Bone mineral content for preterm neonates treated with caffeine using dual energy X-ray absorptiometry: An observational study

1 Introduction

Metabolic bone disease (MBD) also known as osteopenia of prematurity is a common problem encountered in premature infants [1]. Reduced bone mineral density compared to intrauterine bone density at a comparable gestational age defines this condition [2]. MBD is a complex illness that involves environmental, biomechanical, and nutritional variables. Corticosteroids, methyl xanthines and diuretics are among the culprit drugs in the development of MBD in preterm infants [3].

Studies on caffeine citrate showed it to be efficacious as the first line of pharmacotherapy for apnea of prematurity in decreasing the incidence of apneic episodes and the requirement for mechanical ventilation [4,5, 4,5]. Since more than 30 years, caffeine has been utilized as respiratory stimulants for preterm neonates, and as of 2005, it was one of the top 10 drugs prescribed in neonatal critical care units (NICU) [6]. When used to treat apnea of prematurity, caffeine lowers the likelihood of developing bronchopulmonary dysplasia, severe retinopathy of prematurity, cognitive impairments at 18 months, and may even improve motor function at 5 years as well as decreasing the risk of motor impairment at 11 years [7]. Additionally, caffeine can have renal effects such as diuresis, natriuresis, and an increase in the excretion of calcium (Ca) in the urine [8,9, 8,9]. This renal calcium loss aggravates the development of MBD [2].

Predicting the fate of bone mineralization in premature newborns using routine serum alkaline phosphatase (ALP) and phosphate (P) assays is challenging [10]. Radiological studies are insufficiently sensitive to diagnose MBD, revealing only a decline in bone mineralization of > 20–40% [11,12, 11,12]. Contrarily, Dual Energy X-ray Absorptiometry (DEXA) is becoming into one of the standard methods for assessing preterm infants’ mineral status, whole-body composition, and the impact of dietary changes on their weight gain and mineral accretion [13]. DEXA was regarded by Dermerth and Fields as one of the most dependable ways to assess neonatal body composition to this date [14].

The primary objective of this study was to use a DEXA scan to see if caffeine administration in preterm newborns was linked to lower bone mineral content (BMC). Secondary goals included evaluating whether caffeine prescription was linked to a higher risk of nephrocalcinosis or bone fractures.

2 Patients and methods

This prospective observational study enrolled a total of 42 preterm infants admitted in the NICU, Ain Shams University Hospital. The Ethical Committee of Pediatrics department, Faculty of Medicine, Ain Shams University approved the study and an informed consent was obtained from each infant’s legal guardian before their enrollment.

The study comprised preterm infants with gestational ages of 34 weeks or less who had been hospitalized continuously in our NICU. Gestational age was determined by the date of last menstrual cycle and confirmed by antenatal trans-abdominal ultrasonography and modified Ballard scoring system [15]. Exclusion criteria for the trial included preterm infants with congenital defects, suspected chromosomal aberrations, renal or endocrine disorders, diuretic or steroid treatment, or who could not be transported to the DEXA room.

The enrolled preterms were split into 2 groups, the caffeine group: 22 preterm infants received caffeine (from the newborn’s second day and for a period of more than seven days) as per Ain Shams University NICU’s protocol for the prevention or treatment of apnea [apnea prophylaxis for those≤32 weeks’ gestation while apnea treatment is for those 33 or 34 weeks’ gestation], and the control group: 20 premature babies who were not administered caffeine because their parents either objected to its prophylactic usage or it was neither recommended or available. According to our NICU protocol, when the apnea was repeated (more than three times per hour) or life-threatening, mandating frequent bag and mask ventilation.

Preterm neonates in the caffeine group were prescribed a loading dose of caffeine of 20 mg/kg (10 mg/kg caffeine base), then a maintenance dosage of 5 to 10 mg/kg/day (2.5–5 mg/kg/day caffeine base) [16]. According to protocol of Ain Shams University NICU, until 34 weeks’ corrected gestational age and at least seven days without an apnea episode, caffeine medication was continued. Caffeinospire^®, was the utilized caffeine formulation, containing caffeine citrate. (60 mg/3 ml vial), manufactured by Memphis for Pharmaceuticals & Chemical Industries, Egypt; for: Inspire Pharmaceutical Co. (IPC Pharma), USA.

Following the NICU practice at Ain Shams University, all elligible infants were prescribed P and Ca supplements: Ca and P were added during parenteral nutrition, with Ca in the form of Ca gluconate. (Calcionate^@, Memphis company for pharmaceutical and chemical industries-Cairo-Egypt) in the a dosage of 5cc/kg/d (45 mg of elemental Ca /kg/d) and 1.5cc/kg/d of P in the form of Na-Phosphate (Glycophos@, a product of the Fresenius Kabi company) (1.5 mmol P + 3 mmol Na/kg/d). Preterm formula was employed for nutritional care and for the elimination of confounders.

All infants in the caffeine and control groups underwent clinical follow-up that included at least one full physical examination per day. When necessary, plain radiography was employed to confirm a potential fracture.

Laboratory investigations: serum Ca, P, ALP, Mg, Na, K, and creatinine were ordered prior to the initiation of caffeine administration, when necessary as ordered by the neonatologist in charge, and between the fifth and sixth weeks after birth.

DEXA scan and abdominal U/S were performed between the fifth and sixth week post-natal age.

Abdominal ultrasonography U/S (for screening of nephrocalcinosis) was performed with A LOGIA 400 proseries (general electric) Duplex –Pulsed Doppler ultrasound machine with a 7.5 m HZ linear probe and high pass filter at 50 HZ.

In order to assess the total body BMC, a DEXA scan was performed using lunar DPX MD, Type: Whole Body Scanner, Pencil beam, regions: AP Spine, Lateral, Hip, Forearm/ Hand, Ortho Femur Speed: Up To 76 mm Per Sec., Scan Times (min): AP Spine-2, Lateral-12, Hip-4, Forearm-1, Femur-4 Pentium II Computer System, Windows O/S, Lunar Software 15” Hi-Resolution SVGA Monitor, HP Desk Jet Inkjet Printer Auto Centering Laser Guide, Data Analysis: Auto Analysis Software, Smart Scan, Auto Position. Weight and length were assessed and documented. During the assessment of the BMC, to ensure that the position was same for all subjects, the preterm was positioned on the scanning board with the head at the designated start line. The investigation was conducted while the infants slept uninterrupted. A cotton blanket was utilized to supinely position and restrain the infants. Infants were fed a few minutes before the trial to help with sleep induction. When the newborn moved during the measurement or the image quality was inadequate this measurement was excluded from the study.

3 Results

Twenty two preterm neonates received caffeine during the study’s 42 preterm newborns’ participation (10 (45%) males and 12 (55%) females). They were comparable in demographic and initial laboratory data (Table 1).

Table 2 demonstrates increase in the serum ALP in caffeine group compared to control group at 5th–6th week post-natal age. Serum Ca, P, Na, K, Mg and creatinine showed non-significant difference between the two groups before discharge. BMC showed statically significant lower levels in the caffeine compared to control group.

Initial laboratory data were comparable to those at 5th–6th week post-natal age except for higher levels of serum ALP at 5th–6th week post-natal age in both caffeine and control groups as shown in Table 3.

BMC showed positive correlation to birth weight in caffeine (r = 0.913, p = 0.000) and control groups (r = 0.969, p = 0.000) and significant positive correlation to gestational age in caffeine (r = 0.966, p = 0.000) and control groups (r = 0.950, p = 0.000). BMC showed a negative correlation to serum ALP (r = –0.498, p = 0.018) and a positive correlation to serum P (r = 0.522, p = 0.002) in the caffeine group.

The duration for caffeine therapy ranged from 8 to 36 days with a mean of 15.74±7.3 days. Caffeine therapy duration was negatively correlated to BMC (r = –0.370, p = 0.000) and positively correlated to serum ALP levels (r = 0.667, p = 0.001) (data not shown).

Multivariate analysis showed higher ALP levels in preterm infants who received caffeine therapy for more than 14 days compared to control group. Moreover, BMC levels were lower in preterm infants who received caffeine therapy for more than 14 days compared to control group as shown in Table 4.

None of our neonates had nephrocalcinosis. One neonate in the caffeine group (5%) had fracture femur. The fracture was clinically suspected and confirmed by X-ray at day 22 post-natal age. Serum Ca, P, Na, k, Mg and creatinine levels were in normal range, serum ALP was 785 IU/L and BMC was 17 g. Serial follow up of the baby were done by NICU and orthopedic staff. At the age of 3 months the fracture healed. Serum Ca, P, Na, k, Mg and creatinine levels were in normal range, serum ALP decreased to 478 mg/dl and BMC increased to 41 g.

In the present study BMC measurements using DEXA scan were considerably lower in the caffeine. To our knowledge, neonates have not yet been found to be impacted by caffeine’s effects on BMC. Caffeine consumption has been linked to fractures and reduced bone mineral density in adolescent girls and post-menopausal women which is believed to be secondary to its impact on renal and intestinal excretion [17]. Caffeine may potentially function through prostaglandin production, but its antagonistic effects on adenosine signalling in the kidney are thought to account for the majority of its renal effects [18]. Osteopenia and nephrocalcinosis are believed to develop as a result of the cumulative diuretic-induced Ca loss [19].

Our results showed significantly higher levels of serum ALP before discharge compared to initial levels in both caffeine and control groups. Moreover, serum ALP levels before discharge were considerably more elevated in the caffeine group. While serum levels of Ca, P, Na, K, and creatinine were comparable in both groups.

In the past, it has been argued that diagnosing preterm children with metabolic bone disease may be done simply by using straightforward biochemical markers of bone mineralization, such as serum ALP and, to a lesser extent, serum P and serum Ca [20, 21]. Later, Catche and Leon came to the conclusion that serum Ca and P levels are not reliable indicators of mineral shortage in the early stages [22]. Despite the fact that P content in the serum and bone mineral density are connected, this measurement is not sensitive enough to detect infants who are deficient in bone mineral. The sensitivity of the screening and identification of newborns at risk for MBD can be considerably increased by using blood P levels in conjunction with serum ALP levels [23].

After 5 days of caffeine administration, Zanardo et al. observed that preterm newborns’ renal excretion of calcium increased [9]. Since the body maintains normal serum levels by inducing a compensatory hyper-parathyroid condition that results in bone mineral reabsorption and urine phosphate squandering, serum calcium levels often remain normal well into the diseased process. More frequently, chronically low or declining serum phosphate indicates insufficient phosphate nutrition, which prevents PTH from rising and prevents urinary P loss. However, active vitamin D becomes more prevalent in response to hypophosphatemia, which may even result in hypercalcemia and hypercalciuria [24].

In the current study, neonates who took caffeine for more than 14 days had lower BMC and greater ALP than the control group. The length of caffeine therapy also had a negative correlation with BMC and a good correlation with ALP. Miller et al., however, came to the conclusion that prolonged caffeine exposure did not appear to raise the incidence of osteopenia in newborns [25]. They selected ALP as a marker for osteopenia because their investigation was retrospective in nature. Backstrom et al. employed ALP as a biochemical marker for the emergence of osteopenia, with which there has been documented to be a direct association [26]. The value of ALP alone as a biochemical marker of osteopenia has been questioned, though [10]. In the caffeine group, BMC was further inversely linked with serum ALP levels and favourably correlated with serum P levels. In a study including 71 newborns, Ryan et al. compared BMC, serum ALP, and phosphate measurements taken on the same day on 134 times. Although they discovered a relationship between serum ALP, P, and BMC, the biochemical variables could only account for 7% of the variation in BMC, making the prediction of BMC unsatisfactory [27]. According to research conducted on 108 preterm newborns by Awad et al., there is no correlation between whole-body BMC at term and serum ALP or phosphate [28]. Therefore, regular serum ALP and phosphate assays are useless for forecasting bone mineralization in premature newborns [10].

According to the CAP study, there was no discrepancy between the caffeine and control groups in terms of weight at the fifth and sixth week post-natal age [29]. Bauer et al. found that long-term caffeine treatment to premature newborns increases oxygen demand and decreases weight gain [30]. Because methylxanthines have thermogenic effects, they have been used to increase energy expenditure in adults and experimental animals in order to promote and maintain weight loss [31]. Before considering high-calorie supplemental feeding for premature newborns receiving caffeine therapy, these findings must be validated in a broader neonatal population.

While none of our newborns had acquired nephrocalcinosis, one of them who had caffeine for more than 14 days got fractured, a condition that cannot be attributed to caffeine. Neonates under 1500 g have a 30% fracture incidence, according to Koo et al. [32]. The frequency of fractures associated with caffeine medication is not frequently documented in the literature. High caffeine doses have reportedly been linked to a decrease in osteoblast cell viability via promoting intracellular oxidative processes [33].

The limitations of this study were using non-randomized sample because we considered it unethical not to give caffeine for preterm infants, difficult transfer of neonates from NICU to the DEXA room in the radiology unit, measuring serum level of caffeine was not available, and a small sample size of a single institute.

5 Conclusions

Caffeine administration for more than 14 days in preterm neonates may be associated with lower BMC but not nephrocalcinosis or bone fracture. We propose that intensive MBD prevention or treatment can be started earlier in the management process in preterm neonates undergoing caffeine medication. It is suggested that further extensive research be done on this result using larger sample sizes.