2026: Association of Maternal Bone Resorption with Urinary Fluoride Levels During Pregnancy

[pfpcnews]

Bashash M, Peterson KE, Martinez-Mier EA, Lamadrid-Figueroa H, Santiago Á, Ettinger AS, Téllez-Rojo MM, Hu H - "Association of Maternal Bone Resorption with Urinary Fluoride Levels During Pregnancy" Toxics 14(9):798 (2026)
https://doi.org/10.3390/toxics14090798

Abstract

Evidence is accumulating linking prenatal fluoride exposure with adverse neurodevelopment. Maternal urinary fluoride (MUF) is a well-established biomarker of fluoride exposure. Because most fluoride is stored in calcified tissues, pregnancy-associated changes in bone metabolism may contribute to fetal fluoride exposure. We evaluated whether maternal bone resorption, indexed by urinary N-telopeptide of type I collagen (NTX), is associated with MUF across pregnancy. Participants were drawn from the Early Life Exposures in Mexico to Environmental Toxicants (ELEMENT) birth cohorts. Each woman contributed one paired MUF–NTX measurement per trimester. Both biomarkers were creatinine-normalized to account for urine dilution. Trimester-specific generalized linear models with a Gamma distribution estimated associations between NTX and MUF, adjusting for maternal age, birth order, infant birthweight, and cohort. Across the trimesters, MUF was relatively stable (0.89 ± 0.37, 0.91 ± 0.41, and 0.83 ± 0.42 mg/L), whereas NTX increased with gestation (68.0 ± 33.5, 99.0 ± 50.3, and 132.4 ± 65.4 nmol BCE/mM creatinine). Higher NTX was consistently associated with higher MUF: β = 0.00315 (p < 0.0001) in Trimester 1, β = 0.00207 (p = 0.001) in Trimester 2, and β = 0.00253 (p < 0.0001) in Trimester 3, corresponding to 11%, 11%, and 18% higher MUF per 1-SD increase in NTX. Results were unchanged after adjustment for estimated fluoride intake, with no evidence of non-linearity. Maternal bone resorption was independently associated with MUF, consistent with a skeletal contribution to urinary fluoride. Research may be needed on measures that can blunt bone mobilization of fluoride during pregnancy.

[admin]

PFPC Commentary:

Yet another ELEMENT study, again involving Howard Hu, Morteza Bashash, and Esperanza Angeles Martinez-Mier.

Not surprisingly, there is not a word about iodine, despite studies documenting very high iodine intake in the Mexico City population (see: Flores-Rebollar et al., 2014, 2015, 2021). This omission is particularly important in a study specifically concerned with fluoride retention in bone, bone resorption and urinary fluoride. Iodine influences fluoride retention in bone and urinary fluoride excretion, yet maternal iodine status was not considered. Nor is there any discussion of thyroid function, despite the well-established influence of thyroid hormones on bone turnover and urinary NTX.

There are several real problems with this study, and some go directly to the central interpretation rather than being minor limitations.

The most important ones are:

1. NTX is strongly pregnancy-dependent. Their own NTX means almost double across pregnancy, from 68.0 in T1 to 132.4 in T3. They acknowledge pregnancy-related bone resorption, but they do not seriously examine the fact that this very large rise is expected physiologically. They stratify by trimester, but do not adjust for exact gestational week within trimester.

2. They do not account for thyroid status. Their adjustment set is maternal age, firstborn status, birthweight and cohort, with some sensitivity analyses involving BMI, calcium supplementation and estimated fluoride intake. There is no TSH, FT4, FT3, thyroid disease or thyroid autoimmunity. Yet urinary NTX is demonstrably thyroid-sensitive. Hyperthyroidism raises NTX, NTX correlates with FT3/FT4, and experimentally administered T3 raises urinary NTX. That is directly relevant because NTX is their principal predictor.

3. They do not account for iodine status. This is a major omission in a Mexico City population, where studies have documented high iodine intake. Iodine is relevant to both sides of the association they are examining. Iodine intake influences fluoride metabolism, including fluoride retention in bone and urinary fluoride excretion. It can therefore affect both the skeletal fluoride reservoir that the authors propose is being mobilized and the amount of fluoride subsequently appearing in urine. At the same time, iodine status influences thyroid function and bone metabolism, and chronic excess iodine has been shown experimentally to alter osteoclast activity. Thyroid status itself also affects bone turnover and urinary NTX. Thus iodine can influence both MUF and the biological process represented by NTX, yet it is completely absent from their analysis.

4. Their causal direction is not established. They observe:

NTX ↑ ↔ MUF ↑

and interpret that as:

bone resorption ↑ → skeletal fluoride release ↑ → MUF ↑

But their analysis is cross-sectional within trimester. They themselves acknowledge that it "does not establish causation or quantify the bone-derived fraction". More importantly, NTX cannot necessarily be treated as a biologically independent upstream measure of bone resorption. Fluoride itself has dose-dependent effects on osteoclast formation and resorptive function, with experimental studies showing stimulatory and inhibitory effects at different concentrations. Iodine and thyroid status can also alter osteoclast activity and bone turnover. The observed NTX-MUF association therefore cannot distinguish skeletal fluoride release from effects of fluoride, iodine or thyroid physiology on the bone-resorption process itself, nor can this cross-sectional analysis establish the direction of the relationship.

5. The creatinine issue is potentially very important. Both NTX and MUF came from the same urine specimen, and both were divided by the same urinary creatinine concentration. The authors argue that creatinine-adjusting both biomarkers "forecloses urine dilution as a shared cause" of the observed association. But statistically, regressing one creatinine-normalized ratio on another can itself create or strengthen a positive association because the same denominator appears in both variables. This is especially relevant in pregnancy because urinary creatinine concentration changes over gestation, so the common denominator itself is physiologically variable. Creatinine normalization may reduce urine-dilution error, but it does not "foreclose" a spurious correlation.

6. Their trimester means do not behave as their proposed mechanism might naively predict. NTX nearly doubles:

68.0 → 99.0 → 132.4

while MUF does not:

0.89 → 0.91 → 0.83 mg/L.

That does not disprove skeletal fluoride release, because the trimester-specific analytic samples are not identical and these means do not represent a complete within-woman longitudinal comparison. But it certainly deserves discussion if increased bone resorption is being presented as an important determinant of MUF.

7. The external fluoride adjustment is only an estimate. They derive "approximate total daily fluoride intake" from reported water, beverages and foods. That cannot establish that the residual association is independent of total external fluoride exposure. Their own Introduction acknowledges other sources, including dental products. Residual exposure confounding remains entirely possible.

8. They have no measure of skeletal fluoride burden, even though skeletal fluoride burden is central to their proposed mechanism. If bone is the source of the additional urinary fluoride, the amount of fluoride actually stored in bone should be fundamental. Yet skeletal fluoride accumulation is not determined simply by current external fluoride intake. Iodine status influences fluoride retention and deposition in bone, as well as urinary fluoride excretion. The authors therefore have neither the skeletal fluoride measurements nor the iodine data needed to characterize the reservoir they claim is being mobilized. They demonstrate an association between NTX and MUF, but they cannot determine how much fluoride was stored in bone, how iodine status influenced that storage, or how much of the urinary fluoride originated from subsequent skeletal release.

9. The trimester datasets are not the same women. The analytic samples are 423, 197 and 239 women across T1-T3, and the paper describes the models as cross-sectional within trimester. Therefore, the striking gestational increase in mean NTX is not a clean within-woman longitudinal comparison in this analysis.

10. The supplement adds another problem. In every trimester, women excluded from the analytic model had significantly higher urinary NTX than those included. Most exclusions occurred because creatinine-adjusted MUF was unavailable, with smaller additional losses from NTX trimming and incomplete covariate data. As a result, the final regression samples were systematically shifted toward lower NTX values. Because NTX was the study's main predictor, this differential inclusion could affect the estimated NTX-MUF association and limits the representativeness of the analytic sample. The direction and magnitude of any resulting bias cannot be determined from the published data.

See below:

A) Studies showing effects of thyroid status on urinary NTX
B) Studies on the influence of pregnancy on urinary NTX

[admin]

Related Literature

Asakura Y, Adachi M, Tachibana K - "Bone Mineral Density and Bone Metabolic Markers in Children with Hyperthyroidism Before and During Treatment" Clin Pediatr Endocrinol 9(2):97-104 (2000)
https://doi.org/10.1297/cpe.9.97
"Bone resorption rapidly ameliorated, and accelerated bone formation persisted for a few months after normalization of thyroid function."
Untreated hyperthyroid children had markedly high urinary NTX, which fell rapidly toward normal during treatment.

Çulha C, Demirbaş B, Güler S, Şahin İ, Karakurt F, Taşkın S, Kutlutürk F, Serter R, Aral Y - "Bone metabolism in patients with thyrotoxicosis" Medical Network Klinik Bilimler ve Doktor 7(4):429-433 (2001)
No DOI identified
"Thyroid hormones lead to increase in urinary N-telopeptide levels by increasing bone resorption rate and decreasing bone mineral density. Besides normalization of carbonhydrate, lipid and protein meatbolisms, urinary N-telopeptide excretion also decreases along with the increase in bone mineral density."
Thyroid dysfunction itself changes urinary NTX. It therefore cannot automatically be treated as a bone-resorption measure independent of thyroid status.

Heemstra KA, van der Deure WM, Peeters RP, Hamdy NA, Stokkel MP, Corssmit EP, Romijn JA, Visser TJ, Smit JW - "Thyroid hormone independent associations between serum TSH levels and indicators of bone turnover in cured patients with differentiated thyroid carcinoma" Eur J Endocrinol 159(1):69-76 (2008)
https://doi.org/10.1530/EJE-08-0038
"Serum TSH was related to indicators of bone remodeling independently of thyroid hormone levels."
This study specifically measured urinary N-telopeptide of collagen cross-links and found the inverse association between TSH and bone-turnover indices remained after controlling for FT4 and T3.

Karga H, Papaioannou G, Polymeris A, Papamichael K, Karpouza A, Samouilidou E, Papaioannou P - "The effects of recombinant human TSH on bone turnover in patients after thyroidectomy" J Bone Miner Metab 28(1):35-41 (2010)
https://doi.org/10.1007/s00774-009-0098-y
"The study provides evidence that rhTSH produces a transient inhibition of bone resorption."
This is particularly interesting mechanistically. In thyroidectomized postmenopausal women maintained on thyroxine, rhTSH produced a significant decrease in urinary N-terminal and C-terminal telopeptides, while thyroid hormone levels remained essentially stable.

Majima T, Komatsu Y, Doi K, Takagi C, Shigemoto M, Fukao A, Morimoto T, Corners J, Nakao K - "Negative correlation between bone mineral density and TSH receptor antibodies in male patients with untreated Graves' disease" Osteoporos Int 17(7):1103-1110 (2006)
https://doi.org/10.1007/s00198-006-0091-4
"These results indicate a high prevalence of cortical bone loss in male patients with Graves' disease, especially elderly patients."
In 56 newly diagnosed untreated Graves' patients, radial BMD Z-score was negatively associated with urinary NTX/creatinine, FT4, TSH-receptor antibodies, and thyroid-stimulating antibody.

Mora S, Weber G, Marenzi K, Signorini E, Rovelli R, Proverbio MC, Chiumello G - "Longitudinal Changes of Bone Density and Bone Resorption in Hyperthyroid Girls During Treatment" J Bone Miner Res 14(11):1971-1977 (1999)
https://doi.org/10.1359/jbmr.1999.14.11.1971
"Hyperthyroidism in young girls is associated with high urine levels of NTX, indicating enhanced bone resorption rates. Moreover, the degrees of osteopenia and bone resorption seem to be related to the severity of hyperthyroidism."
This paper also states immediately before the conclusion that urinary NTX was "highly correlated with both FT4 and FT3 serum levels at diagnosis, implying a direct relationship between the degree of hyperthyroidism and bone degradation."

Pantazi H, Papapetrou PD - "Changes in parameters of bone and mineral metabolism during therapy for hyperthyroidism" J Clin Endocrinol Metab 85(3):1099-1106 (2000)
https://doi.org/10.1210/jcem.85.3.6457
"The failure of the markers of bone turnover to normalize after 1 yr of therapy indicates an on-going high rate of bone turnover."
In newly diagnosed Graves' disease, the urinary NTX z-score was 9.3 ± 1.3 before treatment. It declined with methimazole but had still not normalized after one year. They also found a significant positive correlation between T3 and urinary NTX early in treatment.

Rosen HN, Dresner-Pollak R, Moses AC, Rosenblatt M, Zeind AJ, Clemens JD, Greenspan SL - "Specificity of urinary excretion of cross-linked N-telopeptides of type I collagen as a marker of bone turnover" Calcif Tissue Int 54(1):26-29 (1994)
https://doi.org/10.1007/BF00316285
"Urinary NTX is responsive to acute thyroid hormone-induced increases and bisphosphonate-induced decreases in bone resorption, and may reflect these changes more accurately than PYD or HYP."
That is particularly strong because they administered T3 experimentally. Urinary NTX/creatinine increased significantly following T3.

[admin]

Besides the findings of Bashash et al., here are other studies showing change of urinary NTX during pregnancy:

Avendaño-Badillo D, Hernández-Ávila M, Hernández-Cadena L, Rueda-Hernández G, Solano-González M, Ibarra LG, Hu H, Téllez-Rojo MM - "High dietary calcium intake decreases bone mobilization during pregnancy in humans" Salud Publica Mex 51(Suppl 1):S100-S107 (2009)
https://doi.org/10.1590/S0036-36342009000700013
"Our results suggest that calcium ingestion, specifically from dairy products, reduces bone resorption during pregnancy."
Urinary NTX: 76.50 ± 38 in T1, 101.02 ± 48.86 in T2, and 144.83 ± 61.33 nmol BCE/mmol creatinine in T3. The paper explicitly reports: "Progressive increases in NTx levels were observed during pregnancy."

Black AJ, Topping J, Durham B, Farquharson RG, Fraser WD - "A Detailed Assessment of Alterations in Bone Turnover, Calcium Homeostasis, and Bone Density in Normal Pregnancy" J Bone Miner Res 15(3):557-563 (2000)
https://doi.org/10.1359/jbmr.2000.15.3.557
"As assessed by these bone markers, in the first 2 trimesters of pregnancy, bone remodeling is uncoupled with a marked increase in bone resorption. A corresponding increase in formation markers is not observed until the third trimester."
Urinary NTX was measured prospectively from preconception through pregnancy. The authors report a significant increase in urinary NTX, with bone resorption already significantly increased by 14 weeks and continuing upward through pregnancy.

Kaji T, Yasui T, Suto M, Mitani R, Morine M, Uemura H, Maeda K, Irahara M - "Effect of bed rest during pregnancy on bone turnover markers in pregnant and postpartum women" Bone 40(4):1088-1094 (2007)
https://doi.org/10.1016/j.bone.2006.11.018
"Immobilization due to bed rest during pregnancy is associated with increases in bone turnover markers in pregnant and postpartum women. Concentrations of bone resorption markers increased rapidly at the start of bed rest, while the concentration of a bone formation marker gradually increased toward puerperium."
Urinary NTX in the normal longitudinal pregnancy group "increased gradually during pregnancy and showed a peak at 36 weeks of pregnancy, followed by a decrease in the postpartum period." Measurements were made at 10, 26, 30 and 36 weeks.

Naylor KE, Iqbal P, Fledelius C, Fraser RB, Eastell R - "The Effect of Pregnancy on Bone Density and Bone Turnover" J Bone Miner Res 15(1):129-137 (2000)
https://doi.org/10.1359/jbmr.2000.15.1.129
"Pregnancy is a high-bone-turnover state. IGF-I levels may be an important determinant of bone turnover during pregnancy. Elevated bone turnover may explain trabecular bone loss during pregnancy."
Urinary NTX was measured before pregnancy and at 16, 26 and 36 weeks. Bone-resorption markers, including urinary NTX, increased progressively. NTX declined postpartum but remained above prepregnancy values.

Naylor KE, Rogers A, Fraser RB, Hall V, Eastell R, Blumsohn A - "Serum Osteoprotegerin as a Determinant of Bone Metabolism in a Longitudinal Study of Human Pregnancy and Lactation" J Clin Endocrinol Metab 88(11):5361-5365 (2003)
https://doi.org/10.1210/jc.2003-030486
"There was no correlation between change in OPG and bone turnover or bone mineral density (P > 0.05), and the physiological importance of elevated OPG in human pregnancy remains uncertain."
Urinary NTX was 219 ± 41% higher at 36 weeks than before conception, P < 0.001. Samples were obtained before conception and at 16, 26 and 36 weeks.

O'Brien EC, Kilbane MT, McKenna MJ, Segurado R, Geraghty AA, McAuliffe FM - "Calcium intake in winter pregnancy attenuates impact of vitamin D inadequacy on urine NTX, a marker of bone resorption" Eur J Nutr 57(3):1015-1023 (2018)
https://doi.org/10.1007/s00394-017-1385-3
"In late pregnancy, during winter months when 25OHD is inadequate, intakes of dietary calcium <1000 mg/day were associated with significantly increased bone resorption (uNTX). Additional dietary calcium is associated with reduced bone resorption in late pregnancy, with greater effect observed in winter."
Urinary NTX was measured at 13 and 28 weeks. The paper reports increasing uNTX during pregnancy, with the magnitude modified substantially by calcium intake, vitamin D status and season.

Vargas Zapata CL, Donangelo CM, Woodhouse LR, Abrams SA, Spencer EM, King JC - "Calcium homeostasis during pregnancy and lactation in Brazilian women with low calcium intakes: a longitudinal study" Am J Clin Nutr 80(2):417-422 (2004)
https://doi.org/10.1093/ajcn/80.2.417
"Calcium homeostasis appears to be attained by a more efficient intestinal calcium absorption during pregnancy and by renal calcium conservation during both pregnancy and lactation. IGF-I and PTH seem to play major roles in the adjustment of calcium metabolism during pregnancy and lactation."
Urinary NTX increased 136% from early pregnancy at 10-12 weeks to late pregnancy at 34-36 weeks, P < 0.001, and remained elevated during early lactation.

Xue Y, Jia W, Zhang H, Dong J, Clemens JD, Tian X - "Urinary cross-linked N-telopeptides of type I collagen and bone metabolic diseases" Chin Med J (Engl) 112(2):149-152 (1999)
https://pubmed.ncbi.nlm.nih.gov/11593583/
"NTx in urine is a specific and sensitive indicator of bone resorption and is able to distinguish normal premenopause from late osteoporotic patients. NTx could be used as diagnostic information about metabolic bone diseases, and to monitor antiresorptive therapy."
Pregnancy was associated with a 3.6-fold increase in urinary NTX compared with age-matched controls.

Yamaga A, Taga M, Minaguchi H - "Changes in Urinary Excretions of C-Telopeptide and Cross-linked N-Telopeptide of Type I Collagen during Pregnancy and Puerperium" Endocr J 44(5):733-738 (1997)
https://doi.org/10.1507/endocrj.44.733
"These results further confirm our previous evidence that bone resorption is enhanced during the 3rd trimester of gestation and puerperium and suggest that urinary CTX and NTX measured by ELISA, which is more convenient than HPLC, are useful markers to assess bone resorption during peripuerperal periods."
Urinary NTX was significantly increased in the third trimester and remained high during the puerperium compared with nonpregnant and early-pregnancy women.

Zeni SN, Ortela Soler CR, Lazzari A, López L, Suarez M, Di Gregorio S, Somoza JI, de Portela ML - "Interrelationship between bone turnover markers and dietary calcium intake in pregnant women: a longitudinal study" Bone 33(4):606-613 (2003)
https://doi.org/10.1016/S8756-3282(03)00203-5
"In summary, not only hormonal changes in calcium metabolism that occur during pregnancy but also other considerations, such as low dietary calcium intake, may lead to an increment in the biological activity of the skeleton."
Urinary NTX and the other bone-resorption markers were elevated relative to nonpregnant values throughout pregnancy, with the largest increase in the third trimester.

[admin]

Here is just one study showing the effects of iodine on bone. There are, of course, other studies such as those by Zhao et al., exploring effects of various doses of both fluoride and iodine on fluoride retention in bone. And there are thousands of studies on thyroid and bone.

• Zhang Y, Zhao X, Zhao N, Meng H, Zhang Z, Song Y, Shan L, Zhang X, Zhang W, Sang Z - "Chronic Excess Iodine Intake Inhibits Bone Reconstruction Leading to Osteoporosis in Rats" J Nutr 154(4):1209-1218 (2024)
  https://doi.org/10.1016/j.tjnut.2024.02.005
  "Chronic excess iodine intake can lead to abnormal changes in skeletal structure, resulting in reduced activity of osteoblasts and osteoclasts, which inhibits the process of bone reconstruction and may lead to osteoporosis."

That finding is important for the ELEMENT issue because it shows that iodine exposure can alter the actual process that generates NTX, namely osteoclastic bone resorption. In this experiment, excess iodine did not simply increase turnover. At sufficiently high chronic exposure, it suppressed osteoclast activity. So an iodine-NTX relationship could itself be non-linear and thyroid-state dependent.

There is also now a large human study specifically examining iodine nutrition and bone metabolism in 1,207 thyroid-disease-free adults. It compared iodine-fortification, iodine-adequate and iodine-excess areas and examined UIC and serum iodine. UIC ≥300 µg/L and serum iodine >90 µg/L were associated with poorer bone outcomes. They did not measure NTX, but they found relationships between iodine status, thyroid hormones, ALP, calcium and BMD even after attempting to account for thyroid dysfunction.

• Zhou Z, Liu J, Li B, Chen Y, He Y, Ren B, Wei Q, Jin M, Chen Y, Gao H, Wan S, Liu L, Shen H - "Relationships between iodine nutrition status, and bone mineral density and bone metabolism: a cross-sectional study of 1207 thyroid disease-free adults in China" Br J Nutr 133(9):1180-1192 (2025)
  https://doi.org/10.1017/S0007114525000790
  "Iodine excess can not only lead to changes in bone metabolism parameters and BMD, but is also a risk factor for osteopenia and osteoporosis."

Perhaps one day someone will check the NHANES 2001-2002 data. They actually measured both urinary iodine and urinary NTX, so a direct iodine-UIC versus NTX analysis is possible in that dataset.