Abstract
Purpose.
Perfluoroalkyl substances (PFAS) are used to make protective coatings on common household products. Prenatal exposures have been associated with developmental outcomes in offspring. Using data from the Avon Longitudinal Study of Parents and Children (ALSPAC), we investigated the association between prenatal concentrations of PFAS and bone health in girls at 17 years of age and whether body composition can explain any associations.
Methods.
We measured concentrations of perfluorooctane sulfonate (PFOS), perfluorooctanoate (PFOA), perfluorohexane sulfonate (PFHxS), and perfluorononanoate (PFNA) in maternal serum samples collected during pregnancy. We obtained bone health outcomes in the girls, such as bone mineral density, bone mineral content, bone area, and area adjusted bone mineral content from whole body dual-energy x-ray absorptiometry (DXA) scans. We used multivariable linear regression to explore associations between each PFAS and each bone health outcome with adjustment for important confounders such as girl’s age at clinic visit, maternal education, and gestational age at sample collection. We also controlled for girl’s height and lean mass to explore the role body composition had on observed associations.
Results.
Prenatal PFOS, PFOA, PFHxS and PFNA concentrations were associated with inverse effects on bone size and mass after adjusting for important confounders. Conversely, PFNA was positively associated with area-adjusted bone mineral content. However, most significant associations attenuated after additional controlling for height and lean mass.
Conclusions.
Prenatal concentrations of some PFAS may be associated with reduced bone mass and size in adolescent girls, although it is not clear whether these associations are driven by body size.
Keywords: PFAS, ALSPAC, bone development, body composition, adolescent
Mini-abstract
Prenatal exposures to perfluoroalkyl substances (PFAS) have been associated with developmental outcomes in offspring. We found that prenatal concentrations of some PFAS may be associated with reduced bone mass and size in 17-year-old British girls, although it is not clear whether these associations are driven by body size.
Introduction
Perfluoroalkyl substances (PFAS) are endocrine disrupting chemicals (EDCs) that can interrupt signaling pathways during fetal development through alteration of hormonal functions [1]. The chemical properties of PFAS can help create protective coatings on many household consumer products, such as food packaging, nonstick cookware, and textiles. PFAS are also used in the production of industrial products. While Europe and the U.S. have phased out production of some PFAS due to potential adverse health effects associated with exposure, they remain a public health concern because of their long half-lives and persistent ability to accumulate in human tissues [2–5]. Common PFAS include perfluorooctane sulfonate (PFOS), perfluorooctanoate (PFOA), perfluorohexane sulfonate (PFHxS), and perfluorononanoic acid (PFNA). During critical periods of development, the developing fetus is particularly vulnerable to chemicals like EDCs that alter hormonal pathways [6]. PFAS are found in circulating blood and cord blood and are transferred to a fetus through the placenta during pregnancy. [6–8]. Toxicological studies indicate that certain PFAS accumulate in bone tissue in the fetus and could cause altered bone development [9].
Several epidemiological studies have suggested a possible inverse relationship between various EDC exposures and bone health [10–13]. Specifically, two cross sectional studies using samples from the U.S. National Health and Nutrition Examination Survey (NHANES) have found lower bone mineral density with increased exposure to certain PFAS [13,12]. However, no studies to date have examined the relationship between in utero exposure to PFAS and bone health in offspring in adolescence.
Our primary objective was to examine associations between maternal exposure to PFAS during pregnancy and their daughters’ bone health at 17 years of age using data from mothers and daughters in the Avon Longitudinal Study of Parents and Children (ALSPAC). Given the close dependency of bone mass acquisition on growth and body composition [14], we were also keen to establish whether any observed associations between prenatal PFAS and measures of bone health could be explained by height, fat or lean mass.
Methods
Study Design and Population
ALSPAC is a birth cohort that recruited pregnant women with expected delivery dates between April 1991 and December 1992. The study enrolled 14,541 pregnant women and their children at birth from three health districts of the former Avon region in South West England. Previous papers have described recruitment methods [15,16]. Enrolled mothers and children provided biological samples, completed questionnaires and participated in clinical assessments. The study website contains additional details for all available data through a fully searchable data dictionary (http://www.bris.ac.uk/alspac/researchers/data-access/data-dictionary/).
The current analysis uses data from an ancillary nested case-control study originally designed to study the association between prenatal EDC exposure and timing of menarche [17]. We selected data for our study from the original sample where we defined case subjects as girls who had early menarche (before 11.5 years of age) and control subjects were a random sample of remaining girls. Among the 448 mother-daughter dyads from the original nested case-control study, 257 have measured maternal gestational blood concentrations for all studied PFAS and a complete whole body dual energy X-ray absorptiometry (DXA) scan done during the daughters’17 year-old clinic visit. We obtained approval for the study from the ALSPAC Ethics and Law Committee, the Local Research Ethics Committees, and the Centers for Disease Control and Prevention (CDC) Institutional Review Board. Mothers provided written informed consent for participation in the study.
Laboratory Analysis
Researchers at ALSPAC transferred blood samples under controlled conditions to the National Center for Environmental Health laboratories of the CDC for analysis. The lab measured serum levels of PFOS, PFOA, PFHxS, and PFNA in maternal blood collected at a median gestational age of 15 weeks. Laboratory methods have been described previously [18]. Limits of detection for the assays were 0.2 ng/mL for PFOS, 0.1 ng/mL for PFOA, 0.1 ng/mL for PFHxS, and 0.08 ng/mL for PFNA. Quality control measures were implemented using standards, reagent blanks, and study samples. Precision of measurements for the analytes, as relative standard deviation, ranged from 8 to 13%.
Data collection
All of the daughters included in our study (n=257) completed a DXA scan at 17 years of age. A Lunar prodigy narrow fan beam densitometer (GE Medical Systems Lunar, Madison, WI, USA) was used to perform a whole body DXA scan, from which total body less head bone mineral content (BMC), bone mineral density (BMD), bone area (BA), fat mass, and lean mass were obtained, as was area-adjusted BMC (ABMC). A DXA scan of the left hip, from which total and femoral neck BMD were obtained, was performed at the same time. Height was measured using a Harpenden Stadiometer (Holtain, Crymych, Pembs, UK). Some daughters (n=230) also completed tibial peripheral quantitative computerized topography (pQCT) using the Statex XCT2000L (Stratec, Pforzheim, Germany) at 17 years of age. pQCT measures the parameters of the tibia in detail and provides a more comprehensive assessment of bone than DXA scans during developmental ages [19]. Cortical BMC (BMCc), cortical BMD (BMDc), and cortical BA (BAc) were measured from the pQCT scan with XCT custom software version 6.00B. A circular ring model was used to derive periosteal circumference (PC) and cortical thickness (CT). A threshold above 650 mg/cm3 was used to define cortical bone [20].
Plasma was separated from sample and frozen at −80 degrees Celsius within 4 hours from collection. Plasma concentrations of β-C-telopeptides of type I collagen (CTX), a marker for bone turnover, were measured using electrochemiluminescence immunoassays (Roche Diagnostics, Lewes, UK) in fasting samples collected around 15 years of age (n=120).
ALSPAC researchers collected covariate data on mothers and daughters through medical records and questionnaires. We characterized low birth weight as less than 2,500 g at delivery. Researchers collected information on maternal pre-pregnancy body mass index (BMI), education, ethnicity, age at delivery, and smoking status during pregnancy. We collected data on whether the child was ever breastfed by questionnaire at 15 months of age. Age at menarche was determined from questionnaires obtained throughout childhood and early menarche was defined as younger than 11.5 years of age [21].
Statistical Analysis
We measured associations between maternal PFAS concentrations and bone health outcomes using weighted linear regression models, with weights inversely proportional to selection probabilities. To account for the original nested case-control study design, we weighted the sample to adjust for under-representation of the true number of girls without early menarche (weight for cases was 1 and for controls was 15.1). All prenatal PFAS concentration variables and bone health variables were continuous. Potential confounding variables were selected a priori for consideration in model selection and backwards elimination was used to determine which confounders were important contributors (Model 1) [22]. We evaluated associations by examining beta coefficients of the exposure, representing the estimated change in bone measure with a one unit (ng/mL) higher PFAS concentration, and their 95% confidence intervals. A negative beta coefficient represents a lower bone measure associated with a one unit (ng/mL) higher PFAS concentration. We then first evaluated the associations between PFAS and bone measures in models that adjusted for the following confounders: maternal education, gestational age at maternal serum sample collection, and daughter’s age at clinic visit. Breastfeeding status and maternal smoking status during pregnancy were not found to be significant confounders. To explore the possible role of growth and body composition in the observed associations, we then examined other models adjusted for height and body composition, guided by results of univariate regression analysis. We used complete cases for all analyses. SAS version 9.2 (SAS Institute Inc., Cary, NC) was used to conduct all analyses.
Results
Table 1 describes demographic characteristics of our sample of 257 mother daughter dyads. Overall, mean (SD) maternal PFOS, PFOA, PFHxS, and PFNA concentrations were 21.92 (10.72) ng/mL, 4.08 (1.83) ng/mL, 2.59 (5.36) ng/mL, and 0.56 (0.22) ng/mL, respectively. Table 2 presents means of the bone health outcomes examined in the study.
Table 1.
Characteristics of study population (n=257) for mothers and daughters enrolled in ALSPAC with maternal serum concentrations and DXA measures at 17 years of age.
| Frequency [n (%)] | PFOS (ng/mL) [Median (IQRa)] | PFOA (ng/mL) [Median (IQRa) | PFHxS (ng/mL) [Median (IQRa) | PFNA (ng/mL) [Median (IQRa))] | |
|---|---|---|---|---|---|
| Overall | 257 (100) | 20.2 (15.6, 25.5) | 3.8 (2.9, 4.9) | 1.7 (1.3, 2.3) | 0.6 (0.4, 0.7) |
| Maternal Prepregnancy BMI | |||||
| Underweight (<18.5 kg/m2) | 13 (5.1) | 17.0 (14.0, 22.6) | 3.4 (2.8, 4.7) | 1.5 (1.3, 2.0) | 0.5 (0.3, 0.6) |
| Normal (18.5–24.9 kg/m2) | 169 (65.8) | 20.5 (15.5, 25.5) | 3.9 (2.9, 5.1) | 1.8 (1.2, 2.4) | 0.6 (0.4, 0.7) |
| Overweight (25.0–29.9 kg/m2) | 40 (15.6) | 21.2 (18.6, 25.9) | 3.4 (3.3, 5.1) | 1.9 (1.6, 3.1) | 0.6 (0.5, 0.7) |
| Obese (≥30.0 kg/m2) | 15 (5.8) | 17.4 (13.1, 21.2) | 3.0 (2.6, 4.3) | 1.3 (1.1, 1.7) | 0.5 (0.3, 0.7) |
| Maternal Educationb | |||||
| Less than O level | 31 (12.1) | 19.3 (17.0, 22.5) | 3.8 (2.9, 4.9) | 1.7 (1.4, 2.4) | 0.6 (0.4, 0.7) |
| O level | 90 (35.0) | 19.9 (14.9, 25.6) | 3.8 (2.9, 4.9) | 1.6 (1.2, 2.3) | 0.6 (0.4, 0.7) |
| Greater than O level | 128 (49.8) | 20.8 (15.8, 25.4) | 4.0 (2.9, 5.0) | 1.8 (1.3, 2.3) | 0.5 (0.4, 0.7) |
| Maternal Race | |||||
| White | 248 (96.5) | 20.4 (15.7, 25.5) | 3.9 (2.9, 5.0) | 1.7 (1.3, 2.3) | 0.6 (0.4, 0.7) |
| Nonwhite | 4 (1.6) | 18.4 (14.7, 22.4) | 2.9 (2.4, 3.2) | 1.6 (1.2, 1.7) | 0.7 (0.5, 0.7) |
| Maternal Age at Delivery (years) | |||||
| <25 | 45 (17.5) | 21.7 (15.6, 25.5) | 4.4 (3.3, 5.1) | 1.7 (1.3, 2.4) | 0.6 (0.4, 0.7) |
| 25–29 | 92 (35.8) | 20.1 (15.9, 25.1) | 3.8 (3.0, 4.9) | 1.6 (1.2, 2.1) | 0.6 (0.4, 0.7) |
| ≥29 | 119 (46.3) | 20.1 (15.5, 25.6) | 3.7 (2.7, 4.8) | 1.9 (1.3, 2.5) | 0.5 (0.4, 0.7) |
| Maternal Smoking Status during first 3 months of pregnancy | |||||
| Yes | 56 (21.8) | 19.0 (13.6, 23.5) | 3.4 (2.8, 4.7) | 1.7 (1.3, 2.5) | 0.5 (0.3, 0.7) |
| No | 197 (76.7) | 21.1 (16.4, 25.5) | 3.9 (2.9, 5.0) | 1.7 (1.3, 2.2) | 0.6 (0.4, 0.7) |
| Low Birth Weight (<2,500g at delivery) | |||||
| Yes | 11 (4.3) | 28.5 (18.9, 38.2) | 4.2 (3.3, 5.6) | 1.6 (1.4, 2.7) | 0.7 (0.6, 0.7) |
| No | 246 (95.7) | 20.2 (15.5, 25.1) | 3.8 (2.9, 4.9) | 1.7(1.3, 2.3) | 0.6 (0.4, 0.7) |
| Preterm Delivery (<37 weeks gestation) | |||||
| Yes | 8 (3.1) | 27.5 (23.1, 40.6) | 4.8 (4.0, 6.1) | 1.8 (1.4, 2.5) | 0.7 (0.6, 0.8) |
| No | 248 (96.5) | 20.2 (15.6, 25.2) | 3.8 (2.9, 4.9)_ | 1.7 (1.3, 2.3) | 0.6 (0.4, 0.7) |
| Ever Breastfed | |||||
| Yes | 209 (81.3) | 20.4 (15.6, 25.5) | 3.8 (2.9, 4.9) | 1.7 (1.3, 2.4) | 0.6 (0.4, 0.7) |
| No | 35 (13.6) | 18.3 (16.7, 23.4) | 3.8 (3.1, 4.9) | 1.7 (1.4, 2.3) | 0.5 (0.4, 0.6) |
| Early Menarche | |||||
| Yes | 131 (51.0) | 19.7 (16.2, 25.2) | 3.9 (2.9, 5.3) | 1.8 (1.3, 2.3) | 0.6 (0.4, 0.7) |
| No | 126 (49.0) | 21.2 (15.2, 25.5) | 3.8 (2.7, 4.7) | 1.7 (1.2, 2.3) | 0.5 (0.4, 0.7) |
Interquartile range
O-level= ordinary level; <O-level=none, Certificate of Secondary Education, or Vocational education; >O-level= Advanced level
Table 2.
Mean and standard deviation of total body less head bone variables measured by dual energy X-ray absorptiometry (DXA) scans at 17 years of age
| N | Mean | SD | |
|---|---|---|---|
| Bone Mineral Density (g/cm2) | 257 | 1.05 | 0.07 |
| Bone Mineral Content (g) | 257 | 2070.02 | 363.31 |
| Bone Area (cm2) | 257 | 1964.63 | 236.79 |
| Area Adjusted Bone Mineral Content (g) | 257 | 2253.85 | 113.74 |
After adjusting for confounders (age at clinic visit, maternal education, and gestational age at sample collection), several of the PFAS studied showed weak associations with parameters obtained from whole body DXA scans. ABMC was positively associated with each of the PFAS. A one ng/mL higher PFOS was associated with a −0.0008 g/cm2 (95% CI: −0.002, 0.000) lower BMD, −5.94 g (95% CI: −10.96, −0.92) lower BMC, −4.07 cm2 (95% CI: −7.38, −0.76) lower BA and 0.76g (95% CI: −0.78, 2.31) higher ABMC (Table 3). Similarly, a one ng/mL higher PFNA was associated with a −0.0152 g/cm2 (95% CI: −0.055, 0.025) lower BMD, −254.44 g (95% CI: −457.08, −51.80) lower BMC,−207.03 cm2 (95% CI: −339.82, −74.25) lower BA, but an 86.33 g (95% CI: 24.61, 148.05) higher ABMC. We also found weak evidence of inverse association between PFOA and bone area, but little or no association with PFHxS concentration.
Table 3.
Regression coefficients (β) for the association between maternal PFAS concentrations (ng/mL) and total body less head bone variables measured by dual energy X-ray absorptiometry (DXA) scans at 17 years of age (n=248)
| Bone Mineral Density (g/cm2) | Bone Mineral Content (g) | Bone Area (cm2) | Area Adjusted Bone Mineral Content (g) | |||||
|---|---|---|---|---|---|---|---|---|
| β | 95% CI | β | 95% CI | β | 95% CI | β | 95% CI | |
| PFOS | ||||||||
| Model 1a | −0.0008 | (−0.002, 0.000) | −5.94 | (−10.96, −0.92) | −4.07 | (−7.38, −0.76) | 0.76 | (−0.78, 2.31) |
| Model 2b | −0.0003 | (−0.001, 0.001) | −2.08 | (−6.20, 2.04) | −1.23 | (−3.75, 1.30) | −0.06 | (−1.49, 1.37) |
| Model 3c | 0.0000 | (−0.001, 0.001) | −2.02 | (−5.98, 1.94) | −1.63 | (−4.33, 1.07) | 0.66 | (−0.91, 2.23) |
| Model 4d | 0.0000 | (−0.001, 0.001) | −0.98 | (−4.69, 2.72) | −0.69 | (−3.05, 1.68) | 0.14 | (−1.26, 1.54) |
| PFOA | ||||||||
| Model 1a | −0.0020 | (−0.007, 0.003) | −21.95 | (−48.52, 4.63) | −16.98 | (−34.47, 0.52) | 6.00 | (−2.13, 14.12) |
| Model 2b | 0.0012 | (−0.004, 0.006) | 3.50 | (−18.36, 25.35) | 1.70 | (−11.70, 15.10) | 0.70 | (−6.90, 8.30) |
| Model 3c | −0.0015 | (−0.006, 0.003) | −19.55 | (−40.03, 0.93) | −15.48 | (−29.40, −1.55) | 5.92 | (−2.21, 14.05) |
| Model 4d | −0.0011 | (−0.005, 0.003) | −5.94 | (−25.68, 13.79) | −2.99 | (−15.60, 9.62) | −1.02 | (−8.48, 6.44) |
| PFHxS | ||||||||
| Model 1a | −0.0002 | (−0.002, 0.002) | −2.24 | (−11.12, 6.63) | −1.61 | (−7.46, 4.25) | 0.40 | (−2.31, 3.11) |
| Model 2b | 0.0001 | (−0.002, 0.002) | −0.08 | (−7.21, 7.05) | −0.01 | (−4.38, 4.35) | −0.06 | (−2.53, 2.42) |
| Model 3c | 0.0006 | (−0.001, 0.002) | 2.48 | (−4.40, 9.37) | 1.35 | (−3.34, 6.05) | 0.26 | (−2.47, 2.99) |
| Model 4d | 0.0006 | (−0.001, 0.002) | 2.26 | (−4.16, 8.67) | 1.15 | (−2.95, 5.24) | 0.37 | (−2.05, 2.79) |
| PFNA | ||||||||
| Model 1a | −0.0152 | (−0.055, 0.025) | −254.44 | (−457.08, −51.80) | −207.03 | (−339.82, −74.25) | 86.33 | (24.61, 148.05) |
| Model 2b | 0.0102 | (−0.028, 0.049) | −52.55 | (−220.64, 115.54) | −59.49 | (−162.33, 43.34) | 45.37 | (−12.85, 103.58) |
| Model 3c | 0.0009 | (−0.032, 0.034) | −164.58 | (−322.21, −6.96) | −151.14 | (−257.81, −44.47) | 84.19 | (22.18, 146.20) |
| Model 4d | 0.0044 | (−0.029, 0.038) | −77.08 | (−227.76, 73.60) | −71.73 | (−167.72, 24.27) | 40.98 | (−15.82, 97.78) |
Adjusted for age at clinic visit, maternal education, and gestational age at sample collection
Adjusted for age at clinic visit, maternal education, gestational age at sample collection, and height
Adjusted for age at clinic visit, maternal education, gestational age at sample collection, and lean mass
Adjusted for age at clinic visit, maternal education, gestational age at sample collection, height, and lean mass
To explore any effects of body size on these associations, we examined the association between maternal PFAS serum concentrations and fat mass, lean mass, and height of the daughter individually using linear regression to help design regression models looking at effects on bone health. We found moderate associations between certain PFAS and total lean mass and height, however weaker or no associations with fat mass (Supplemental Table 1). Based on these preliminary results, we adjusted for height and lean mass individually and together with important confounders to examine whether associations seen between PFAS and bone health were direct effects independent of body composition.
Following adjustment for height alone, associations with PFOS, PFOA and PFNA and bone parameters were all attenuated. Adjustment for lean mass alone, or height plus lean mass, led to similar attenuation of associations between PFOS and BMC and BA to that seen for height adjustment. Associations between PFOA and BA, and between PFNA and BMC, BA and ABMC, were unaffected by lean mass adjustment.
Associations with additional bone measures at 15 and 17 years of age available for a subset of the population can be found in supplemental material. Mean and standard deviation of plasma CTX levels, pQCT, and hip DXA measures can be found in Supplemental Table 2. After adjusting for important confounders, weak or no associations were found between maternal PFAS concentrations and CTX levels at 15 years of age (Supplemental Table 3), tibial pQCT outcomes at 17 years of age (Supplemental Table 4), or total hip or femoral neck BMD measured with hip DXA at 17 years of age (Supplemental Table 5).
Discussion
In our study of the potential association between prenatal PFAS concentrations and several markers of bone health in girls during adolescence, we found moderate inverse associations with bone size and mass in confounder-adjusted analyses. These associations attenuated after additional adjustment for body composition variables, height and lean mass. Conversely, we found PFNA to be directly related to ABMC.
Measures of body size including fat mass, lean mass, and height, have been found to be associated with bone mineral density [23,14]. Our results suggest that the effect of prenatal PFAS exposure on body size may influence the effect of PFAS exposure on bone health measures, given the attenuation of the associations we observed by height adjustment. However, the precise mechanisms appeared to differ since associations with PFOS were attenuated to a similar extent by height and lean mass adjustment, whereas those for PFOA and PFNA were only attenuated by height adjustment.
The mechanism through which body size influences the association between maternal PFAS concentrations on bone health remains unclear. In a previous ALSPAC study of 447 mother daughter dyads, maternal PFAS concentrations were associated with fetal and postnatal growth. While a negative association was seen between PFAS exposure and birthweight, a positive association was seen between PFOS and weight at 20 months [17]. Addition of birthweight to the models presented in this analysis did not change the results (data not shown). An ancillary study (n=359) found no association between prenatal PFAS exposure and percent body fatness in ALSPAC girls at 9 years of age overall [24].
However, previous studies have reported positive associations between PFAS and body size, such as body fatness, lean mass, and height, which is highly related to bone measures. A Danish cohort study with similar PFAS exposure reported an association between maternal serum PFOA concentrations during pregnancy and adiposity in female offspring at 20 years of age (n=345) [25]. The direct association seen in other cohort studies between body size and BMD would suggest that maternal PFAS concentrations would also be positively association with BMD. In the present study, after controlling for body size, such as height and lean mass, most of the associations between maternal PFAS exposure and BMD, BMC, and BA remain negative. However, positive associations were seen between maternal PFAS concentrations and AMBC, a bone measure less dependent on body size, possibly reflecting a reciprocal relationship between bone size and volumetric bone mineral density as previously reported in this cohort on pQCT scan [26]
There are many risk factors for fractures and discussing early life factors in relation to fracture risk later in life is particularly important when discussing implications of our findings on future skeletal health. Bone size is an important determinant of fracture risk in biomechanical terms, however, it is unclear how total body measures used in our study relate to fracture risk in childhood and later in life [27]. A previous ALSPAC study found total body BMD to be associated with fracture risk in childhood; however, the actual parameters used in the previous study, such as BMD adjusted for body size, were not significantly associated with prenatal PFAS concentrations in the current study [28]. Additionally, while hip BMD is known to predict subsequent fracture risk, we did not observe a significant association with prenatal PFAS exposure. Toxicological studies have reported that maternal exposure can result in PFAS accumulation in bone tissues and causes decreases in bone mineral density, particularly with PFOA [9].
To our knowledge, no previous epidemiologic studies have examined the association between in utero PFAS exposure and bone health later in life. However, using cross sectional data from NHANES, two studies reported inverse associations between PFAS exposure and bone health in women, results of which are consistent with those reported here. The first study used data on 2,339 adults from the 2005–2006 and 2007–2008 cycles of NHANES to examine the association between serum concentrations of PFOA and PFOS and total lumbar spine and total hip BMD in adults. This study found inverse associations between serum PFOS concentrations and total lumbar spine BMD in premenopausal women [12]. A second study using samples from the 2009–2010 NHANES population found similar results. This study found among adult women overall, higher serum PFOS and PFHxS concentrations were associated with lower BMD and higher prevalence of osteoporosis [13]. Although these studies suggest a relationship between PFAS exposure and bone health in adults through the possible disruption of hormonal functioning, they do not examine the potential prenatal effects of PFAS in utero exposure on subsequent bone development through adolescence.
There are potential limitations to the current analyses. We weighted linear regression models to account for study design; however, bias may be introduced by analyzing data from a sample population originally selected for a nested case-control study for another outcome. When comparing the sample used for the current analysis (n=257) to the original sample 448 mother daughter dyads, the current sample included more mothers in the higher education categories. Blood samples for maternal PFAS concentrations were only collected once during pregnancy and were assumed to be both indicators of exposure throughout pregnancy and a proxy for fetal exposure. We controlled for gestational age at sample collection to account for variability in the timing of sample collection across pregnancies.
In conclusion, prenatal PFAS exposure may be associated with bone mass and size in adolescent girls; however, it is unclear whether these associations are mostly explained by the effects of PFAS on body size. Additional research about the implications of these results for skeletal health, including fracture risk, can help further explain the relationship between prenatal exposures to environmental exposures, such as PFAS, and bone health.
Supplementary Material
Acknowledgements
We are extremely grateful to all the families who took part in this study, the midwives for their help in recruiting them, and the whole ALSPAC team, which includes interviewers, computer and laboratory technicians, clerical workers, research scientists, volunteers, managers, receptionists and nurses. The UK Medical Research Council and the Wellcome Trust (Grant ref: 102215/2/13/2) and the University of Bristol provide core support for ALSPAC. This research was also funded by the US Centers for Disease Control and Prevention (CDC). The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the CDC, the Public Health Service, or the US Department of Health and Human Services.
Footnotes
Conflict of Interest
Zuha Jeddy, Jonathan H. Tobias, Ethel V. Taylor, Kate Northstone, W. Dana Flanders, Terryl J. Hartman declare that they have no conflict of interest.
References
- 1.Diamanti-Kandarakis E, Bourguignon J-P, Giudice LC, Hauser R, Prins GS, Soto AM, Zoeller RT, Gore AC (2009) Endocrine-disrupting chemicals: an Endocrine Society scientific statement. Endocrine reviews 30 (4):293–342 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Harbison RD, Bourgeois MM, Johnson GT (2015) Hamilton and Hardy’s industrial toxicology. John Wiley & Sons, [Google Scholar]
- 3.Wang Z, Cousins IT, Scheringer M, Hungerbühler K (2013) Fluorinated alternatives to long-chain perfluoroalkyl carboxylic acids (PFCAs), perfluoroalkane sulfonic acids (PFSAs) and their potential precursors. Environment international 60:242–248 [DOI] [PubMed] [Google Scholar]
- 4.Lau C, Butenhoff JL, Rogers JM (2004) The developmental toxicity of perfluoroalkyl acids and their derivatives. Toxicology and applied pharmacology 198 (2):231–241 [DOI] [PubMed] [Google Scholar]
- 5.Olsen GW, Burris JM, Ehresman DJ, Froehlich JW, Seacat AM, Butenhoff JL, Zobel LR (2007) Half-life of serum elimination of perfluorooctanesulfonate, perfluorohexanesulfonate, and perfluorooctanoate in retired fluorochemical production workers. Environmental health perspectives:1298–1305 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Newbold RR (2011) Developmental exposure to endocrine-disrupting chemicals programs for reproductive tract alterations and obesity later in life. The American journal of clinical nutrition 94 (6 Suppl):1939S–1942S [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Apelberg BJ, Witter FR, Herbstman JB, Calafat AM, Halden RU, Needham LL, Goldman LR (2007) Cord serum concentrations of perfluorooctane sulfonate (PFOS) and perfluorooctanoate (PFOA) in relation to weight and size at birth. Environmental health perspectives:1670–1676 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Calafat AM, Wong L-Y, Kuklenyik Z, Reidy JA, Needham LL (2007) Polyfluoroalkyl chemicals in the US population: data from the National Health and Nutrition Examination Survey (NHANES) 2003–2004 and comparisons with NHANES 1999–2000. Environmental health perspectives:1596–1602 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Koskela A, Finnilä M, Korkalainen M, Spulber S, Koponen J, Håkansson H, Tuukkanen J, Viluksela M (2016) Effects of developmental exposure to perfluorooctanoic acid (PFOA) on long bone morphology and bone cell differentiation. Toxicology and applied pharmacology 301:14–21 [DOI] [PubMed] [Google Scholar]
- 10.Beard J, Marshall S, Jong K, Newton R, Triplett-McBride T, Humphries B, Bronks R (2000) 1, 1, 1-trichloro-2, 2-bis (p-chlorophenyl)-ethane (DDT) and reduced bone mineral density. Archives of Environmental Health: An International Journal 55 (3):177–180 [DOI] [PubMed] [Google Scholar]
- 11.Wallin E, Rylander L, Hagmar L (2004) Exposure to persistent organochlorine compounds through fish consumption and the incidence of osteoporotic fractures. Scandinavian journal of work, environment & health:30–35 [DOI] [PubMed] [Google Scholar]
- 12.Lin L-Y, Wen L-L, Su T-C, Chen P-C, Lin C-Y (2014) Negative association between serum perfluorooctane sulfate concentration and bone mineral density in US premenopausal women: NHANES, 2005–2008. The Journal of Clinical Endocrinology & Metabolism 99 (6):2173–2180 [DOI] [PubMed] [Google Scholar]
- 13.Khalil N, Chen A, Lee M, Czerwinski SA, Ebert JR, DeWitt JC, Kannan K (2016) Association of perfluoroalkyl substances, bone mineral density, and osteoporosis in the US population in NHANES 2009–2010. Environmental Health Perspectives (Online) 124 (11):81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Clark EM, Tobias JH (2011) Fat mass is a positive predictor of bone mass in adolescents. Journal of Bone and Mineral Research 26 (3):673–673 [DOI] [PubMed] [Google Scholar]
- 15.Fraser A, Macdonald-Wallis C, Tilling K, Boyd A, Golding J, Smith GD, Henderson J, Macleod J, Molloy L, Ness A (2013) Cohort profile: the Avon Longitudinal Study of Parents and Children: ALSPAC mothers cohort. International journal of epidemiology 42 (1):97–110 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Boyd A, Golding J, Macleod J, Lawlor DA, Fraser A, Henderson J, Molloy L, Ness A, Ring S, Smith GD (2012) Cohort profile: the ‘children of the 90s’—the index offspring of the Avon Longitudinal Study of Parents and Children. International journal of epidemiology:dys064 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Maisonet M, Terrell ML, McGeehin MA, Christensen KY, Holmes A, Calafat AM, Marcus M (2012) Maternal concentrations of polyfluoroalkyl compounds during pregnancy and fetal and postnatal growth in British girls. Environmental health perspectives 120 (10):1432–1437 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kuklenyik Z, Needham LL, Calafat AM (2005) Measurement of 18 perfluorinated organic acids and amides in human serum using on-line solid-phase extraction. Analytical chemistry 77 (18):6085–6091 [DOI] [PubMed] [Google Scholar]
- 19.Stagi S, Cavalli L, Cavalli T, de Martino M, Brandi ML (2016) Peripheral quantitative computed tomography (pQCT) for the assessment of bone strength in most of bone affecting conditions in developmental age: a review. Italian journal of pediatrics 42 (1):88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Sayers A, Tobias JH (2010) Fat mass exerts a greater effect on cortical bone mass in girls than boys. The Journal of Clinical Endocrinology & Metabolism 95 (2):699–706 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Christensen KY, Maisonet M, Rubin C, Holmes A, Calafat AM, Kato K, Flanders WD, Heron J, McGeehin MA, Marcus M (2011) Exposure to polyfluoroalkyl chemicals during pregnancy is not associated with offspring age at menarche in a contemporary British cohort. Environment international 37 (1):129–135 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Kleinbaum DG, Kupper LL, Morgenstern H (1982) Epidemiologic research: principles and quantitative methods. John Wiley & Sons, [Google Scholar]
- 23.Edelstein SL, Barrett-Connor E (1993) Relation between body size and bone mineral density in elderly men and women. American journal of epidemiology 138 (3):160–169 [DOI] [PubMed] [Google Scholar]
- 24.Hartman TJ, Calafat AM, Holmes AK, Marcus M, Northstone K, Flanders WD, Kato K, Taylor EV (2017) Prenatal Exposure to Perfluoroalkyl Substances and Body Fatness in Girls. Childhood Obesity; [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Halldorsson TI, Rytter D, Haug LS, Bech BH, Danielsen I, Becher G, Henriksen TB, Olsen SF (2012) Prenatal exposure to perfluorooctanoate and risk of overweight at 20 years of age: a prospective cohort study. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Steer CD, Sayers A, Kemp J, Fraser WD, Tobias JH (2014) Birth weight is positively related to bone size in adolescents but inversely related to cortical bone mineral density: Findings from a large prospective cohort study. Bone 65:77–82 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Clark EM, Tobias JH, Ness AR (2006) Association Between Bone Density and Fractures in Children: A Systematic Review and Meta-analysis. Pediatrics 117 (2):e291–e297. doi: 10.1542/peds.2005-1404 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Clark EM, Ness AR, Bishop NJ, Tobias JH (2006) Association Between Bone Mass and Fractures in Children: A Prospective Cohort Study. Journal of Bone and Mineral Research 21 (9):1489–1495. doi: 10.1359/jbmr.060601 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.