11.4 Biomonitoring, Internal Dose, Fate/Transport & BCF

Internal dose is not the bottle concentration

Applied dose (what was in the chamber or the gavage needle) and potential dose (what contacted the skin or the GI lumen) are not internal dose (what reached blood or the target). Handbook III.B.2–4 asks you to reconstruct internal dose from biomarkers, to know which environmental fate processes change the concentration that ever reaches a receptor, and to use bioaccumulation metrics such as a bioconcentration factor (BCF) without confusing them with biomagnification. Independent OpenExamPrep teaching in this section covers those tools at a conceptual level. NHANES as a national public-health database is developed in the applied-toxicology chapter on environmental and public-health practice; here you only need to know how a urine or blood number becomes a dose. This material is not an ABT product and does not claim official approval, review, or partnership with ABT, CDC, OECD, or EPA.

Biomarkers of exposure, effect, and susceptibility

Biomarker of exposure. A chemical, metabolite, or adduct measured in a biological matrix that indicates contact: blood lead, urinary trichloroacetic acid after trichloroethylene, urinary cotinine after nicotine, hemoglobin adducts of ethylene oxide. These close the pathway in a CSM: if the metabolite is in urine, a complete path existed, even if the air model was uncertain. They still require the parent versus metabolite and unbound versus total discipline from section 11.1. A urinary metabolite is not parent air concentration.

Biomarker of effect. A measurable alteration in a biochemical, physiologic, or other process: red-cell acetylcholinesterase inhibition after an organophosphate, ALA dehydratase inhibition after lead, DNA strand breaks, heme-synthesis changes. Effect biomarkers sit closer to hazard and clinical toxicology. Using an effect biomarker as if it were a kilogram-per-day intake, without a calibration, mixes III.B with III.A.

Biomarker of susceptibility. A host factor that changes internal dose or response for the same applied exposure: GSTM1 deletion, PON1 status for some organophosphates, ALDH2*2 for acetaldehyde, age-related creatinine excretion, pregnancy-related plasma volume. Susceptibility biomarkers do not prove that exposure occurred; they change how a given exposure is read.

Adducts can straddle exposure and effect: a hemoglobin adduct documents that reactive parent reached blood, and it is also a molecular alteration. On the examination, name the question the marker is being asked to answer.

From urine and blood to dose: creatinine and reverse dosimetry

Spot urine concentrations swing with hydration. Creatinine adjustment (µg/g creatinine) reduces that noise because creatinine excretion is more stable than water excretion in a given person, though it still varies with muscle mass, age, sex, and renal function. Specific gravity adjustment is the cousin used when creatinine is a poor denominator (very low muscle mass). 24-hour urine is the gold-volume method and is rare in population surveys.

A simple mass-balance reverse dosimetry (steady state, one-compartment teaching form) is:

Estimated daily intake ≈ (C_urine × V_urine) / F_excreted

or, with creatinine: (C_µg/g × creatinine excretion in g/day) / F_excreted, where F_excreted is the fraction of parent intake that appears as that analyte in urine. If F is 0.5, the urine amount is half the intake, so intake is twice the excreted mass. If you measured a metabolite, F must be the molar yield of that metabolite, not 1.0 by default. Blood reverse dosimetry usually needs clearance or a PBPK model: concentration = intake × bioavailability / clearance at steady state, rearranged for intake. Treating a µg/L blood number as if it were already an RfD in mg/kg-day skips every term in that rearrangement.

NHANES, run by CDC, is the later-chapter survey that reports nationally representative blood and urine concentrations for many environmental chemicals. Use it here only as the reason those creatinine-adjusted tables exist: they are biomarker distributions, not ready-made ADDs. Converting an NHANES 95th percentile urine value into a population intake is reverse dosimetry plus survey weights, not a lookup of the RfD.

Fate and transport: four words that are not synonyms

Persistence. How long the parent remains in a medium (hydrolysis, photolysis, biodegradation half-lives). A short air half-life can coexist with a long sediment half-life.

Bioavailability. The fraction of contacted mass that is absorbed. Soil lead phosphate is less bioavailable than lead acetate in water; a total-soil milligram is not an absorbed milligram.

Bioaccumulation. Net uptake into an organism from all routes (water, diet, sediment, air) relative to the surrounding medium. Bioconcentration is the water-only subset (dissolved chemical across the gill, in the OECD 305 aqueous design).

Biomagnification. Concentration increases with trophic level in a food web (classic organochlorines in piscivorous birds). High BCF does not automatically prove biomagnification if the chemical is metabolized in the predator or is not transferred in diet.

log Kow (octanol–water partition coefficient) screens hydrophobicity. Very low log Kow: stays in water, little lipid storage. Mid-range (often cited around 3–6): bioaccumulation concern. Very high log Kow (often >6): uptake can fall because of reduced bioavailability, size, or metabolism—the “superhydrophobic” caveat, not a safety certificate.

Koc (organic-carbon–water partition coefficient) describes sorption to soil and sediment organic carbon. Kd = Koc × foc. High Koc means the chemical rides organic carbon: less in the dissolved water column, more in sediment and in soil that children ingest. Dissolved-water BCFs can look modest while benthic receptors still see a sediment path.

BCF, OECD 305, and sentinels

BCF = C_organism / C_water at steady state, typically L/kg on a whole-fish wet-weight basis. OECD Test Guideline 305 is the conceptual fish test: aqueous exposure to measure BCF (uptake rate k1 and depuration k2, with BCF ≈ k1/k2), and a dietary exposure design for poorly water-soluble substances that reports a biomagnification factor (BMF) from spiked feed. It is not a mammalian TK study and not a human biomonitoring protocol.

Regulatory flags differ by program; do not memorize one number as universal. Many TSCA-style PBT discussions treat BCF/BAF of 1,000 as bioaccumulative and 5,000 as very bioaccumulative. REACH Annex XIII uses 2,000 (B) and 5,000 (vB). Read the statute you are actually in. A BCF is still not a human ADD.

Sentinel species are organisms whose tissue or health signals contamination for a media or food web: historically canaries in mines; mussels in coastal monitoring; piscivorous birds for persistent organochlorines; lichens for atmospheric metals. A sentinel documents that a pathway is complete in the field. It does not replace a human CSM, and a clean sentinel in the wrong habitat does not clear a different receptor.

Monte Carlo / probabilistic exposure (method, not software)

Point-estimate ADD uses one C, one IR, one BW. Probabilistic exposure draws those inputs from distributions (for example lognormal concentration, normal body weight, triangular or empirical intake rate), computes ADD thousands of times, and reports a distribution of dose—median, mean, 95th percentile—plus sensitivity (which input drives the variance). That is Monte Carlo sampling. It does not require naming a commercial package. It does not fix a wrong CSM: sampling an incomplete pathway still yields a precise estimate of a dose that does not occur. It is also not a substitute for a PBPK reverse-dosimetry model; it is a way to propagate variability and uncertainty through the same equation taught in section 11.2. EPA RAGS Volume III is the Superfund-language cousin of this idea. On the examination, know what the output is (a dose distribution) and what it is not (proof that the 95th percentile equals the RME point estimate, or a license to ignore correlation between body weight and intake).

Scenario

A community has blood lead measurements, soil lead, and a fish advisory. Blood lead is a biomarker of exposure (and, at high levels, sits next to effect markers such as ALA-D). Converting the geometric mean blood lead to an incremental soil intake requires a biokinetic model (IEUBK-style thinking), not dividing µg/dL by 70 kg. Urinary creatinine-adjusted cadmium in NHANES tables is not an RfD; reverse dosimetry would need excretion fraction and an intake reconstruction, then comparison with a toxicity value in matching units. Fish-tissue concentrations in a piscivorous bird are a sentinel/food-web signal of biomagnification for a persistent organochlorine; the human subsistence-fisher ADD still needs a consumption rate × fillet concentration, which is dietary exposure, not a BCF. Running Monte Carlo on IR and BW while leaving C as a single maximum detection in an unused well produces a tight distribution of a pathway the CSM already ruled incomplete.

Traps

• Calling AChE inhibition a biomarker of exposure, or GSTM1 genotype proof that exposure occurred.
• Treating µg/g creatinine as mg/kg-day without reverse dosimetry.
• Equating BCF with biomagnification, or OECD 305 with a human urine study.
• Using log Kow as if it were Koc, or dissolved BCF as if sediment were irrelevant.
• Treating a Monte Carlo 95th percentile as automatically equal to a RAGS RME point estimate, or as a reason to skip the CSM.