BPA-Free Is Not Safe

Evidence Review

Hormonal activity equivalence — Rochester & Bolden (2015)

The Rochester and Bolden systematic review [1] is the defining reference for this topic. Searching literature through 2014, the authors identified 32 in vitro and in vivo studies evaluating the hormonal activity of BPS and BPF. Their analysis found that both compounds demonstrate estrogenic activity broadly comparable to BPA, with activity detected across estrogen receptor alpha, estrogen receptor beta, and membrane estrogen receptor pathways. They also found antiestrogenic, androgenic, and antiandrogenic activity, indicating broad endocrine disruption rather than a narrow estrogenic effect. The authors concluded there is "no evidence that BPS and BPF are safer alternatives to BPA." A key limitation is that most studies available at the time of review were in vitro; the authors noted the urgent need for more in vivo and epidemiological data.

Population exposure — Lehmler et al. (2018)

Using NHANES 2013–2014 urinary data from a nationally representative sample of U.S. children and adults, Lehmler et al. [2] found BPS detectable in 89.4% of samples and BPF in 66.5%. Median urinary BPS concentration was 0.18 µg/L; the 95th percentile reached 3.4 µg/L. These detection rates are comparable to BPA's ubiquity at the height of its use. The data confirm that regulatory action on BPA has not meaningfully reduced total bisphenol body burden — it has only shifted which bisphenol is present. Children tended to have higher BPS levels than adults, consistent with greater hand-to-mouth behavior and proximity to treated surfaces.

Neurodevelopmental effects — Kinch et al. (2015)

Kinch and colleagues [3] exposed zebrafish embryos to BPA and BPS at concentrations ranging from 0.0068 µg/L to 68 µg/L during the first 24 hours of development. Both compounds produced a 240% increase in hypothalamic neuron production at concentrations as low as 0.0068 µg/L — a concentration within the range of detected environmental levels. The mechanism involved estrogen receptor signaling: blocking estrogen receptors abolished the effect. This is significant because it shows that BPS is not merely weakly estrogenic at high doses but is biologically active at concentrations that overlap with real-world exposure, and at the earliest stages of neural development. The authors acknowledged that zebrafish models have limitations in predicting human outcomes, but noted the conservation of estrogen receptor pathways across vertebrates makes the finding relevant.

Cardiac effects — Gao et al. (2015)

Working with ex vivo rat heart preparations and isolated ventricular myocytes, Gao and colleagues [4] applied BPS at concentrations of 1 nM and 1 µM — representing low and moderate environmental/occupational exposures respectively. BPS induced rapid, non-genomic responses in cardiac tissue: at 1 nM, it increased the frequency of spontaneous calcium release events (sparks) — a marker of arrhythmia susceptibility — by approximately 60% in female myocytes. In intact heart preparations under isoproterenol stress (simulating sympathetic activation), BPS-treated female hearts showed a marked increase in ventricular arrhythmia. Effects were sex-specific, with female tissue consistently more sensitive, a pattern also seen with BPA. The researchers proposed that BPS activates estrogen receptor-mediated signaling in cardiac cells through a non-transcriptional pathway, directly altering ion channel behavior.

Developmental toxicity of BPF — Liang et al. (2017)

This study [5] compared BPA, BPS, BPF, and BPAF in zebrafish embryos across multiple endpoints: acute toxicity, developmental malformations, and estrogenic gene expression. BPF was found to be estrogenically equivalent to BPA based on yolk sac estrogenic gene induction (ranking: BPAF > BPA = BPF > BPS). All four bisphenols produced statistically significant developmental deformities — cardiac edema, craniofacial abnormalities, spinal curvature, and trunk defects — at concentrations of 1 mg/L and above. The study is valuable for directly comparing the four analogs head-to-head in the same experimental model, confirming that the structural analogy translates to biological equivalence across multiple endpoints.

Carcinogenicity assessment — Edaes & de Souza (2022)

This narrative review [6] examined the oncogenic potential of BPS and BPF relative to BPA, focusing on mechanisms including estrogen-receptor-mediated cell proliferation, oxidative stress induction, and epigenetic modification. The authors concluded that the weight of evidence does not support treating BPS or BPF as safer replacements from a carcinogenicity standpoint. They called for regulatory frameworks that assess bisphenol analogs as a class rather than evaluating each in isolation after it has already achieved widespread use — a regulatory pattern that has allowed BPS and BPF to repeat BPA's trajectory of widespread deployment before adequate safety evaluation.

Overall evidence assessment

The evidence base for BPS and BPF health risks is growing but still substantially smaller than the literature on BPA. Most mechanistic studies use in vitro cell models or zebrafish embryos; robust human epidemiological data directly linking BPS or BPF exposure to disease outcomes are limited. However, the structural similarity to BPA, the consistent findings of estrogenic activity across independent laboratories, the near-universal population exposure confirmed by NHANES, and the multiple pathways of effect (endocrine, neurological, cardiac) together constitute a credible precautionary concern. The history of BPA — approved and widely used for decades before evidence of harm accumulated — provides a strong argument for applying the precautionary principle to its structural analogs rather than waiting for definitive human disease data to emerge.