Lead Service Lines and Drinking Water Infrastructure

What the evidence says

The clearest individual-level evidence on lead service lines (LSLs) comes from Milwaukee, where 54.1% of residences (653 of 1,207) had LSLs. After adjusting for covariates, living in a residence with an LSL was associated with higher infant blood lead levels: an odds ratio of 1.60 (95% CI 1.04–2.46) for reaching a blood lead level (BLL) of 3.5 μg/dL or greater, and a 0.13 μg/dL increase (95% CI 0.04–0.22) in continuous log-transformed BLL. This is the only claim here that quantifies an individual-level dose-response between service line presence and blood lead.

A separate body of evidence describes the corrosion mechanism that puts lead into water in the first place. Switching a city's water source to corrosive, untreated water caused drinking water pipes to corrode and leach lead. In Washington, DC, switching water treatment chemistry from chlorine to chloramine in 2001 increased water corrosivity and caused lead to leach from aging pipes. Pressure fluctuations and system flushing were described as disturbing lead-bearing pipes and mobilizing lead particles, and elevated water temperature was described as increasing lead's solubility, exacerbating corrosion. None of these four mechanism claims carries a reported magnitude — all are directional only. The reverse process was also observed: after a city switched back to treated Lake Huron water, water lead levels and blood lead levels began to decline, again without a reported magnitude.

At the geographic level, lead pipe density emerged as the strongest predictor of blood lead levels across DC wards in a cross-validated model (AUC 0.90–0.95), and housing and schools built before the 1986 US ban on lead plumbing were both described as carrying elevated contamination risk, though neither claim gives a magnitude. In direct testing of DC schools, 60 schools in one testing round ("Draw 1") and 21 in a second round ("Draw 2") had lead concentrations above 5 ppb, and 14 (Draw 1) and 9 (Draw 2) schools exceeded the EPA's 15 ppb action level. A model-based estimate put 11% of DC's total area in a "very high" lead contamination risk zone, 13.88% in "high," and 38% in "very low." Within schools, the type of device used to collect the water sample was reported as the single most influential predictor of measured lead levels, ahead of building age or other factors.

Outside the US cases, cross-sectional sampling in Pakistan found lead above permissible limits in Punjab groundwater, "high" lead contamination in Peshawar drinking water (urban and rural), and lead above WHO limits in Swabi district drinking water — none of these three claims gives a numeric concentration or share of samples affected, only that limits were exceeded. A broader claim states that heavy metals including lead in groundwater and surface sources contribute to carcinogenic and non-carcinogenic health risk, again without a magnitude.

The evidence base also covers non-lead water-infrastructure risk that is directionally relevant to aging-pipe policy: intermittent water supply is associated with higher fecal contamination and increased diarrhoeal risk, and maintaining positive pressure in water mains, physical separation from sewer lines, rapid leak detection, and systematic replacement of aging pipes is described as preventing fecal contamination ingress during low-pressure events. No source here quantifies the effect of an actual LSL replacement program (as opposed to LSL presence/absence) on blood lead levels, no source gives cost figures for replacement, and no source reports a magnitude for how much lead concentrations change under any of the corrosivity mechanisms (source switch, chloramine, pressure events, temperature) — the evidence establishes direction for these mechanisms but not size.

Where the evidence splits

No contradictions in the current sources.

Thresholds and dose-response

DC school testing used 5 ppb as the exceedance threshold reported in results — the source notes this itself exceeds the American Academy of Pediatrics' recommendation of 1 ppb. Under that 5 ppb threshold, 60 schools (Draw 1) and 21 schools (Draw 2) tested above it.

A second, higher threshold — the US EPA's action level of 15 ppb — was exceeded by 14 schools (Draw 1) and 9 schools (Draw 2), a smaller subset than exceeded the 5 ppb mark.

For blood lead, the individual-level threshold used was a BLL of 3.5 μg/dL or greater, at which residence in an LSL home carried an odds ratio of 1.60 (95% CI 1.04–2.46).

Open empirical questions

• No claim here measures the effect of an actual lead service line replacement program (before/after) on blood lead levels — only the association between LSL presence and blood lead.
• No source quantifies the magnitude of lead-concentration change from the corrosivity mechanisms identified (source switching, chloramine treatment, pressure events, temperature) — only direction.
• No source gives a cost figure, timeline, or completion-rate for infrastructure replacement.
• The disproportionate exposure of in-utero and formula-fed infants to lead-in-water is described qualitatively; no source quantifies how much greater that exposure or resulting harm is relative to other age groups.