Progress Report/What we have learned: hsCRP as a possible direct clinical indicator for BPA health effects in in a human dietary intervention study

NOTE: This is a draft (January 22, 2020). Further edits and citations will follow.

In addition, the format below has been adapted from that used by Environmental Health Perspectives. This structure is used because it is an effective and familiar shorthand.

This should not be misinterpreted as a draft intended for submission to a scientific publication.


This proof-of-concept trial  (not intended for publication) was designed to explore two goals in order to see if (and how) the outcomes might be worthwhile in a larger, more conventional study.

(Goal 1) Determine if hsCRP — an accepted and well-regarded clinical blood human health indicator of inflammation — could be a direct human marker for levels of BPA, phthalates and other environmental chemicals.

Most importantly, could reducing BPA levels “move the needle” on C-Reactive Protein?

Outcomes: The sourcing, preparation and serving protocols used in the intervention phase reduced inflammation. Unfortunately, the BPA blood samples have been mishandled by a UCSF research lab making it unlikely that accurate results will be known..

However, the BPA blood sample data is likely irrelevant in this case. That is due to a number of unexpected anomalies that revealed substantial potential confounding factors that would require protocol changes that would be necessary to make a larger trial replicable.

(Goal 2) Determine if it was possible to increase the replicability of dietary intervention trials by developing methods to structure and assess the difficulty of bringing the discipline of standard laboratory practices to the sourcing, preparation, and serving of human food. A wide variety of tasks and goals (still in draft mode) can be found in the links at: The Study – Deconstructed (right-hand column)

Outcomes: Useful information and techniques were successfully developed that could improve the replicability of dietary intervention trials.

Sourcing foods with transparency and an eye to availability for replicability was successful but required extraordinary adaptation and more efforts than anticipated. See this link for a small sampling that we have only begun to describe.

With staff training, laboratory best practices were efficiently adapted to food preparation activities.


The lack of direct human data on the health effects of low-level, plastic-derived environmental chemicals such as Bisphenol A (BPA) has contributed to an internationally divisive scientific controversy that has prevented health professionals and consumers from making scientifically valid health decisions.

On one side of the controversy are university and independent scientists who contend that hundreds of published studies prove that micro- and pico-molar concentrations of BPA are unhealthy. On the other side, corporate and some federal regulatory scientists point to the recent CLARITY study whose results contend that low levels are safe.

CLARITY and most other published studies have been based upon an in vivo murine model. As a result, the current debate centers on protocol flaws, confounding factors, sources of contamination and the finer points of how the murine studies were conducted.

All of the arguments on both sides are unlikely to sway opinion one way or another for four reasons:

(1) There are no direct human studies measuring health effects.

(2) The murine model results almost always fail to translate accurately to humans. This is frequently the case in drug trials where up to 92% have been shown to fail in humans following successful murine results.

(3) The human dietary intervention studies published so far have focused only on reducing subject levels of marker chemicals such as BPA and have not measured any direct health outcome indicators.

(4) The few human dietary intervention study results published so far have failed to show consistent results. Much of that can be attributed to the fact that human dietary intervention studies frequently fail to apply fundamental laboratory standards to food sourcing, preparation, and serving, and make little or no effort to minimize confounding factors such as air quality and other environmental contamination.

A possible solution waiting in clinical blood tests?

Significantly, a source of ethical, respected and clinically valuable direct human data may lurk in standard laboratory blood profiles.

One potential blood profile candidate is High-Sensitivity C-Reactive Protein (hsCRP) which offers valuable clinical insights into numerous inflammation-linked conditions including cardiovascular disease, Type 2 Diabetes, cancer, Alzheimer’s Disease, depression, suicide, and auto-immune diseases including IBD, rheumatoid arthritis, and lupus.

Further, hsCRP is a likely candidate for assessing the human health effects of Bisphenol A because that chemical (and its analogs) has been clinically shown to increase inflammation and also has been credibly associated with those conditions.

Given that the use of clinical blood tests as possible direct human health effects indicators of environmental chemical contamination is an unknown field, investigators felt that a small (N=1) proof-of-concept trial was warranted to test the validity of the concept and protocols before beginning a larger, far more expensive study.


The main objective of this study was to determine if measurable changes in hsCRP could serve as a direct human health effects marker for Bisphenol A (BPA) and other plastic-chemical based contamination in a brief (six-day) dietary intervention trial.

Because BPA-related human dietary intervention studies have been shown to be non-replicable, a second objective of this study was to enforce strict laboratory standard protocols for food sourcing, preparation, and serving in order to enable the efficient replication of this and of future studies.


This n=1, proof-of-concept, six-day study was approved by the Committee on Human Research/IRB at the University of California San Francisco School of Medicine (UCSF). The intervention consisted of a three-day “typical” American diet with known sources of plastic contamination followed by a three-day intervention in which food sourcing, preparation, and serving were subjected to extreme scrutiny and methods to reduce BPA contamination from source to the subject’s mouth. An extensive protocol for this strict regime was also approved by CHR/UCSF.

Two blood samples — one to determine hsCRP levels and the second to determine BPA concentrations — were taken each day at baseline, the second on the morning of the “typical” menu leg and the third on the morning after the extreme decontamination intervention leg. All blood draws were taken at Sonoma Valley Hospital (an affiliate of UCSF) in light-green-topped tubes with standard concentrations of Li-heparin. The hsCRP sample was centrifuged by SVH and the test was run by the UCSF Parnassus laboratory.

The samples for BPA analysis were hand-delivered as whole blood to a UCSF lab where they were centrifuged and frozen for LC-MS/MS analysis. All samples were drawn using a vacutainer kit supplied by the lab and composed of polymers known not to leach BPA or other contaminants into the blood samples. As directed by the lab, samples were hand-delivered within four hours of blood draws. Results have been delayed because of an administrative complication with the lab.

The menu for the two legs of the study was as close to identical as possible. Food items for the “typical” leg of the study were selected because they appeared on (a) the USDA’s list of frequently consumed items, (b) national food frequency consumption surveys and (c) they were readily available as top-selling items in supermarket chains.

Choosing popular national-branded items that are widely available in supermarket chains makes replication easier because the items are easier for future investigators to obtain and are likely to have consistent compositions due to their rigidly controlled processes for sourcing and production.

Once the menu was set for the “typical” leg, we then “reverse-engineered” the typical food choices to use in the intervention menu by substituting identical or highly similar items which were minimally processed, organic and (when possible) were not packed in plastic.

When plastic was inevitable, we developed and employed a set of protocols designed to greatly reduce or eliminate contamination from plastic packaging.

Foods for the intervention study were also selected for transparency of the producer and because they were either available in national distribution or came from a source who would ship nationwide. Dairy posed a unique problem because milk is sourced locally or regionally even by large chains. For that reason, we had milk samples analyzed via LC-MS/MS which offers future investigators the advantage of choosing milk with the same or similar BPA quantification.

Beyond sourcing foods with transparency and replication, we extended the discipline to preparation and serving.

Recipes were considered formulae and cooking methods and times as experimental protocols.

All main ingredients were measured to the nearest gram. Spices and other items were measured to the nearest tenth of a gram. Culinary assistants were trained and monitored for compliance.

Items with known plastic contamination (such as micro-plastic contamination of table salt) were replaced where possible with reagents from Sigma-Aldrich.

A 400-square-foot professional kitchen with stainless steel countertops and carbon-filtered water was cleared of all plastic and polymer items utensils, containers, and cookware. Only glass, stainless steel, aluminum foil, and maple cutting boards were allowed. Vinyl gloves were used in food preparation for sanitary control in both legs of the trial, but actual food contact with the gloves was minimal and incidental.


Levels of hsCRP decreased 21.4% from baseline to end of the minimally contaminated leg: 1.1 mg/L from 1.4 mg/L. An anomaly occurred when hsCRP unexpectedly declined at the end of the “typical” diet leg.

Significantly, a major NIH-funded dietary intervention by Hall, et. al. published in 2019 exhibited this same anomaly.

NIH/Hall data from: “Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain: An Inpatient Randomized Controlled Trial of Ad Libitum Food Intake”


As noted in the results, above, inflammation levels are measurably affected by a dietary intervention designed to reduce levels of BPA and phthalates and are worth further study. Serum BPA may be a valid direct health indicator of human BPA contamination.

That decrease may be related to the reduction of food additives or reductions in BPA. Because food additives are most prevalent in ultra-processed foods, the outcomes measured are most logically due to both BPA and food additives — either together or separately.

Significantly, moving BPA causality forward without confounding factors will probably require another proof-of-concept investigation — something that is discussed at the end of the following section.

Possibilities for further investigation

Further insights may be forthcoming from the analysis of serum BPA. That analysis has been delayed because of an administrative complication with the UCSF lab tasked with the analysis.

In addition, this n=1, proof-of-concept investigation has raised multiple confounding factors that need to be addressed in a larger trial. Among those are the need for stringent safeguards to avoid confounders such as air pollution, dust, transdermal absorption, and other factors,

In addition, this n=1 trial points out two further factors to make a larger trial better able to more accurately determine causality between human clinical health indicators and BPA and other plastic-derived chemicals.

Markers for ultra-processed food additives

The first factor would be to determine a blood marker associated with one or more of the most common additives found in ultra-processed foods.

The second would be to include hsCRP along with other indicators in an expanded diagnostic blood panel.

Expanded blood profiling

Because high inflammation is associated with many metabolic health problems, this expanded blood panel would include additional indicators such as ghrelin, adiponectin, and insulin resistance.

Because hsCRP is non-specific to the sources of inflammation, further accuracy in human causality determinations may also be enhanced by the employment of select cytokines and chemokines to determine direct BPA influences on inflammation sources such as immune system disorders and tumorigenesis.

Moving BPA causality forward will probably require another proof-of-concept investigation: Controlled dosing

If alternatives 1 and 2 are not satisfactory — or are likely to prove inadequate in determining human health effects specifically as relates to BPA — a third method would be to conduct a dietary intervention with exactly the same menu in the typical diet and BPA dosing in the intervention diet.

All ingredients would be pre-tested for BPA levels at the sourcing stage and alternatives selected if levels are determined to be above an acceptable threshold.

Subsequent BPA levels would be established after meal preparation and a total daily BPA dosage calculated.

Then, prepared (and tested) meals would be divided into two segments: one for the meals as prepared and a second — intervention menu –to be dosed with BPA at the federal acceptable “safe” level.

This regime would isolate BPA as the only contributor to the inflammation and other clinically relevant manifestations.

With a sufficient number of markers measured — similar or identical to those on the NIH/Hall study — it may be possible to develop a profile of markers that is specific to BPA.

This study protocol would be — by far — the most expensive alternative. That means that a proof-of-concept investigation will probably be wise to determine any further confounding factors.

Partial list of relevant links