Ferguson and colleagues analyzed soil arsenic measurements from two Miami baseball parks investigated in 2014, then modeled exposure for hypothetical visitors and players. Soil concentrations differed by depth and field location. The health-risk results are scenario calculations, with several internal reporting inconsistencies; no children’s biomarkers or health outcomes were measured.
Key numbers
Source-reported surface-soil summaries in mg/kg distinguish mixed sampling depths from equivalent 0–15 cm composites:
| Soil grouping | Chapman | Colonial |
|---|---|---|
| Infield, mixed surface depths (Table 2) | 4.8; n = 4 | Not sampled |
| Outfield, mixed surface depths (Table 2) | 94.6; n = 21 | 47.9; n = 40 |
| Adjacent areas, mixed surface depths (Table 2) | 33.0; n = 36 | 12.0; n = 13 |
| Mean, equivalent 0–15 cm across areas (Table 3) | 43.9; n = 60 | 40.7; n = 53 |
| 95% upper confidence limit of that mean | 63.4 | 47.7 |
| Highest equivalent 0–15 cm concentration | 118 | 120 |
| Highest single surface concentration | 262 | 120 |
Chapman’s 262 mg/kg maximum came from a 0–5 cm sample. The 95% upper confidence limits describe uncertainty in means; they are not 95th-percentile sample concentrations. Table groups and mapped measurements overlap and cannot be added to obtain a total independent sample count.
The authors report chronic hazard quotients of 2.37 for a hypothetical 1–2-year-old under the highest-concentration scenario and 9.69 for a 2–3-year-old pica scenario (Table 12). These are source-model outputs, not measured disease rates, individual diagnoses or validated predictions for present-day park users.
Methods (brief)
Sampling occurred from April through September 2014 during an adaptive investigation of Chapman and Colonial fields. Certified laboratories analyzed soils, but this paper does not specify the arsenic digestion method, moisture basis, detection limits or species fractions. Some depth-paired measurements were combined into equivalent 0–15 cm concentrations. The article supplies both grouped results and labeled surface-sample maps.
Exposure scenarios assume soil ingestion, dermal contact and inhalation, using literature-derived activity and absorption factors. Typical attendance was modeled as two hours per day on 169 days per year; some cancer scenarios extend across a 78-year lifetime. The soil arsenic measurements were not speciated, although the toxicity calculations use inorganic-arsenic assumptions.
Evidence fitness
The study supports qualified environmental comparisons among sampled depths and field areas. It also illustrates how chosen exposure assumptions affect model outputs. The suggested origins—contaminated fill or historical herbicides—were not confirmed by source apportionment. The study contains no baby-wipe, cereal or fish measurements.
Limitations
Targeted sampling intensified after contamination was located, so the measurements do not represent a random survey of all park soils. Colonial lacked infield samples. Repeated map labels and a discrepancy between an infield count and its note limit reconstruction of individual sample identities.
Several model tables contain conflicting exponents, aggregates smaller than their component doses, inconsistent cumulative values, or mismatched field/concentration headers. Oral reference values were applied to aggregate exposure across routes. Assumed lifetime attendance, constant contamination and generic activity patterns were not verified for actual individuals. These limitations prevent treating the paper’s risk calculations or reassurance as validated health or safe-play guidance. Regulatory comparisons are secondary references from the paper, not current primary-instrument determinations.
Related evidence
Arsenic · Source attribution and environmental burden apportionment · Cat 2 (Children Personal Care) non-ingestion exposure pathways
Update history
No substantive edit history is available in this build. The full commit record is available in git.