neurol 15, 451–452.

27.

Altmann

, Tian

, Henderson

, Greicius

, Alzheimer’s Disease Neuroimaging Investigators (2014) Sex modifies the APOE-related risk of developing Alzheimer disease. Ann Neurol 75, 563–573.

28.

Alzheimer’s Association Report. (2023) 2023 Alzheimer’s disease facts and figures. Alzheimers Dement 19, 1598–1695.

29.

Jia

, Quan

, Fu

, Zhao

, Li

, Wei

, Tang

, Qin

, Wang

, Qiao

, Shi

, Wang

Y-J

, Du

, Zhang

, Luo

, Qu

, Zhou

, Gauthier

, Jia

, Group for the Project of Dementia Situation in China (2020) Dementia in China: Epidemiology, clinical management, and research advances. Lancet Neurol 19, 81–92.

30.

Bassendine

, Taylor-Robinson

, Fertleman

, Khan

, Neely

(2020) Is Alzheimer’s disease a liver disease of the brain? J Alzheimers Dis 75, 1–14.

31.

Cheng

, Tian

D-Y

, Wang

Y-J

(2020) Peripheral clearance of brain-derived Aβ in Alzheimer’s disease: Pathophysiology and therapeutic perspectives. Transl Neurodegener 9, 16.

32.

Peng

, Duggan

, Dark

, Daya

, An

, Davatzikos

, Erus

, Lewis

, Moghekar

, Walker

(2022) Association of liver disease with brain volume loss, cognitive decline, and plasma neurodegenerative disease biomarkers. Neurobiol Aging 120, 34–42.

33.

Zhang

, Zhang

, Wang

, Cheng

, Wang

, Qiao

, Peng

, Alzheimer’s Disease Neuroimaging Initiative (2024) Associations of liver function with plasma biomarkers for Alzheimer’s disease. Neurol Sci 45, 2625–2631.

34.

(2014) Urine reflection of changes in blood. Adv Exp Med Biol 845, 13–19.

35.

Gao

(2013) Urine—an untapped goldmine for biomarker discovery? Sci China Life Sci 56, 1145–1146.

36.

, Gao

(2015) Urinary biomarkers of brain diseases. Genom Proteom Bioinform 13, 345–354.

37.

Wang

Y-R

, Wang

Q-H

, Zhang

, Liu

Y-H

, Yao

X-Q

, Zeng

, Li

, Zhou

F-Y

, Wang

, Yan

J-C

, Zhou

H-D

, Wang

Y-J

(2017) Associations between hepatic functions and plasma amyloid-beta levels—implications for the capacity of liver in peripheral amyloid-beta clearance. Mol Neurobiol 54, 2338–2344.

38.

Nho

, Kueider-Paisley

, Ahmad

, MahmoudianDehkordi

, Arnold

, Risacher

, Louie

, Blach

, Baillie

, Han

, Kastenmüller

, Trojanowski

, Shaw

, Weiner

, Doraiswamy

, Duijn

C van

, Saykin

, Kaddurah-Daouk

, Consortium ADNI and the ADM (2019) Association of altered liver enzymes with Alzheimer disease diagnosis, cognition, neuroimaging measures, and cerebrospinal fluid biomarkers. JAMA Netw Open 2, e197978.

39.

Hansson

, Edelmayer

, Boxer

, Carrillo

, Mielke

, Rabinovici

, Salloway

, Sperling

, Zetterberg

, Teunissen

(2022) The Alzheimer’s Association appropriate use recommendations for blood biomarkers in Alzheimer’s disease. Alzheimers Dement 18, 2669–2686.

40.

Syrjanen

, Campbell

, Algeciras-Schimnich

, Vemuri

, Graff-Radford

, Machulda

, Bu

, Knopman

, Jack

, Petersen

, Mielke

(2022) Associations of amyloid and neurodegeneration plasma biomarkers with comorbidities. Alzheimers Dement 18, 1128–1140.

41.

Akamine

, Marutani

, Kanayama

, Gotoh

, Maruyama

, Yanagida

, Sakagami

, Mori

, Adachi

, Kozawa

, Maeda

, Otsuki

, Matsuoka

, Iwahashi

, Shimomura

, Ikeda

, Kudo

(2020) Renal function is associated with blood neurofilament light chain level in older adults. Sci Rep 10, 20350.

42.

Mielke

, Dage

, Frank

, Algeciras-Schimnich

, Knopman

, Lowe

, Bu

, Vemuri

, Graff-Radford

, Jack

, Petersen

(2022) Performance of plasma phosphorylated tau 181 and 217 in the community. Nat Med 28, 1398–1405.

43.

, Garcia-Ptacek

, Trevisan

, Evans

, Lindholm

, Eriksdotter

, Carrero

(2021) Kidney function, kidney function decline, and the risk of dementia in older adults: A registry-based study. Neurology 96, e2956–e2965.

44.

Kjaergaard

, Johannesen

, Sørensen

, Henderson

, Christiansen

(2021) Kidney disease and risk of dementia: A Danish nationwide cohort study. BMJ Open 11, e052652.

45.

Stocker

, Beyer

, Trares

, Perna

, Rujescu

, Holleczek

, Beyreuther

, Gerwert

, Schöttker

, Brenner

(2023) Association of kidney function with development of Alzheimer disease and other dementias and dementia-related blood biomarkers. JAMA Netw Open 6, e2252387.

46.

Gabin

, Romundstad

, Saltvedt

, Holmen

(2019) Moderately increased albuminuria, chronic kidney disease and incident dementia: The HUNT study. BMC Nephrol 20, 261.

47.

Helmer

, Stengel

, Metzger

, Froissart

, Massy

Z-A

, Tzourio

, Berr

, Dartigues

J-F

(2011) Chronic kidney disease, cognitive decline, and incident dementia. Neurology 77, 2043–2051.

48.

O’Hare

, Walker

, Haneuse

, Crane

, McCormick

, Bowen

, Larson

(2012) Relationship between longitudinal measures of renal function and onset of dementia in a community cohort of older adults. J Am Geriatr Soc 60, 2215–2222.