Sex differences in body composition and bone mineral density in phenylketonuria: A cross-sectional study

Background Low bone mineral density (BMD) and subsequent skeletal fragility have emerged as a long-term complication of phenylketonuria (PKU). Objective To determine if there are differences in BMD and body composition between male and female participants with PKU. Methods From our randomized, crossover trial [1] of participants with early-treated PKU who consumed a low-phenylalanine (Phe) diet combined with amino acid medical foods (AA-MF) or glycomacropeptide medical foods (GMP-MF), a subset of 15 participants (6 males, 9 females, aged 15–50 y, 8 classical and 7 variant PKU) completed one dual energy X-ray absorptiometry (DXA) scan and 3-day food records after each dietary treatment. Participants reported lifelong compliance with AA-MF. In a crossover design, 8 participants (4 males, 4 females, aged 16–35 y) provided a 24-h urine collection after consuming AA-MF or GMP-MF for 1–3 weeks each. Results Male participants had significantly lower mean total body BMD Z-scores (means ± SE, males = − 0.9 ± 0.4; females, 0.2 ± 0.3; p = 0.01) and tended to have lower mean L1–4 spine and total femur BMD Z-scores compared to female participants. Only 50% percent of male participants had total body BMD Z-scores above − 1.0 compared to 100% of females (p = 0.06). Total femur Z-scores were negatively correlated with intake of AA-MF (r = − 0.58; p = 0.048). Males tended to consume more grams of protein equivalents per day from AA-MF (means ± SE, males: 67 ± 6 g, females: 52 ± 4 g; p = 0.057). Males and females demonstrated similar urinary excretion of renal net acid, magnesium and sulfate; males showed a trend for higher urinary calcium excretion compared to females (means ± SE, males: 339 ± 75 mg/d, females: 228 ± 69 mg/d; p = 0.13). Females had a greater percentage of total fat mass compared to males (means ± SE, males: 24.5 ± 4.8%, females: 36.5 ± 2.5%; p = 0.047). Mean appendicular lean mass index was similar between males and females. Male participants had low-normal lean mass based on the appendicular lean mass index. Conclusions Males with PKU have lower BMD compared with females with PKU that may be related to higher intake of AA-MF and greater calcium excretion. The trial was registered at www.clinicaltrials.gov as NCT01428258.


Introduction
PKU (PKU; OMIM 261600) is an autosomal recessive genetic disease that results in a deficiency of phenylalanine hydroxylase (PAH; EC 1.14.16) to hydroxylate Phe to tyrosine (Tyr), using tetrahydrobiopterin as a cofactor [2]. Early identification of PKU with newborn screening and initiation of a low-Phe diet within the first weeks of life are essential to prevent severe cognitive impairment caused by the neurotoxicity of high Phe concentrations in the brain [3,4]. Primary treatment for PKU involves lifelong adherence to a low-Phe diet, restricted in protein from natural foods, in combination with low-Phe amino acid medical foods (AA-MF) or glycomacropeptide medical foods (GMP-MF) to meet daily protein and micronutrient needs [4,5].
We recently conducted a randomized, controlled crossover trial to investigate the safety and efficacy of Phe-free AA-MF and low-Phe GMP-MF in 30 participants with early treated PKU and concluded that GMP-MF did not significantly increase plasma Phe concentrations [1]. From this clinical trial, two sub-studies were conducted: 1) A cross-sectional study in which one dual energy X-ray absorptiometry (DXA) scan was obtained from 15 participants with PKU; 2) A crossover pilot study in which 24-h urine collections and food records were obtained from 8 participants with PKU consuming AA-MF and GMP-MF to determine the impact of dietary acid load on excretion of renal net acid and minerals [9]. In our recent crossover pilot study, we demonstrated that ingestion of high-acid AA-MF significantly increased urinary excretion of renal net acid, calcium, and magnesium and concluded that this may negatively affect bone health in PKU [9]. Unexpectedly, we identified that 2 of 8 participants (both males) had low BMD-for-age based on DXA. We hypothesize that higher intake of AA-MF needed to support an intense pubertal growth spurt, may increase urinary calcium excretion and reduce bone accretion in males with PKU. Our objective was to investigate whether there are differences in BMD and body composition between males and females with PKU.

Study design and protocol
As stated in the introduction, this manuscript presents data from two sub-studies of our previously reported randomized, controlled, crossover trial [1]. First, utilizing a cross-sectional study design, we obtained one whole-body DXA scan from 15 of 30 participants who completed our randomized, controlled, crossover trial where participants consumed their typical low-Phe diet in combination with average intake of 0.74-0.76 g protein equivalents/kg/day from AA-MF or Gly-tactin™ GMP-MF for 3-wk each [1]. Intakes of medical foods composed of primarily elemental amino acids are described as protein equivalents. Participants reported lifelong intake of AA-MF prior to the trial. Thus, DXA scans reflect intake with AA-MF.
Second, a crossover pilot study was conducted in 8 of 30 particpants with early treated PKU [9]. Participants consumed a low-Phe diet in combination with a Glytactin™ GMP-MF with a low potential renal acid load and an AA-MF with a high potential renal acid load for 1-3 weeks each. Participants provided one 24-h urine collection for each dietary treatment and two-three 24-h food records before and during the 24-h urine collection. Food records for all studies were analyzed using Food Processor SQL (version 10.12.0, ESHA) [1,9].
Briefly, for the crossover pilot study, the nutrient intakes of the low-Phe diet with AA-MF and GMP-MF treatments were generally similar except for the dietary protein source of medical foods, such that amino acids were consumed with AA-MF and primarily intact protein with GMP-MF. Mean intakes of total energy (2266-2566 kcal/d), total protein (79-81 g/d), and protein equivalents from medical foods (55-57 g protein equivalents/d) were similar between dietary treatments [9]. Despite similar intakes of total calcium (1745-1898 mg/d) and magnesium (568-684 mg/d), participants excreted more urinary calcium and magnesium with AA-MF than GMP-MF [9]. Consistent with the greater potential renal acid load (mEq/d) with AA-MF compared to Glytactin™ GMP-MF (means ± SE, AA-MF, 39 ± 5; GMP-MF, − 43 ± 6; p < 0.0001), excretion of renal net acid was 3-fold high with ingestion of AA-MF compared with GMP-MF [9].
Inclusion criteria included PKU diagnosis that was early-treated with medical food, a current prescribed diet providing > 50% of daily protein needs from AA-MF, and enrollment or completion of our clinical trial at the Waisman Center site [1]. Classical and variant PKU were defined based on genotype and response to sapropterin dihydrochloride [1]. The University of Wisconsin-Madison Health Sciences review board approved the study protocol. All participants provided written informed consent. The trial was registered at www.clinicaltrials.gov as NCT01428258.

Statistical analysis
All statistical analyses were performed using SAS version 9.4 and assumptions of normality and equal variance were tested. Most analyses used PROC MIXED (SAS Institute Inc.). Participant characteristics and DXA scan data were analyzed using ANOVA with effects for sex and genotype (classical or variant PKU). The Kruskal-Wallis test was used to test for differences due to diet or genotype, if data was skewed. Statistical significance was set at p < 0.05.

Participants
Fifteen (6 males, 9 females) participated in this sub-study, including 12 adults (aged 19-50 y) and 3 adolescents (aged 15-17 y). Participant characteristics are summarized in Table 1. Of the 8 participants categorized with classical PKU, all were adults (4 males, 4 females). Of the 7 participants categorized with variant PKU, 4 were adults (1 male, 3 females) and 3 were adolescents (1 male, 2 females). Although we tended to have more females than males with variant PKU, BMD or BMD Z-scores were similar between participants with classical and variant PKU (Supplemental Table 1). Two participants (both female adolescents) used a consistent dose of sapropterin dihydrochloride throughout the study.

Bone mineral density
Bone mineral density parameters are summarized in Table 2. Male participants had significantly lower mean total body BMD Z-scores (means ± SE: − 0.9 ± 0.4 vs. 0.2 ± 0.3, p = 0.01) and tended to have lower mean L1-4 spine and total femur BMD Z-scores compared to female participants (Fig. 1). Furthermore, only 50% of male participants had total body BMD Z-scores above − 1.0, compared to 100% of female participants (p = 0.06). Two of 15 participants were diagnosed with low BMD-for-age, i.e. a Z-score ≤ −2.0 (1 adult male with classical PKU and 1 adult male with variant PKU). Of note, four additional participants (2 males, 2 females) had at least 1 BMD Z-score that was between − 1.0 and − 2.0 and one female participant had four BMD Zscores between − 1.0 and −2.0. There were no male versus female differences in BMD (g/cm 2 ) for total body, L1-4 spine, total femur, femoral neck and femoral trochanter. BMD (g/cm 2 ) and BMD Z-scores were similar in classical and variant PKU (Supplemental Table 1). Because of our small sample size and lack of age-and sex-matched controls, BMD Z-scores were used to compare differences due to sex and PKU genotype.
Participants with classical and variant PKU reported lifelong compliance with AA-MF and similar protein intake from medical foods during the course of our previously reported study (means ± SE, g protein equivalents from medical foods, classical PKU = 56 ± 3; variant PKU = 46 ± 3; p = 0.13) [9]. Male participants tended to have higher intakes of protein equivalents from AA-MF and greater urinary calcium excretion compared to female participants, in spite of similar intake of protein equivalents from AA-MF when adjusted for body weight or total lean mass.
We utilized correlation coefficients to test the hypothesis that intake of AA-MF may negatively impact bone. Total femur Z-scores were negatively correlated with intake of AA-MF (r = − 0.58; p = 0.048), but not spine L1-4 Z-scores (r = − 0.17; p = 0.55) or total body Z-scores (r = −0.42; p = 0.12) (Fig. 2). Due to the cross-sectional design and small sample size of this study, etiology for the significant correlation between intake of AA-MF and total femur Z-scores, but not spine L1-4 or total body Z-scores, is not clear. Regardless, this finding supports evidence from our previous study that the high dietary acid load of AA-MF increases urinary excretion of renal net acid, calcium, and magnesium and may increase bone resorption [9]. However, in the current study, we did not find significant differences between males and females in potential renal acid load from AA-MF nor urinary excretion of renal acid, magnesium, and sulfate (Table 3). Nonetheless, males tended to excrete more calcium than females, which could potentially have a negative impact on bone mineral density over time.

Body composition
Body composition assessments are summarized in Table 2. Consistent with the general population, female participants had significantly greater total fat mass compared to male participants (means ± SE, females, 36.5 ± 2.5%; males, 24.5 ± 4.8, p = 0.047). Three of 6 male participants and 4 of 9 female participants had excess fat mass, defined as fat mass ≥ 25% in males and ≥ 35% in females per the World Health Organization [31]. As expected, males had   [30][31][32][33][34][35] significantly more total lean mass and ALM. Unlike the general population where men have greater ALMI (also known as ALM/ht 2 ), there was no significant difference in the ALMI and also no difference in ALMI Z-scores between male and female participants. Notably, the mean ALMI for our young male participants was close to the cut point of ≤ 7.26 kg/m 2 , a suggested cut-off to identify sarcopenia in older men [28,32]. In summary, male participants had low-normal lean mass and female participants had significantly greater fat mass.

Discussion
The prevalence of low BMD-for-age in PKU, defined as a Zscore ≤ −2.0, is high, reportedly affecting approximately 20% of individuals with PKU in comparison to 2% of a general population [7,33,34]. We have previously reported that AA-MF provide a high dietary acid load that increases urinary excretion of renal net acid and bone-related minerals (calcium and magnesium), which may contribute to skeletal fragility in PKU [9]. To build on this, in this study, we observed significantly lower BMD Z-scores for total body and trends for lower BMD Z-scores for L1-4 spine and total femur in male participants compared to females. Although we found no significant differences between males and females in the urinary excretion of renal net acid, magnesium or sulfate, males tended to excrete more calcium compared to females. This finding is important given that we have previously shown that urinary calcium excretion is negatively correlated with L1-4 spine BMD in 8 participants with lifelong compliance with high-acid AA-MF [9].
Individuals with PKU and others who consume the majority of their dietary protein from amino acids instead of intact dietary proteins are estimated to have 20-40% higher protein requirements [4,5], due to evidence of lower protein retention and protein synthesis with amino acids compared with intact protein [35][36][37]. Adequate intake of dietary protein or protein equivalents is important to optimize growth, especially given evidence of low-normal lean mass and excess fat mass observed in our adolescent and adult PKU participants. The observation that our young male participants had a mean ALMI that was similar to our female participants and was close to a suggested cut-off point to identify sarcopenia in older men was surprising. Although the etiology of this observation is unclear, possible explanations may include increased protein degradation compared to protein synthesis related to lifestyle and differences in protein metabolism associated with the PKU   [30][31][32][33][34][35] genotype or hormones [35][36][37]. Life style factors that might impact muscle may include suboptimal physical activity, limited intake of intact protein from natural foods, and consumption of elemental amino acids as opposed to intact protein from medical foods [32,[35][36][37]. While it is tempting to recommend higher intake of protein equivalents from AA-MF for individuals with PKU, our data suggest negative effects on bone health when AA-MF with a high potential renal acid load provides the primary source of protein [9]. We observed a negative correlation between protein intake from AA-MF and total femur Z-scores similar to that reported in a larger data set [10]. Moreover, our male participants showed lower BMD Z-scores, tended to consume more protein equivalents from AA-MF, and excreted more calcium compared to female participants. Given the limited understanding of bone health in PKU, a conservative approach is to encourage adequate intake of AA-MF or GMP-MF that have a low potential renal acid load in combination with weight-bearing exercise to support bone remodeling [6,9,38]. Acquisition of lean mass to promote bone strength is important given recent evidence of reduced bone strength in relation to muscle force in patients with PKU assessed with peripheral quantitative computed tomography [6].
In conclusion, our data demonstrate that males with PKU have lower total body BMD Z-scores and may be at greater risk for osteoporosis than females with PKU. We hypothesize that the lower total body BMD Z-scores found in our male participants may be related to low-normal lean mass and/or higher intakes of AA-MF with a correspondingly greater loss of urinary calcium. The study supports existing Genetic Metabolic Dietitian International Guidelines that recommend periodic measurement of BMD in individuals with PKU [4,5]. Although limited by a small sample size (as is often the case in research studies in patients with rare diseases) and use of whole body DXA scans, our study suggests that AA-MF had a negative effect on BMD. Furthermore, our findings highlight the need to separately report indicators of bone health for males and females with PKU. Additional research, particularly interventional studies, to investigate the impact of AA-MF in the acquisition of bone mass in individuals with PKU are needed, especially in males.

Ethical approval
The University of Wisconsin-Madison Health Sciences review board approved the study protocol. All subjects provided written informed consent. The trial was registerd at www.clinicaltrials.gov as NCT01428258.

Funding
This work was supported by Department of Health and Human Services grants R01 FD003711 from the FDA Office of Orphan Product Development to Ney, P30-HD-03352, and by the Clinical and Translational Science Award (CTSA) program, through the NIH National Center for Advancing Translational Sciences (NCATS), grant UL1TR000427. Cambrooke Therapeutics, Inc. donated the GMP medical foods used in this study, but was not involved in the design or conduct of the study or in the collection, analysis or interpretation of the data.