Serum PCSK9 levels in infants with deviant birth weight: a biomarker of the lipoprotein metabolism

Abstract

Objective

Proprotein Convertase Subtilisin/Kexin-Type 9 (PCSK9), a modulator of low-density lipoprotein (LDL) cholesterol metabolism, has been reported to be a promising biomarker for evaluating lipoprotein metabolism; however, evidence in infants is limited. In the current study, we sought to investigate potential differences in serum PCSK9 levels between infants with deviant birth weight and controls.

Methods

We enrolled 82 infants, classified into 33 small (SGA), 32 appropriate (AGA), and 17 large for gestation (LGA) infants. Serum PCSK9 was measured on routine blood analysis within the first postnatal 48 h.

Results

PCSK9 was significantly higher in SGA as compared to AGA and LGA infants [322 (236–431) as compared to 263 (217–302) and 218 (194–291) ng/ml respectively, p = .011]. In comparison to term AGA infants, PCSK9 was significantly elevated in preterm AGA and SGA infants. We also found a significantly higher level of PCSK9 in term female SGA infants as compared to term male SGA infants [325 (293–377) as compared to 174 (163–216) ng/ml, p = .011]. PCSK9 was significantly correlated with gestational age (R = –0.404, p < .001), birth weight (R = –0.419, p < .001), total cholesterol (R = 0.248, p = .028) and LDL cholesterol (R = 0.370, p = .001). SGA status (OR 2.56, p = .004, 95% CI 1.83–4.28) and prematurity (OR 3.10, p = .001, 95% CI 1.39–4.82) were strongly related to serum PCSK9 levels.

Conclusion

PCSK9 levels were significantly associated with total and LDL cholesterol. Moreover, PCSK9 levels were higher in preterm and SGA infants, suggesting that PCSK9 might be a promising biomarker for evaluating infants with increased later cardiovascular risk.

Highlights

• What’s already known? Proprotein Convertase Subtilisin/Kexin-Type 9 (PCSK9) is a promising biomarker for evaluating lipoprotein metabolism; however, evidence in infants is limited. Infants that were born with a deviant birth weight have a unique lipoprotein metabolism profile.

• What this study adds? Serum PCSK9 levels were significantly associated with total and LDL cholesterol. PCSK9 levels were higher in preterm and small for gestation infants, suggesting that PCSK9 might be a promising biomarker for evaluating infants with increased later cardiovascular risk.

Keywords:

• Cardiovascular risk
• lipoprotein
• metabolism
• prematurity
• small for gestational age

Introduction

Lipoprotein is a crucial component of fetal neuronal development and a precursor to the synthesis of steroid hormones [1]. Furthermore, low-density lipoprotein (LDL) is a well-recognized risk factor for atherosclerosis in childhood, adolescence, and adulthood [2]. Lipoprotein homeostasis during fetal life is prone to disruption in pathological conditions such as prematurity or growth deviation [3]. Infants born small for gestational age (SGA), i.e. those that fail to achieve a standard weight threshold, usually the tenth centile on growth charts [4], have a unique lipoprotein metabolic profile, namely markedly greater triglycerides, and LDL cholesterol [5,6], whereas, large for gestational age born infants (LGA) are also in an increased risk of metabolic syndrome, in comparison to appropriate for gestational age born infants (AGA) [7,8].

It has been suggested that the intrauterine environment during critical periods of fetal development, either in the terms of malnutrition or over-nutrition, could program components of offspring of dyslipidemia or a later metabolic syndrome [9]. According to the Developmental Origin of Health and Disease theory, adverse intrauterine events could lead to the disruption of fetal metabolic programming [10], whereas epigenetic modification, i.e. changes in cellular or tissue function without genetic changes, has been proposed as the potential key factor [11]. However, the mechanisms regulating lipoprotein homeostasis during pregnancy and under pathological conditions have not been fully evaluated.

Proprotein Convertase Subtilisin/Kexin-Type 9 (PCSK9) is a serine protease that was discovered in 2003 [12]. Since then, extensive research on this protein has emerged PCSK9 as a significant modulator of LDL metabolism [13,14]. PCSK9 acts by binding to the LDL receptors, leading to lower LDL turnover rates and higher circulating LDL cholesterol concentrations [15]. To date, limited studies have examined the serum concentrations of PCSK9 in infants, and thus, its potential role in lipoprotein metabolism in early life remains inconclusive [1,16].

We hypothesized that term and preterm infants with deviant birth weights might have elevated serum PCSK9 levels in comparison to infants with appropriate birth weights. Therefore, in the current study, we examined the serum concentrations of PCSK9 amongst SGA, LGA, and AGA infants, and we also examined the correlation of PCSK9 with significant perinatal factors, including gestational age, birth weight, gender, bodyweight deviation, and the lipid profile.

Materials and methods

We conducted a cross-sectional study at the University Hospital of Ioannina, from December 2020 until July 2021. The University Hospital is a regional tertiary center where an average of 1500 infants are born annually; the neonatal unit is a referring level III center with a mean of 450 admissions per year.

In the current study, we enrolled both term and preterm (<37 weeks of gestational age) infants that were further classified based on their birth weight for gestation. Infants were defined as SGA based on a birth weight below the tenth centile, according to the population Fenton charts [17]. Infants with a birth weight above the 90th centile, according to the same growth charts, were defined as LGA. Finally, children with a birth weight between the 10–90th centile were defined as AGA. For every infant with deviant birth weight, we enrolled an AGA counterpart matched for gestational age (±one week) and gender. Infants with congenital anomalies, liver disease, clinical or laboratory evidence of early-onset sepsis, or perinatal asphyxia were excluded.

The perinatal and neonatal characteristics were recorded, including gestational age, anthropometrics at birth, gender, mode of delivery, gestational diabetes, hypertension, mode of conception, feeding initiation, and length of stay.

Serum PCSK9, as well as triglycerides, total cholesterol, LDL cholesterol, and high-density lipoprotein (HDL) cholesterol levels, were measured on routine blood analysis within the first postnatal 48 h, whereas no additional blood samples were obtained from the infants. The remaining centrifuged plasma supernatants were stored at −80 °C. Upon completion of the sample collection, the PCSK9 was measured in duplicate by a sandwich enzyme-linked immunosorbent assay kit (R&D Systems, Minneapolis, MN, USA).

Statistical analysis

Continuous variables were expressed as mean ± standard deviation or median (interquartile range), as appropriate. The normality of the distributions of continuous variables was assessed by the Kolmogorov–Smirnov test. Comparisons of continuous variables were performed utilizing the student’s unpaired t-test, the one-way ANOVA with Bonferroni posthoc analysis, the non-parametric Mann–Whitney test, or the non-parametric Kruskal–Wallis test. Categorical variables were expressed as n (percentage %) and compared with the chi-square test or Fisher’s exact test. An analysis between infants with deviant and appropriate birth weight in the subgroups of term and preterm infants was also performed. Spearman’s rho correlation was used to evaluate the association of PCSK9 levels with continuous variables including gestational age, birth weight, total cholesterol, LDL cholesterol, and HDL cholesterol. A linear regression analysis was used to examine the association of PCSK9 levels (continuous variable) with deviant birth weight adjusted for prematurity and gender.

All tests were two-sided and a p-value less than 0.05 was considered statistically significant (alpha 0.05). A power analysis revealed that a sample size of at least 32 infants for each subgroup analysis would be sufficient to detect a difference of 25% in serum PCSK9 levels between groups, with a power of 0.8 and a type-I error of 0.05. The data were analyzed using SPSS Statistics (IBM SPSS Statistics for Windows, Version 24.0. Armonk, NY, USA). Informed consent was obtained from the caregivers of all infants, and the study was approved by the Ethical Committee of the Institution (No 860/05.11.2020).

Results

A total of 82 infants were enrolled: 33 SGA, 32 AGA, and 17 LGA infants. The perinatal and neonatal characteristics of the study cohort are depicted in Table 1. Apart from the anthropometrics, no significant differences were noted regarding the remaining perinatal characteristics. Of the total cohort, 48 infants had been born at term (16 SGA, 15 AGA, 17 LGA), whereas 34 infants were preterm (17 SGA and 17 AGA).

Table 1. Perinatal characteristics in the total cohort, and the subgroup of term and preterm infants.

Total cohort	SGA (n = 33)	AGA (n = 32)	LGA (n = 17)	p Value
Gestational age, weeks	35.6 ± 2.4	35.6 ± 3.3	38.6 ± 1.1	.001
Birth weight, g	2075 ± 552	2585 ± 696	4127 ± 265	<.001
Birth length, cm	44.9 ± 3.9	47.4 ± 4.1	52.6 ± 2.4	<.001
Head circumference, cm	31.5 ± 2.3	32.5 ± 2.5	35.9 ± 0.9	<.001
Gender, male	18 (55%)	18 (56%)	12 (70%)	.549
Mode of delivery, cesarean section	28 (85%)	21 (66%)	8 (47)	.021
Gestational diabetes	6 (18%)	10 (31%)	2 (12%)	.243
Gestational hypertension	4 (12%)	2 (6)	–	.310
Singleton pregnancy	24 (73%)	27 (84%)	17 (100%)	.121
Conception, in vitro fertilization	8 (24%)	9 (28%)	–	.056
Breast milk feeding	25 (76%)	21 (66%)	13 (77%)	.734
Length of stay, days	6 (3–24)	6 (3–18)	3 (3–5)	.043
Term infants	SGA (n = 16)	AGA (n = 15)	LGA (n = 17)	p value
Gestational age, weeks	38 ± 0.7	38.4 ± 0.9	38.6 ± 1	.108
Birth weight, g	2519 ± 230	3207 ± 302	4127 ± 265	<.001
Birth length, cm	48 ± 1.8	50.4 ± 2	52.6 ± 2.4	<.001
Head circumference, cm	33.3 ± 0.9	34.4 ± 0.8	35.9 ± 0.9	<.001
Gender, male	6 (38%)	9 (60%)	12 (70%)	.167
Mode of delivery, cesarean section	11 (69%)	6 (53%)	8 (47)	.475
Gestational diabetes	4 (25%)	3 (20%)	2 (12%)	.243
Gestational hypertension	1 (6%)	–	–	.310
Singleton pregnancy	13 (81%)	15 (100%)	17 (100%)	.059
Conception, in vitro fertilization	2 (13%)	1 (7%)	–	.402
Breast milk feeding	10 (63%)	5 (33%)	13 (77%)	.045
Length of stay, days	3 (3–4)	4 (3–5)	3 (3–5)	.411
Preterm infants	SGA (n = 17)	AGA (n = 17)		p Value
Gestational age, weeks	33.6 ± 1.9	33.3 ± 2.5		.484
Birth weight, g	1658 ± 422	2036 ± 420		.013
Birth length, cm	42 ± 3	44.3 ± 3.4		.044
Head circumference, cm	29.8 ± 1.9	30.9 ± 2.4		.162
Gender, male	12 (71%)	9 (53%)		.481
Mode of delivery, cesarean section	17 (100%)	13 (77%)		.103
Gestational diabetes	2 (18%)	7 (41%)		.118
Gestational hypertension	3 (17%)	2 (12%)		1.000
Singleton pregnancy	11 (65%)	12 (71%)		1.000
Conception, in vitro fertilization	6 (35%)	8 (47%)		.728
Breast milk feeding	15 (78%)	16 (86%)		1.000
Length of stay, days	24 (14–34)	15 (6–30)		.474

SGA: small of gestation; AGA: appropriate for gestation; LGA: large for gestation. Continuous variables are expressed as mean ± standard deviation, or median (interquartile range). Categorical variables are expressed as n (percentage).

Overall, serum PCSK9 was found significantly higher in SGA infants as compared to AGA and LGA counterparts [322 (236–431) as compared to 263 (217–302) and 218 (194–291) ng/ml respectively, p = .011]. Among term infants, no significant differences were noted in serum PCSK9 between SGA, AGA, and LGA subjects; however, in the subgroup of preterms, SGA infants had a significantly higher serum PCSK9 in comparison to AGA infants [379 (308–551) as compared to 284 (221–336) ng/ml, p = .006] (Table 2). In comparison to term AGA infants, serum PCSK9 was significantly elevated in preterm AGA [284 (221–336) as compared to 242 (207–287) ng/ml, p = .049] and in preterm SGA infants [379 (308–551) as compared to 242 (207–287) ng/ml, p < .001].

Table 2. PCSK9 and lipids in the total cohort, and the subgroup of term and preterm infants.

Total cohort	SGA (n = 33)	AGA (n = 32)	LGA (n = 17)	p Value
PCSK9, ng/ml	322 (236–431)	263 (217–302)	218 (194–291)	.011
Triglycerides, mg/dl	116 ± 61	90 ± 46	101 ± 37	.152
Total cholesterol, mg/dl	93.4 ± 26.3	104.3 ± 41.8	97.7 ± 44.7	.518
HDL cholesterol, mg/dl	27.3 ± 10.1	31.6 ± 12.5	27.9 ± 8.2	.265
LDL cholesterol, mg/dl	48.9 ± 24.7	56.6 ± 36.2	45.8 ± 13	.385
Term infants	SGA (n = 16)	AGA (n = 15)	LGA (n = 17)	p Value
PCSK9, ng/ml	264 (174–352)	242 (207–287)	218 (194–291)	.625
Triglycerides, mg/dl	148 ± 62	106 ± 39	101 ± 37	.017
Total cholesterol, mg/dl	93.7 ± 24.9	86.8 ± 29.3	97.7 ± 44.7	.681
HDL cholesterol, mg/dl	25.1 ± 6.3	23.9 ± 9.5	27.9 ± 8.2	.398
LDL cholesterol, mg/dl	41.5 ± 21.4	45.9 ± 13	45.8 ± 13	.758
Preterm infants	SGA (n = 17)	AGA (n = 17)		p Value
PCSK9, ng/ml	379 (308–551)	284 (221–336)		.006
Triglycerides, mg/dl	86.7 ± 43.3	77.4 ± 48.7		.573
Total cholesterol, mg/dl	93.1 ± 28.3	119.7 ± 45.9		.058
HDL cholesterol, mg/dl	29.4 ± 12.7	38.9 ± 10.6		.035
LDL cholesterol, mg/dl	48.9 ± 24.7	56.6 ± 36.2		.273

PCSK9: Proprotein Convertase Subtilisin/Kexin-Type 9; SGA: small of gestation; AGAL: appropriate for gestation; LGA: large for gestation; HDL: high-density lipoprotein; LDL: low-density lipoprotein. Continuous variables are expressed as mean ± standard deviation or median (interquartile range).

In the subgroup analysis based on gender, significantly higher levels of PCSK9 were evident in term female SGA infants compared with term male SGA infants [325 (293–377) as compared to 174 (163–216) ng/ml, p = .011]. No differences were recorded, whatsoever, between males and females based on their birth weight for gestation status (Supplemental Table 1, Supplemental Figure 1).

A significant correlation was found between serum PCSK9 and gestational age (R=–0.404, p < .001), birth weight (R=–0.419, p < .001), total cholesterol (R = 0.248, p = .028) and LDL-cholesterol (R = 0.370, p = .001). In multivariate analysis, birth weight deviation (SGA status) (OR 2.56, p = .004, 95% CI 1.83–4.28) and prematurity (OR 3.10, p = .001, 95% CI 1.39–4.82), but not the gender (OR 0.48, p = .574, 95% CI 0.21–2.17), were strongly related to serum PCSK9 levels (Supplemental Table 2).

Discussion

In the current study, we found a significant correlation between PCSK9 levels and both total and LDL cholesterol in infants. Our findings also showed that PCSK9 was elevated in SGA-born infants, as well as in preterm SGA in comparison to AGA-born infants. Besides lipidemic parameters, PCSK9 correlated with gestational age, birth weight, and SGA status.

Previous studies have suggested that PCSK9 might have an important role in regulating total, and LDL-cholesterol levels during the fetal period [18]. Besides, lipoprotein is a crucial component of fetal neuronal development, and previous reports have shown that maternal serum PCSK9 levels were lower in pregnant women carrying fetuses with open neural tube defects, in comparison to those of healthy subjects, suggesting that PCSK9 could be used as a molecular biomarker for the noninvasive prenatal screening [19]. Moreover, there is evidence that circulating PCSK9 is unlikely to be transported from maternal to fetal circulation due to its relatively large molecular size [18]. Therefore, fetuses most probably regulate their lipid levels, whereas serum levels of PCSK9 at birth might reflect the ability of the fetus to carry out protein synthesis [20]. Genetic factors regarding PCSK9 expression have been found to contribute to individual circulating PCSK9 and serum LDL cholesterol levels [21,22]. Furthermore, in SGA-born infants, epigenetic modification without genetic changes is more easily induced by the adverse intrauterine environment during germ cells and fetal development [11,23]. In fact, events occurring during the critical period in early life, and epigenetic rather than genetic antecedents may have the most significant impact on metabolic programming [24]. Studies in adults have supported that serum PCSK9 is an important contributor to LDL-cholesterol homeostasis, and research has advanced to developing monoclonal antibodies for the inhibition of PCSK9 action in order to treat hypercholesterolemia; however, evidence in infants remains scarce [25]. In that aspect, our findings showing a significant correlation between PCSK9 levels and both total and LDL cholesterol in infants might add to the current evidence indicating that PCSK9 could be a promising biomarker of lipoprotein metabolism in infants [1,16].

In a previous study, Araki et al. evaluated serum PCSK9 levels in term and preterm infants soon after birth [1]. Among 81 neonates, 23 were SGA whereas the remaining were AGA-born infants. The authors found that serum PCSK9 levels correlated with triglycerides and LDL cholesterol, whereas gestational age and PCSK9 concentrations were independently associated with the serum LDL cholesterol [1]. Although there was a direct correlation between PCSK9 levels and gestational age with the lipid profile, no direct association between serum PCSK9 and gestational age could be established. Also, the authors reported that there were no significant differences in the PCSK9 levels between SGA and AGA infants [1]. Our findings were in part in agreement with the above study, showing that PCSK9 presented a significant correlation with gestational age, birth weight, total cholesterol, and LDL cholesterol. Furthermore, in our study, SGA status and prematurity were independently associated with serum PCSK9 levels. Contrary to our study, in the study by Araki et al. term and preterm infants were examined altogether with no subgroup analysis based on prematurity [1]. Differences in the study methodology might explain the discrepancies between our and the aforementioned study.

A subsequent study that was performed by Pecks et al. compared PCSK9 levels in the umbilical cord blood of 172 infants, namely 70 infants with intrauterine growth restriction and 102 control infants [16]. Neonatal serum PCSK9 and LDL-cholesterol concentration levels were negatively associated with gestational age. In multivariate analysis, PCSK9 levels remained an independent predictor of fetal LDL-cholesterol concentrations. Interestingly, the authors found significantly lower concentration levels of PCSK9 in the umbilical cord blood of growth-restricted compared to non-growth-restricted infants [16]. Of note, the authors utilized robust definitions of intrauterine growth restriction and SGA infants using both population and customized centiles; however, as per the study design, the groups of growth-restricted and control infants had not been matched for gestational age, whereas multiple gestations, or maternal diabetes mellitus, gestational diabetes, or other severe maternal metabolic disorders comprised some of the exclusion criteria [16]. The contradictive results between our and the latter study should be interpreted in light of differences in sample characteristics. Notwithstanding, lower gestational age has been associated with higher serum PCSK9 levels, whereas previous evidence has suggested that pregnancies complicated by gestational diabetes were associated with lower placental PCSK9 expression and lower maternal LDL cholesterol levels [26].

In our study, we recorded a higher serum PCSK9 level in females as compared to males, only amongst term SGA infants. Our findings are in line with the previous study by Araki et al. which suggests that serum PCSK9 levels are influenced by gender [1]. Indeed, Araki et al. demonstrated that serum PCSK9 levels in female infants were significantly higher than in male infants [1]. The impact of the female gender on the circulating PCSK9 levels has been also suggested in previous studies in adults [22,27,28], where plasma PCSK9 levels of women were significantly higher in comparison to men [22]. Moreover, gender and age had a significant influence on PCSK9 levels in puberty, similar to the effects on triglycerides and LDL cholesterol [27]. Based on current evidence, it seems that sex hormones might affect PCSK9 levels in serum, and thus, influence the lipid profile.

The limitations of our study should be acknowledged. The main limitation could be the relatively small sample size of the study population, although it was comparable to previous studies. Given that this was a single-center study, further evidence is warranted before generalizing our findings. It should be also noted that there were no maternal data regarding PSCK9 levels and lipid profile; however, available evidence as mentioned supports the notion that circulating PCSK9 is unlikely to be transported from the maternal to fetal circulation [18].

Conclusion

Serum PCSK9 levels were significantly associated with total and LDL cholesterol. Moreover, PCSK9 levels were higher in preterm and SGA infants, suggesting that PCSK9 might be a promising biomarker for evaluating infants with increased later cardiovascular risk. Further studies are warranted to evaluate the role of genetic variations in PCSK9 individual expression.

Ethics approval

Informed consent were obtained from the caregivers of all infants, and the study was approved by the Ethical Committee of the Institution (No 860/05.11.2020).

Acknowledgments

The authors acknowledge the technical assistance of Afroditi Papagianni.

Disclosure statement

HM reports receiving honoraria, consulting fees, and non-financial support from healthcare companies including Amgen, Pfizer, and Sanofi. The other coauthors report no conflict of interest.

Additional information

Funding

This research received no specific grant from any funding agency, commercial or not-for-profit sectors.