Table 3.

Results and the studies selected for the scoping review.

Study:	Chemical:	Population:	Study Design:	End Measurement(s):	Exposure Effect or Biomarker Concentration Measured:	Results:	Main Conclusions:
Åkesson et al. (2014) (review) [28]	Cd	Cd exposure and bone effects: 15 studies with study population from n = 270 to n = 10,978. Cd exposure and fractures: 6 studies with study population from n = 506 to n = 22,173.	Prospective, retrospective, or cross-sectional.	Bone effects or fractures. Most of the studies used DXA and defined low bone mass/osteoporosis based on the z- or T-score.	Threshold of bone effects measured mostly by U-Cd, sometimes from B-Cd. Exposure measured through biomonitoring, in two studies also dietary exposure was measured. Use of different exposure assessment methods (urine, blood and dietary) reduces the possible confounding.	In most of the studies, associations between Cd exposure and low BMD as well as an increased risk of osteoporosis were observed. In four studies no statistically significant Cd-related effect on BMD were observed.	There is an association with exposure to low concentrations of Cd and effects on bone (incl. increased risk of osteoporosis and fractures).
Lv et al. (2017) [29]	Cd	n = 1116 subjects	Population-based study.	BMD and the levels of urinary markers of early renal impairment.	U-Cd	U-Cd concentrations of subjects: 0.21 to 87.31 µg/g creatinine (median 3.97 µg/g). A significant negative association of U-Cd concentrations with BMD. A positive association with U-Cd concentrations with osteoporosis. The odds of osteoporosis increased with increase of U-Cd concentration.	An inverse association between the body burden of Cd and osteoporosis was observed. Toxic effect of Cd on bone may take place together with nephrotoxicity.
Engström et al. (2012) [30]	Cd	n = 2676 women (aged 56–69 years)	Population-based prospective cohort study.	BMD at the total body, femoral neck, and lumbar spine by using DXA. Risk of osteoporosis: hip or spine. Risk of any first incident fracture.	U-Cd	An inverse association of dietary Cd and BMD at the total body (p = 0.045) and the lumbar spine (p = 0.004). No association at the femoral neck (p = 0.89). When adjusting for dietary factors (calcium, magnesium, iron and fiber), the inverse association became more pronounced. Comparison of high dietary Cd exposure (≥13 µg/day, median) with lower exposures (˂13 µg/day) resulted in a 32% increased risk of osteoporosis (95% CI: 2–71%) and 31% increased risk for any first incident fracture (95% CI: 2–69%). Comparison of high dietary Cd with high U-Cd (≥0.50 µg/g creatinine) among never-smokers resulted in OR = 2.65 (1.43–4.91) for osteoporosis and OR = 3.05 (1.66–5.59) for fractures.	Even low-level Cd exposure from food was associated with low BMD and an increased risk of osteoporosis and fractures.
Wallin et al. (2016) [31]	Cd	n = 936 men (aged 70–81 years)	Prospective cohort study.	BMD at the total body, hip, and lumbar spine by using DXA. Incident fractures.	U-Cd	Significant negative associations between U-Cd and BMD were observed; lower BMD (4% to 8%) in all sites was detected for in the fourth quartile of U-Cd. Positive associations between U-Cd and incident fractures, especially non-vertebral osteoporosis fractures in the fourth quartile of U-Cd was observed. U-Cd was significantly associated with non-vertebral osteoporosis fractures (adjusted hazard ratio 1.3 to 1.4 per µg creatinine) also among never-smokers, but not with the other fracture groups.	Among elderly men, relatively low Cd exposure through diet and smoking was associated with increased risk of low BMD and osteoporosis-related fractures.
Campbell & Auinger (2007) [32]	Pb	About 40,000 people (≥2 months of age). Analysis included subjects ≥50 years of age (n = 8654) and the final analysis 4689 subjects from NHANES III-survey.	A secondary analysis of a cross-sectional study.	Primary outcome: BMD of the total hip measured by DXA. Clinical outcomes: the presence of back pain and history of osteoporotic-related fracture.	B-Pb	The adjusted mean total hip BMD: Non-Hispanic white males with a blood Pb-level in the lowest tercile versus the highest tercile was 0.961 g/cm2 and 0.934 g/cm2, respectively (p ˂ 0.05); White females with marginally significant difference (0.05 ˂ p ˂ 0.10 in comparison to lowest tercile).	Among white subjects, significant inverse association between Pb exposure and BMD was detected. Between blood Pb-level tercile and clinical outcomes no associations were observed.
Silbergeld et al. (1988) [33]	Pb	2981 women (both black and white) from NHANES II-survey.	-	Pb-status in women, both before and after menopause.	B-Pb (both whole blood and plasma).	After menopause, a highly significant increase in whole blood and calculated plasma Pb concentrations was detected.	Bone Pb is not an inert storage site for Pb. Pb may interact with other factors within post-menopausal osteoporosis aggravating the course of disease by inhibiting activation of vitamin D, uptake of dietary calcium and several regulatory aspects of bone cell function.
Sun et al. (2008) [21]	Pb	n = 249 (191 males and 58 females).	-	BMD measured by monophoton absorptiometry. Osteoporosis defined by Z-score (Z score ˂ −2).	U-Pb and B-Pb.	In both genders, a significant decrease in the groups of the high U-Pb-level compared with the low U-Pb-level was observed. No significant difference between B-Pb and BMD was detected. The prevalence of osteoporosis increased significantly with the increase of both U-Pb and B-Pb. A dose-response relationship between Pb exposure and prevalence of osteoporosis was observed.	U-Pb had closer association with osteoporosis caused by Pb in comparison to B-Pb. Occupational Pb exposure was associated with osteoporosis.
Nash et al. (2004) [34]	Pb	2575 women aged 40–59 years, and the final analysis on 1914 subjects from NHANES III-survey.	Cross-sectional design.	BMD measured in five regions of the femur by DXA.	B-Pb.	A significant inverse relationship between BMD and B-Pb level that remained even after adjusting for other factors traditionally associated with B-Pb. A one-unit change in BMD resulted in 0.6-µg/dL lower geometric mean B-Pb level. The association remained after adjusting for menopausal status.	Among perimenopausal women, due to post-menopausal bone mineral resorption, Pb stored in bone may significantly increase B-Pb levels.
Wong et al. (2015) [35]	Pb	n = 38, post-menopausal women (mean age 76 +/−8).	A cross-sectional observational cohort study.	Volumetric BMD and structural parameters obtained from peripheral quantitative computed tomography images.	B-Pb (whole blood) and XRF scan to obtain bone Pb content at the mid-tibia and calcaneus. Blood Pb and bone Pb were expressed as a blood:bone Pb partition coefficient (PBB) (blood-to-bone).	Higher amounts of bone Pb at the tibia were associated with thinner distal tibia cortices (−0.972, (−1.882–0.061) per 100 µg Pb/g of bone mineral) and integral volumetric BMD (−3.05 (−6.05–0.05) per µg Pb/g of bone mineral. A higher PBB was associated with larger trabecular separation (0.115 (0.053, 0.178), lower trabecular volumetric BMD (−26.83 (−50.37, −3.29) and trabecular number (−0.08 (−0.14, −0.02), per 100 µg Pb/g of bone mineral. Total Pb exposure activities significantly related to bone Pb at the calcaneus (8.29 (0.11–16.48)).	Pb accumulation in bone can have a small harmful effect on bone structure. Greater partitioning of Pb in blood versus bone manifested more dramatic effects on both microstructure and volumetric BMD.
Min & Min (2014) [36]	Phthalates	n = 398, post-menopausal women ≥50 years from from NHANES-surveys (2005–2006 and 2007–2008).	-	Total hip and femur neck BMD measured by DXA and osteoporosis defined by the WHO criteria.	U-phthalate metabolites (11 different).	Increasing of the urinary mono-n-butyl phthalate (MnBP), mono-(3-carboxypropyl) phthalate (MCPP) and monobenzyl phthalate (MBzP) quartiles was significantly associated with reduced total hip or femur neck BMD. Subjects with the highest levels of MCPP phthalate, mono(carboxyoctyl) (MCOP) phthalate and the sum of three di(2-ethylhexyl) phthalate (ƩDEHP) metabolites were more likely to have an increased risk for total hip or femur neck osteoporosis than subjects with the lowest levels of these metabolites.	Increases in the urinary phthalate metabolites (except MCNP, MECPP, MEP and MiBP) were significantly associated with low BMD and high risk of osteoporosis in post-menopausal women. Phthalate exposure may adversely affect bone homeostasis and BMD in humans.
DeFlorio-Barker & Turyk (2016) [37]	Phthalates	n = 480, post-menopausal women from NHANES-survey (2005–2010).	A hypothesis-generating study, a cross-sectional study design.	BMD at the femoral neck and spine.	U-phthalate metabolites.	Mono-ethyl phthalate (MEP), molar sum of low molecular weight metabolites (mono-n-butyl phthalate (MNBP), mono-isobutyl phthalate (MIBP), MEP), molar sum of estrogenic metabolites (MNBP, MIBP, MEP, mono-benzyl phthalate (MBZP)) and an estrogenic equivalency factor were negatively associated with spinal BMD.	Due to the cross-sectional study design, uncertainty concerning the critical time window of exposure, the potential for exposure misclassification and residual confounding, no conclusions about association between phthalate metabolites and BMD in post-menopausal women could be drawn.
Khalil et al. (2016) [38]	PFASs	NHANES-survey (2009–2010): n = 1914, subjects of 12–80 years of age with BMD measurements for total femur (TFBMD), its sub-region femoral neck (FNBMD, n = 1914) and lumbar spine (LSBMD, n = 1605).	-	BMD (total femur, femoral neck, and lumbar spine) measured by DXA and physician-diagnosed osteoporosis.	Four PFASs (PFOA, PFOS, PFHxS, and PFNA) from blood serum.	Among men, higher serum PFAS (except for PFNA) concentrations were observed than among women (p < 0.001). In both sexes, serum PFOS concentrations were inversely associated with FNBMD (p < 0.05). Among women, significant negative associations between PFOS exposure and TFBMD and FNBMD and between PFOA exposure and TFBMD (p < 0.05) was observed. Among post-menopausal women, serum PFOS was negatively associated with TFBMD and FNBMD, and PFNA was negatively associated with TFBMD, FNBMD and LSBMD (all p < 0.05). One log unit increase in serum PFOA, PFHxS and PFNA increased osteoporosis prevalence in women as follows (aOR’s and 95% CI:s reported): 1.84 (1.17–2.905), 1.64 (1.14–2.38) and 1.45 (1.02–2.05), respectively. Among women, the prevalence of osteoporosis was significantly higher in the highest versus the lowest quartiles of PFOA, PFHxS and PFNA: 2.59 (1.01–6.67), 13.20 (2.72–64.15), and 3.23 (1.44–7.21), respectively.	Association between serum PFASs concentrations and lower BMD was observed; varied according to the specific PFAS and bone site assessed. Most of the associations were limited to women. In women, osteoporosis was associated with PFAS exposure.
Lin et al. (2014) [39]	PFASs	NHANES-survey (2005–2006, 2007–2008), n = 2339 (aged ≥ 20 years).	Cross-sectional design.	Total lumbar spine and total hip BMD measured by DXA and history of fractures.	The blood serum samples of PFOA and PFOS.	Among non-menopausal women, a 1-U increase in the natural log-transformed serum PFOS level was associated with a decrease in total lumbar spine BMD by 0.22 g/cm2 (95% CI 0.038–0.007, p = 0.006). No association was detected between PFOA and PFOS concentrations and femoral neck BMD or self-reported fractures.	Among non-menopausal women, a modest effect of serum PFOS concentration with decreased total lumbar spine BMD was observed.