Overview
All three reviewers at Impact 5. No meaningful scoring disagreement remained.
The unanimous final score reflects meaningful exploratory importance tempered by moderate limitations in prediction coverage, informative data loss, and demonstration of added predictive value, together with the modest safety-monitoring concern. These weaknesses lessen expected influence without establishing that the work cannot succeed.
-
01
Bound the proposed mechanism to what the study can test
Without skin-perfusion measurement, an association between RASL and whole-body strain would not establish the proposed glandular oxygen-delivery mechanism. Research Strategy A.3, p. 4, following the discussion of the forearm occlusion model
-
02
Remove the unsupported prevention claim
The Specific Aims say this observational, no-alert R21 will produce a sensor that prevents heat stroke, although prevention effects are reserved for a future trial. Specific Aims, Expected outcomes and impact, p. 3
-
03
Resolve prediction coverage and refitting after patch replacement
A new 60-minute RASL fit for each patch excludes at least two hours per eight-hour shift, including time after midday replacement. Refitting may absorb developing strain, and an early narrow sweat-rate range may create apparent later trends. Research Strategy C.5, Predictor, p. 7
-
04
Justify event-level precision and the incremental-value comparison
The proposal gives no event-count or precision justification for an AUC increment of at least 0.05. Its comparator omits the cited heart-rate-based core-temperature estimate, while heart rate contributes to the PSI outcome. Research Strategy C.5, Analysis, p. 7
This is the short list. The panel filed three independent reviews, argued across three rounds, and produced four groups of recommended revisions — including items not shown here. Each finding is tied to the section it came from, so you can check the claim against your own document.
Document intelligence
Detected documents:
- Cover/title block (p. 1)
- Project Summary/Abstract and Project Narrative (p. 2)
- Specific Aims (p. 3)
- Research Strategy: Significance, Innovation, Approach, timeline (pp. 4–8)
- Protection of Human Subjects, including Data and Safety Monitoring (pp. 9–10)
- Inclusion of Women and Minorities, with planned enrollment table (p. 10)
- Inclusion of Individuals Across the Lifespan (p. 10)
- Bibliography and References Cited (p. 11)
- Facilities and Other Resources (p. 12)
- Data Management and Sharing Plan (p. 13)
- Budget Justification (Modular) (p. 14)
- Authentication of Key Biological and/or Chemical Resources (p. 14)
- Resource Sharing Plan (p. 14)
- Letter of Support, Marlow Creek Civil and Line Services (p. 15)
Conditional requirements:
- DMS Plan (Section IV.2, "required for any NIH-funded... research that will generate scientific data"): present.
- Resource Sharing Plan (Section IV.2, PHS 398 Research Plan): present.
- Human subjects study record (Section IV.2, required if "Are Human Subjects Involved?" is Yes): not supplied for review. A Protection of Human Subjects narrative is present.
- Inclusion of Women, Minorities and Across the Lifespan (Section V, human subjects inclusion criteria): present.
- Clinical trial status (Section II, "Clinical Trial? Not Allowed"): present. The applicant states the study is not an NIH-defined clinical trial (p. 9).
- Authentication of Key Biological/Chemical Resources (Section V, Additional Review Considerations): present.
- Vertebrate Animals (Section V): not applicable.
- Foreign subawards (Section III.1): not applicable. None are indicated.
- SF424 forms and Senior/Key Person profiles (Section IV.2): not supplied for review.
Document quality:
- Figures 1–3 (pp. 5–6) appear as captions only. No graphics are visible in the page images. Captions 2–3 describe pilot data whose key values are also given in the text.
- Minor ligature artifact on p. 13: "af ecting".
- Internal figures are consistent. The enrollment table sums to 40, with 8 women and 22 Hispanic or Latino participants (55%), matching the text. The letter's 2028–2029 field seasons align with the timeline.
Author questions: None.
Compliance & Alignment
Status: Compliant with concerns
Verified issues:
- DMS Plan (Section IV.2): "Where the DMS Plan Format Page requires a 'Yes or No' response, no additional narrative is allowed." Element 5(B) (p. 13) reads "Controlled access. Yes." and then adds narrative ("Access requires an approved data use agreement…"). The packet does not show whether this field is a Yes/No field on the Format Page, so this is a concern only.
Verified compliant:
- Specific Aims: 1 page (p. 3), within 1.
- Research Strategy: 5 pages (pp. 4–8), within the R21 limit of 6.
- Project Summary is under 30 lines; Narrative is 3 sentences.
- Budget: $150,000 and $125,000, total $275,000. This meets the "may not exceed $275,000" and "no more than $200,000" in any year limits.
- Project period: 2 years, within 2.
- Application type "New" is allowed.
- Clinical trials not allowed: the narrative describes an observational design (p. 9).
- DMS Plan and Resource Sharing Plan are present.
Not verifiable from supplied documents:
- SF424 forms, including the human subjects answer, application type and budget form entries.
- Human subjects study record, required if Human Subjects = Yes, including the formal clinical-trial questionnaire.
- Senior/Key Person profiles, eRA Commons IDs, ORCID linkage, and organizational registrations (SAM, UEI, Grants.gov).
- Due date and on-time submission.
- Whether the DMS Plan used the Format Page.
- Appendix contents.
- Foreign subawards, which can only be confirmed from budget forms.
- Font size, margins and line density.
- Signature on the letter of support.
Not audited (no stated rule):
- Typeface.
- Length of Bibliography, Facilities, Protection of Human Subjects, Inclusion sections, Budget Justification and Letters of Support.
- Number of letters.
Notes: Audited against R21 (PA-25-304) limits. Figures 1–3 render as captions without graphics. No rule addresses this, but the applicant should confirm that the submitted PDF includes them.
Alignment: Strong
Solicitation-specific rubric:
| ID | Type | Weight | Requirement | Source |
|---|---|---|---|---|
| SC-01 | eligibility | mandatory | Direct costs may not exceed $275,000 combined; no more than $200,000 in any single year | Sec. II, Award Budget |
| SC-02 | eligibility | mandatory | Total project period may not exceed 2 years | Sec. II, Award Project Period |
| SC-03 | eligibility | mandatory | No clinical trials accepted | Part 1 Purpose; Sec. II, Clinical Trial? |
| SC-04 | program-priority | scored | Exploratory, novel, higher-risk work, distinct from R01-type or well-established projects | Sec. I |
| SC-05 | program-priority | scored | Fits a participating IC's mission; program-priority relevance considered at funding | Part 1 Components; Sec. V.2 |
| SC-06 | document | mandatory | DMS Plan on NIH Format Page addressing all elements; no narrative after Yes/No responses | Sec. IV.2; Sec. VI.3 |
| SC-07 | document | mandatory | Resource Sharing Plan per Application Guide | Sec. IV.2 |
| SC-08 | document | mandatory | At least one human subjects study record when Human Subjects = Yes | Sec. IV.2 |
| SC-09 | document | mandatory | Appendix limited to blank questionnaires or surveys | Sec. IV.2 |
| SC-10 | eligibility | mandatory | No foreign subawards or subcontracts | Sec. III.1 |
| SC-11 | eligibility | mandatory | SAM, eRA Commons, Grants.gov and ORCID registrations completed | Sec. III.1; IV.7 |
| SC-12 | eligibility | mandatory | Activities align with applicable law; race not used as a participation selection criterion | Sec. VI.2 |
Proposal mapping:
| ID | Status | Evidence |
|---|---|---|
| SC-01 | addressed | $150,000 + $125,000 = $275,000 (p1, p14) |
| SC-02 | addressed | 2-year period; timeline months 1–24 (p1, p8) |
| SC-03 | addressed | Observational design; non-trial determination with rationale (p9) |
| SC-04 | addressed | A.4 risk rationale; RASL reframing; Innovation section (p4–5) |
| SC-05 | addressed | NIEHS environmental-exposure fit stated (p4); actual IC assignment is not verifiable |
| SC-06 | partial | Elements 1–6 covered (p13); Element 5B adds narrative after "Yes" |
| SC-07 | addressed | No-cost academic license for LTP-2 design files, protocols and firmware (p14) |
| SC-08 | not-verifiable | PHS Human Subjects form evidence not supplied for review |
| SC-09 | not-verifiable | Appendix not supplied for review |
| SC-10 | not-verifiable | Modular budget lists no subawards (p14); subaward forms not supplied |
| SC-11 | not-verifiable | Registration and cover-sheet evidence not supplied for review |
| SC-12 | addressed | Race and ethnicity not used for selection (p10) |
Rubric caveats:
- The call follows the Simplified Review Framework: Factors 1–3 plus standard additional review criteria and considerations, none weighted beyond NIH defaults. These are not repeated as rows.
- The call states one R21-specific emphasis. Reviewers stress the conceptual framework, innovation and potential to advance knowledge, and preliminary data are not required.
- The call states no rating scale of its own.
- On timing, the support letter commits the partner to the summers of 2028 and 2029. Combined with the December 1 start, this implies a February 2027 due date (earliest start December 2027).
- The Aims page claims the sensor "will prevent exertional heat stroke." This overstates what an observational, no-alert study can show, and reviewers may note it under Factor 1.
Read: the budget justification and its budget table, 2 years (Proposal, p. 14); no separate budget form was supplied, so the figures in that table are checked. Figures are copied from the documents; every sum and comparison is computed, never estimated.
Not verifiable (1)
- The budget's periods are within the maximum durationThe budget has 2 periods, but their dates or lengths are not stated legibly; the call allows 24 months.
Consistent (11) arithmetic by year 2 · solicitation and funder 3 · narrative and justification 6
- The budget lines sum to the total direct costs Years 1 and 2
- The direct costs requested are within the award ceiling
- Within the yearly ceiling Years 1 and 2
Months match the text:
- Tomas Reinholt, PhD
- Mateo Velarde-Kim, PhD
- Priya Ostrander, PhD
- Field research coordinator, to be named
- Research technician, to be named
- Data manager, to be named
Technical analysis
Key technical claims
Pilot lactate accuracy: CCC 0.87, bias −0.6 mM, limits of agreement (LoA) −3.9 to 2.7 mM, n = 131 pairs (C.2, p.5). Evidence: mirror-site pads, laboratory assay, 6 workers, 22 shifts [14]. Partly supported. The numbers agree with each other: (2.7 + 3.9)/2 = 3.3 mM, so SD ≈ 3.3/1.96 ≈ 1.68 mM. Nothing says the pilot LoA were adjusted for clustering (131/22 ≈ 6 pairs per shift), so they are probably too narrow. A CCC computed across a 4 to 32 mM between-person range flatters the within-person tracking that RASL depends on. The Aim 1 benchmark (CCC ≥ 0.85) leaves a margin of only 0.02.
Usable signal for 86% of patch-hours; 3 of 44 patches lost adhesion (p.3, p.5). Supported as feasibility. The margin over the ≥85% benchmark is 1 point, from 22 shifts.
Bench compensation: drift 3.1%/h; pH error reduced from 11% to 3% (p.5, [13]). Partly supported. Tested in artificial sweat at a fixed 34 °C, reported in the team's own paper, with no field drift data. Over the 20-min window, drift ≈ 3.1 × 0.33 ≈ 1%, about 0.01 ln units, which is negligible against the 0.15 threshold. Biofouling in the field is untested.
Glandular origin, supported by blood vs sweat lactate r = 0.08 (p.5). Questionable as evidence. Blood lactate spanned only 1.1 to 2.6 mM, and range restriction alone predicts r ≈ 0. The claim really rests on refs [6,10].
RASL rose >0.15 log units before 7 of 9 episodes, median lead 24 min; false rises on 5 of 15 non-episode shifts (p.5). Questionable. The threshold was derived from these same unblinded data (C.5, "taken from the pilot"), and the 9 episodes come from ≤6 workers. Wilson 95% CI for 7/9 is 0.45 to 0.94; for 5/15 it is 0.15 to 0.58. Sensitivity is counted per episode but false rises per shift, and false alerts on episode shifts are not reported. The proposal admits the pilot cannot estimate accuracy (p.6).
Mechanism: RASL marks glandular hypoxia from constrained skin perfusion (p.3, A.3). Unsupported in this setting. It is inferred from a forearm occlusion model [9] and a perfusion plateau [8]. Skin blood flow is never measured, so a positive Aim 2 result cannot separate this mechanism from sweat-rate or skin-temperature artefacts.
Aim 2 benchmarks (AUC ≥ 0.75, AUC gain ≥ 0.05) are testable with 320 shifts (p.3, C.5). Unsupported. There is no power or precision calculation. At the pilot rate, 9/22 ≈ 0.41 episodes per shift, Season 1 gives about 160 × 0.41 ≈ 65 clustered episodes in 20 workers. An AUC gain of 0.05 over a model that already contains heart rate, which is part of PSI, is unlikely to be estimated precisely.
"Will produce a wearable sensor that will prevent exertional heat stroke" (Specific Aims, p.3). Unsupported. It contradicts C.9: "This R21 does not deliver alerts or test their effect."
Figures and tables
- Figs 1 to 3 (pp.5 to 6): the page images show the captions only. No graphic is rendered, so axes, units, data points and panels cannot be checked.
- Fig 1: illustrative schematic.
- Fig 2 (Bland–Altman, 131 pairs): the "no visible trend" claim cannot be verified. Nothing shows points marked by worker, LoA adjusted for repeated measures, or proportional error.
- Fig 3: one selected shift showing a 27-min lead. This is a best-case illustration, not evidence of accuracy, and it is not stated whether this shift helped set the threshold.
- Timeline table (p.8): the arithmetic is consistent (20 × 8 = 160 shifts; 20 × 3 = 60 reference shifts).
- Enrollment table (p.10): rows and columns sum correctly (40 total, 8 women, 22/40 = 55% Hispanic). Not technical evidence.
Methods and analysis
- Baseline window. RASL is referenced to a log–log fit over the first 60 min of each patch.
- With two patches per shift, maximum coverage is (8 − 2 × 1)/8 = 75%, before counting time for flow to stabilise.
- The hour lost after the midday patch change falls in the early afternoon, when strain peaks, and the afternoon fit absorbs any strain already present.
- The early fit covers a narrow, low sweat-rate range and is then extrapolated to higher rates. A wrong slope will produce RASL trends that follow sweat rate, which itself rises with strain.
- Signal versus noise. The log base of "0.15 log units" is not stated: natural log means a 16% rise, log₁₀ a 41% rise. The pilot SD of 1.68 mM is about 17% at 10 mM, the same size as a natural-log threshold. Sweat-rate agreement is reported only as r = 0.83, not as agreement, and its error feeds directly into RASL.
- Reservoir saturation. A 120 µL reservoir filling in 4 h through 0.5 cm² implies 120/(240 × 0.5) = 1.0 µL·cm⁻²·min⁻¹. At 2 µL·cm⁻²·min⁻¹, which is plausible on the upper back during heavy work in heat, it fills in 120/(2 × 0.5) = 120 min. Data would then be lost on exactly the high-strain shifts, so the missing data are informative.
- Comparator. The comparison model leaves out the heart-rate-based core-temperature estimate that A.2 names as the current method [5], as well as patch skin temperature and sweat rate alone. Any AUC gain could therefore reflect thermal or sweat-rate information.
- Outcome coupling. PSI contains heart rate, so landmarks just below PSI 7 are trivially predictable. Performance should also be reported excluding those landmarks.
- Safety stops. Notification at 39.5 °C or a symptom rated 4 triggers the employer's protocol, which truncates episodes. No censoring rule is given.
- Reference method. Occlusive pads alter local skin temperature and sweat composition. The 45-min pad intervals and continuous readings with a 3-min lag need a stated rule for aligning them in time.
- Sex. With 8 women and no planned sex-specific analysis, sex effects on sweat rate and lactate go unexamined.
Feasibility
The timeline matches the support letter (December start; field seasons May to September 2028 and 2029). Tasks most likely to slip:
- Field coverage. 160 shifts in about 22 weeks is about 7 shifts per week for one coordinator (7.5 calendar months). That person also does real-time core-temperature safety monitoring, 45-min pad exchanges and pre-shift procedures, across dispersed trench and overhead-line crews. Telemetry range to the tablet is not addressed.
- Technician effort. 2.4 calendar months in Year 1 covers fabricating 380 patches with QC plus 60 × 8 × 2 = 960 duplicate assays.
- Patch supply. 380 patches against 320 needed gives 19% spare. If wear is shortened to 3.5 h (the Aim 1 fallback), an 8-h shift needs 3 patches, or 480 per season.
Technical risks not addressed
- The mechanism cannot be tested in the field. A chamber sub-study with laser-Doppler skin blood flow would address this. The chamber is currently listed for device testing only (p.12).
- Skin-surface contamination from residual lactate, dust, sunscreen or initial washout. A skin-cleaning protocol and discarding the first volume would address it.
- Hidromeiosis late in long shifts lowers sweat rate independently of perfusion and will move RASL. It needs a sensitivity analysis.
- Step changes when patches are replaced, given the 6.2% batch CV in sensitivity. The proposal does not report whether carrying the morning fit over to the afternoon patch is viable.
- No predefined handling of reservoir saturation. Fill time should be modelled against sweat rate, and a patch-change trigger added.
Strongest technical points
- Pre-registration before Season 1 and a blinded outcome pipeline with separate predictor and outcome files (C.6).
- Season 2 uses new workers in a new season, so validation is both temporal and person-independent, on top of leave-one-worker-out cross-validation.
- Measuring lactate and sweat rate in the same fluid path directly addresses the dilution confound [7] that undermined earlier sweat-lactate work.
- Aim 1 delivers value even if Aim 2 fails, and the proposal states the pilot's limits honestly (p.6).
Individual reviews
Summary
This observational R21 tests whether a sweat-lactate signal adjusted for local sweat rate can provide an early indication of physiological heat strain during outdoor work. It also evaluates the patch’s measurement accuracy under field conditions.
Factor 1, Importance of the Research — Strengths
- Reinterpreting gland-produced sweat lactate relative to concurrent sweat rate addresses a limitation of earlier biomarker work and could establish a useful new research direction (pp. 3–5).
- Field validation against sweat assays and core-temperature-based strain would be valuable even if the proposed warning signal fails (pp. 3–4).
Factor 1, Importance of the Research — Weaknesses
- A positive association would not establish the proposed glandular-perfusion mechanism, because skin blood flow is not measured (pp. 3–4, 6–7). This narrows the conceptual advance.
- The claim that this project will produce a sensor that prevents heat stroke exceeds an observational study with no alerts or prevention test (p. 3; C.9, p. 8).
Factor 2, Rigor and Feasibility — Strengths
- Same-path lactate and sweat-rate measurement, paired field references, prespecification, blinded outcome processing, and validation in new workers in Season 2 strengthen the design (pp. 5–7).
- The pilot supports field feasibility, although its nine strain episodes cannot establish predictive accuracy (C.2, pp. 5–6). Figures 1–3 render as captions only, so their depicted evidence cannot be independently checked.
Factor 2, Rigor and Feasibility — Weaknesses
- RASL requires a new 60-minute fit after each patch change. This removes prediction time, including after midday, and may absorb strain already developing; the plan does not establish that an early, potentially narrow sweat-rate range supports later predictions (C.3–C.5, pp. 6–7).
- There is no event-based precision justification for the AUC increment. The comparison omits the heart-rate-based temperature estimate identified as an existing approach, while heart rate is itself part of the PSI outcome (A.2, p. 4; C.5, p. 7).
- Reservoir filling in approximately four hours at typical sweat rates raises the possibility of earlier, informative data loss during heavy sweating. The missing-data plan does not specifically address saturation (C.2, p. 5; C.6, p. 7). Eight planned women permit little assessment of sex-related signal differences (p. 10).
Factor 3, Expertise and Resources
Appropriate
Additional review criteria
Human-subject protections address capsule and fingerstick risks, voluntary participation, employer confidentiality, and real-time safety notification (pp. 9–10). Fingerstick precautions address the relevant biohazard (p. 9). No additional concern worsened the score.
Solicitation alignment
- SC-01: $275,000 total; neither year exceeds $200,000 (pp. 1, 14).
- SC-02: Two years, months 1–24 (pp. 1, 8).
- SC-03: Observational measurements, with no assigned intervention (p. 9).
- SC-04: The untested RASL concept suits exploratory R21 work (pp. 3–5).
- SC-05: Occupational heat exposure supports the stated NIEHS fit (p. 4).
- SC-06: The DMS plan covers six elements; whether narrative after “Yes” in Element 5B conflicts with the required format is unclear (p. 13).
- SC-07: A resource-sharing plan offers design files and firmware under a no-cost academic license (p. 14).
- SC-08: The human-subjects study record was not supplied for review; a protections narrative is present (pp. 9–10).
- SC-09: Appendix materials were not supplied for review.
- SC-10: No foreign subaward appears in the justification; subaward forms were not supplied for review (p. 14).
- SC-11: Registration evidence was not supplied for review.
- SC-12: Race and ethnicity are not selection criteria (p. 10).
Overall impact
The potentially consequential new signal and useful field-validation dataset support a strong Importance score. Uncertain prediction coverage, comparator choice, and event-level precision temper confidence that this study will establish its added warning value.
- Factor 1: 3
- Factor 2: 5
- Factor 3: Appropriate
Summary
This exploratory study asks whether a sweat-rate-adjusted lactate signal can provide an individual early warning of physiological heat strain during outdoor work. It also seeks to establish how accurately the patch measures sweat lactate in that setting.
Factor 1, Importance of the Research — Strengths
- Interpreting sweat lactate as a glandular signal rather than a blood-lactate surrogate addresses a well-defined gap, with potential value even if prediction fails (pp. 3–4).
- Measuring lactate and sweat rate in one fluid path directly addresses a major confound in prior work (pp. 4–5).
Factor 1, Importance of the Research — Weaknesses
- The claim that this project will produce a sensor that prevents heat stroke exceeds its observational scope: alerts and their effects are reserved for a future trial (p. 3; C.9, p. 8).
Factor 2, Rigor and Feasibility — Strengths
- Laboratory reference assays, repeated-measures agreement analyses, preregistration, blinded outcome processing, and validation in new workers the following season support reproducibility (C.4–C.6, pp. 6–7).
- The pilot establishes field feasibility, but appropriately is not presented as an accuracy estimate for prediction: nine high-strain episodes are too few (C.2, pp. 5–6).
Factor 2, Rigor and Feasibility — Weaknesses
- The first hour of each patch establishes its RASL baseline, removing at least two hours per eight-hour shift from prediction, including time after midday replacement. Fitting each baseline over its own sweat-rate range may also create apparent trends as rates change (C.2–C.5, pp. 5–7).
- No event-count or precision justification supports the proposed AUC increment. The comparator includes heart rate, a component of the PSI outcome, but not the named heart-rate-based core-temperature method; thus incremental utility against current individual monitoring remains uncertain (A.2, p. 4; C.5, p. 7).
- Sweat-rate accuracy is reported as correlation rather than agreement, although its error enters RASL. The reservoir may fill before the planned midday change during heavy sweating, preferentially losing high-strain data; no saturation rule is given (C.2–C.6, pp. 5–7). Figure 2 (p. 5) and the selected trace in Figure 3 (p. 6) appear only as captions, so their graphical claims cannot be checked. Enrollment includes eight women, but no sex-specific analysis is planned (p. 10; C.5, p. 7).
Factor 3, Expertise and Resources
Appropriate
Additional review criteria
Human-subject protections address capsule and fingerstick risks, voluntary participation, confidentiality, and safety notification (pp. 9–10); no score-worsening concern identified. No vertebrate animals or significant biohazard concern is identified. This is a new application.
Solicitation alignment
- SC-01: $150,000 plus $125,000 meets both direct-cost limits (pp. 1, 14).
- SC-02: The period is two years, months 1–24 (pp. 1, 8).
- SC-03: Observational measurements, with no assigned intervention or displayed patch readings, support the non-trial determination (p. 9).
- SC-04: The untested RASL concept fits exploratory, higher-risk work (pp. 4–5).
- SC-05: Occupational heat exposure is linked to NIEHS interests (p. 4); assignment is not established here.
- SC-06: The DMS plan addresses six elements (p. 13); whether narrative after “Yes” in Element 5B conforms to the required Format Page is unclear.
- SC-07: A resource-sharing plan offers design files, protocols, and firmware under a no-cost research license (p. 14).
- SC-08: A human-subjects narrative is present (pp. 9–10); the study record was not supplied for review.
- SC-09: Appendix contents were not supplied for review.
- SC-10: No foreign subaward is listed in the justification (p. 14); subaward forms were not supplied for review.
- SC-11: Registration evidence was not supplied for review.
- SC-12: Race and ethnicity are not selection criteria (p. 10).
Overall impact
The distinctive signal concept and field-validation design give this R21 meaningful potential. Moderate weaknesses in baseline construction, informative sensor loss, and the ability to establish incremental predictive value reduce confidence in robust findings.
- Factor 1: 3
- Factor 2: 5
- Factor 3: Appropriate
Summary
This observational R21 tests whether a sweat-rate-adjusted lactate signal can provide early warning of physiological heat strain during outdoor work. It would establish field sensor performance and assess whether the signal warrants a future alerts trial.
Factor 1, Importance of the Research — Strengths
- Interpreting gland-derived lactate against concurrent sweat rate addresses a documented limitation of earlier approaches and could create a useful noninvasive signal (pp. 3–5).
- Field validation against core-temperature-based strain addresses a gap left by laboratory exercise studies (pp. 3–4).
Factor 1, Importance of the Research — Weaknesses
- Without skin-perfusion measurement, even a positive association cannot establish the proposed glandular oxygen-delivery mechanism (pp. 3–4, 7).
- The claim that this study will produce a sensor that prevents heat stroke exceeds its observational design; alerts and prevention are reserved for a later trial (pp. 3, 8).
Factor 2, Rigor and Feasibility — Strengths
- Same-path lactate and sweat-rate measurement, blinded reference assays, prespecification, and validation in new workers during a second season strengthen the design (pp. 5–7).
- The pilot supports field feasibility, while the application appropriately acknowledges that nine strain episodes cannot establish predictive accuracy (pp. 5–6).
Factor 2, Rigor and Feasibility — Weaknesses
- Each replacement patch requires a new 60-minute baseline after flow stabilizes. This sacrifices coverage around midday and may absorb strain already developing; a narrow early sweat-rate range could also create spurious later RASL trends (pp. 5–7).
- No event-based precision justification supports the proposed AUC increment. Sweat-rate error enters RASL, yet pilot sweat-rate performance is reported as correlation rather than agreement; the comparator also omits sweat rate alone and the cited heart-rate-based core-temperature estimate (pp. 5–7).
- Pilot thresholds were selected from few episodes, and false rises occurred on 5 of 15 non-episode shifts. Figures 2–3 are captions without visible graphics, so the claimed absence of proportional error and illustrated lead time cannot be checked (pp. 5–6).
Factor 3, Expertise and Resources
Appropriate
Additional review criteria
Human subjects: Occupational recruitment, capsule and fingerstick risks, confidentiality, consent without supervisors, and a core-temperature safety stop are addressed (pp. 9–10, 15). The real-time monitoring plan does not specify how safety stops operate during telemetry loss or across dispersed crews (pp. 6, 9); this modest concern worsened the score. Vertebrate animals: none. Biohazards: routine fingerstick precautions are specified (p. 9).
Solicitation alignment
- SC-01: $150,000 plus $125,000 meets both direct-cost limits (pp. 1, 14).
- SC-02: The 24-month timeline meets the period limit (pp. 1, 8).
- SC-03: No intervention effect is tested; the stated observational design is consistent with the no-trial requirement (p. 9).
- SC-04: RASL is a novel, explicitly risky exploratory application (pp. 4–5).
- SC-05: Occupational heat exposure has a stated NIEHS rationale; assignment is not shown (p. 4).
- SC-06: The DMS plan covers six elements (p. 13); Format Page use cannot be verified, and whether text after “Yes” violates a Yes/No field requirement is unclear.
- SC-07: A resource-sharing plan offers design files, protocols, and firmware under a no-cost academic license (p. 14).
- SC-08: A human-subjects narrative is present (pp. 9–10); the study record was not supplied for review.
- SC-09: Appendix materials were not supplied for review.
- SC-10: No foreign subaward appears in the justification (p. 14); subaward forms were not supplied for review.
- SC-11: Registration evidence was not supplied for review.
- SC-12: Race and ethnicity are not selection criteria (p. 10).
Overall impact
The potentially valuable new signal and strong field-validation framework are tempered by baseline-related loss and confounding of the predictor, uncertain precision for incremental prediction, and a modest safety-monitoring concern. These lessen, but do not foreclose, the prospect of influential exploratory findings.
- Factor 1: 3
- Factor 2: 5
- Factor 3: Appropriate
Panel discussion
I see a strong exploratory case for testing sweat lactate relative to local sweat rate, and the field validation could be useful even if the warning signal fails, as laid out in A.4 and C.2. My main concern is that the 60-minute refit after each patch change in C.3–C.5 may leave too little lead time to establish the claimed early warning value. I also think the prevention claim on page 3 goes beyond what this observational study can show, so I start at an overall impact score of 5.
I land at a 5: the RASL concept is a worthwhile exploratory test, and the reference assays and next-season validation in C.4–C.6 make the field study credible. My main concern is that the first hour of each patch is used to establish its baseline, so C.2–C.5 excludes early-shift and post-replacement time from prediction and may introduce sweat-rate-dependent trends. I am also unsure the proposed AUC increment will establish useful prediction, because C.5 gives no event-count or precision justification and does not compare against the heart-rate-based core-temperature method named in A.2.
I’m initially at a 5: this is a worthwhile exploratory field test, with blinded reference assays and validation in new workers in the second season. My main feasibility concern is that each replacement patch requires a new 60-minute baseline, which could leave midday strain unobserved or absorb a rise already underway into the baseline (pp. 5–7). On protections, the core-temperature safety stop is appropriate, but I need to know how it will work during telemetry loss or across dispersed crews (pp. 6, 9).
I agree with B that C.5 does not show whether the proposed AUC increment would demonstrate added value over the heart-rate-based temperature estimate named in A.2. C’s point about telemetry loss in the safety plan on pages 6 and 9 is worth clarifying, though I see it as a modest protections concern rather than a reason to discount the research question. For Importance, I still find the field test of the RASL concept compelling, while the absence of a skin-perfusion measure in A.4 limits what a positive result could establish about its proposed mechanism.
I agree with A that the field validation remains valuable even if the warning signal fails, but C.3–C.5 does not yet show that a baseline fitted during the first hour of each patch will leave enough usable time to test early warning. C’s point about the safety stop changes my view of the protections plan: pages 6 and 9 do not explain how staff would act on a high core temperature during telemetry loss or across dispersed crews. I still see that as an addressable concern, rather than a reason to discount the exploratory value of the study.
I agree with A and B that the repeated 60-minute baseline in C.3–C.5 is the clearest threat to the early-warning claim, not just a loss of recording time: a rise already underway after the midday patch change could be incorporated into the new baseline. B’s point about reservoir saturation in C.2 also changes how I view missing data, because heavy sweating could preferentially remove observations near high strain. I still see the safety stop on page 9 as a separate, modest protections concern: the plan does not say how staff will act on a core-temperature threshold if telemetry is lost or workers are spread across crews.
I remain at 5 because the RASL concept in A.4 and the field validation could yield valuable exploratory findings, but the repeated baseline fitting in C.3–C.5 may limit the study’s ability to establish an early warning signal. Final rating: 5
I see substantial exploratory value in the field validation and next-season testing in C.4–C.6, but the repeated baseline in C.3–C.5, possible reservoir saturation in C.2, and lack of an event-count justification or comparison with the method named in A.2 limit confidence in the early-warning claim. Final rating: 5
I remain at a 5 because the exploratory signal and blinded field validation in C.4–C.6 are compelling, but the repeated baseline in C.3–C.5 and potential saturation in C.2 weaken the early-warning test, while the safety plan on page 9 leaves a modest concern about telemetry loss across dispersed crews. Final rating: 5
Panel summary
Panel consensus — The panel agreed that this exploratory R21 addresses an important occupational heat-exposure problem through a distinctive sweat-rate-adjusted lactate signal. Field validation could yield useful findings even if prediction fails. Moderate weaknesses in baseline construction, informative sensor loss, and evaluation of incremental predictive value lessen confidence that the study will establish early-warning utility. A modest, addressable safety-monitoring concern also remained. All reviewers finalized overall impact scores of 5.
Factor 1, Importance of the Research — Consensus strengths
- Interpreting gland-derived lactate relative to concurrent sweat rate addresses a recognized limitation of earlier sweat-biomarker approaches and could establish a useful research direction.
- Field validation against reference assays and core-temperature-based strain would remain valuable even without a successful warning signal.
Factor 1, Importance of the Research — Consensus weaknesses
- Without skin-perfusion measurement, a positive association would not establish the proposed glandular oxygen-delivery mechanism, limiting the conceptual advance.
- The claim that this project will produce a sensor that prevents heat stroke exceeds its observational scope; alerts and prevention effects are not tested.
Factor 2, Rigor and Feasibility — Consensus strengths
- Same-path lactate and sweat-rate measurement addresses dilution confounding; laboratory reference assays support field measurement validation.
- Preregistration, blinded outcome processing, and validation in new workers during a second season strengthen reproducibility and independence.
- Pilot results support field feasibility, with appropriate acknowledgment that nine strain episodes cannot establish predictive accuracy.
Factor 2, Rigor and Feasibility — Consensus weaknesses
- Moderate: Each patch requires a new 60-minute baseline, excluding at least two hours per eight-hour shift from prediction. Midday refitting may absorb developing strain, while extrapolation from a narrow initial sweat-rate range may generate apparent signal trends.
- Moderate: No event-count or precision justification supports the proposed AUC increment. The comparator omits the cited heart-rate-based core-temperature estimate, while heart rate also contributes to the PSI outcome, weakening interpretation of incremental value.
- Moderate: Reservoir saturation during heavy sweating could preferentially remove high-strain observations; no saturation-specific missing-data rule is provided.
- Sweat-rate performance is reported as correlation rather than agreement, despite its measurement error entering RASL. Eight planned women and no sex-specific analysis limit assessment of sex-related signal differences.
- Figures 2–3 appear as captions without visible graphics, preventing independent examination of the claimed absence of proportional error and illustrated lead time; numerical pilot results remain available in the text.
Factor 3, Expertise and Resources — Appropriate
Additional review criteria — Human-subject protections leave a modest, addressable concern affecting impact: the safety plan does not specify how staff would implement core-temperature safety stops during telemetry loss or across dispersed crews. No other additional-criteria concern affected the score.
Solicitation-specific criteria
- SC-01: Addressed; stated direct costs meet annual and combined limits.
- SC-02: Addressed; the project period is two years.
- SC-03: Addressed; the observational design tests no assigned intervention.
- SC-04: Addressed; the novel, higher-risk RASL concept fits exploratory R21 priorities.
- SC-05: Addressed scientifically through occupational heat exposure and stated NIEHS relevance; actual IC assignment is not verifiable.
- SC-06: Substantive elements addressed; Format Page conformity is not verifiable. Narrative after “Yes” in Element 5B remains a formatting concern, not established noncompliance.
- SC-07: Addressed; the resource-sharing plan offers design files, protocols, and firmware under a no-cost academic license.
- SC-08: Not verifiable; formal human-subjects study-record compliance cannot be established.
- SC-09: Not verifiable; appendix restrictions cannot be assessed.
- SC-10: Not verifiable; no foreign subaward is indicated in the supplied justification.
- SC-11: Not verifiable; registration completion cannot be established.
- SC-12: Addressed; race and ethnicity are not participation selection criteria.
Points of disagreement
- Reviewers initially differed on whether safety-monitoring details warranted a score-affecting protections concern. Discussion resolved this as a modest, addressable concern without changing final scores. No meaningful scoring disagreement remained.
Impact assessment
- Factor 1: 3
- Factor 2: 5
- Factor 3: Appropriate
- Overall impact: 5
Comparative limitations — Competing applications, portfolio priorities, and funding constraints were unavailable; the recommendation is therefore an estimate with reduced confidence. Human-subjects study records, appendix materials, SF424/subaward forms, Senior/Key Person profiles, and registration/cover-sheet evidence were not supplied for review; this does not establish submission deficiencies. Actual IC assignment and DMS Format Page use remain unverified. Overall impact alone does not determine funding; NIH funding decisions depend on percentile, institute paylines, and priorities.
Justification — The unanimous final score reflects meaningful exploratory importance tempered by moderate limitations in prediction coverage, informative data loss, and demonstration of added predictive value, together with the modest safety-monitoring concern. These weaknesses lessen expected influence without establishing that the work cannot succeed.
All three reviewers at Impact 5.
Panel decision confidence: Moderate2 of 3
Recommended revisions
The panel's weaknesses and gaps, each with how to fix it, grouped by review stage. Where the fix is wording, GrantPanel suggests the words; where it is your research (aims, methods, results), it shows what to write and leaves the writing to you. It never edits your document.
Factor 1 — Importance of the Research
Address next- Bound the proposed mechanism to what the study can test Without skin-perfusion measurement, an association between RASL and whole-body strain would not establish the proposed glandular oxygen-delivery mechanism.Research Strategy A.3, p. 4, following the discussion of the forearm occlusion modelHow to fix
- Distinguish the mechanism supported by prior work from the association this field study can test.
- State the limit on mechanistic interpretation without changing the study's aims.
NowIn a forearm occlusion model, restricting local blood flow raised sweat lactate at constant sweat rate [9], linking sweat lactate to local oxygen delivery
Answer firstHow will you distinguish evidence for early warning from evidence for the proposed perfusion mechanism?
What to writeIn a forearm occlusion model, restricting local blood flow raised sweat lactate at constant sweat rate [9], linking sweat lactate to local oxygen delivery. [State what an association in this study could—and could not—establish about the proposed glandular oxygen-delivery mechanism.] - Remove the unsupported prevention claim The Specific Aims say this observational, no-alert R21 will produce a sensor that prevents heat stroke, although prevention effects are reserved for a future trial.Specific Aims, Expected outcomes and impact, p. 3How to fix
- Replace the prevention claim with the proposed assessment of early-warning potential.
- Retain the stated future-trial distinction.
NowBy turning an inexpensive skin patch into an early warning of dangerous strain, this work will produce a wearable sensor that will prevent exertional heat stroke among the millions of U.S. workers who labor outdoors through the summer. A positive result will support an R01 that tests real-time alerts in a prospective trial
Suggested wordingThis work will test whether an inexpensive skin patch can provide an early-warning signal of dangerous strain among outdoor workers. A positive result will support an R01 that tests real-time alerts in a prospective trial.
Factor 2 — Rigor and Feasibility
Address next- Resolve prediction coverage and refitting after patch replacement A new 60-minute RASL fit for each patch excludes at least two hours per eight-hour shift, including time after midday replacement. Refitting may absorb developing strain, and an early narrow sweat-rate range may create apparent later trends.Research Strategy C.5, Predictor, p. 7How to fix
- Specify how prediction coverage after each patch change will be quantified and reported.
- Supply a prespecified way to assess midday refitting and extrapolation beyond the baseline sweat-rate range.
NowFor each patch, during the first 60 minutes after flow stabilizes, we fit log lactate on log sweat rate. RASL is the observed minus the predicted log lactate from then on
Answer firstWhat prespecified checks will determine whether the post-replacement fit supports valid predictions during the remainder of the shift?
What to writeFor each patch, during the first 60 minutes after flow stabilizes, we fit log lactate on log sweat rate. RASL is the observed minus the predicted log lactate from then on. [Specify prediction coverage after each fit and how the analysis will assess developing strain at midday and predictions outside the baseline sweat-rate range.] - Justify event-level precision and the incremental-value comparison The proposal gives no event-count or precision justification for an AUC increment of at least 0.05. Its comparator omits the cited heart-rate-based core-temperature estimate, while heart rate contributes to the PSI outcome.Research Strategy C.5, Analysis, p. 7How to fix
- Supply an event-based precision justification for the AUC increment.
- Decide how to evaluate incremental value against the cited individual-monitoring approach and interpret heart-rate overlap with PSI.
NowThe comparison model includes WBGT, heart rate, accelerometry-derived work intensity, acclimatization status (days worked outdoors in the previous 14), pre-shift urine specific gravity, age, and body mass index. The full model adds the RASL change
Answer firstWhat anticipated event count and comparison analysis support interpreting an AUC increment of at least 0.05 as added warning value?
What to writeThe comparison model includes WBGT, heart rate, accelerometry-derived work intensity, acclimatization status (days worked outdoors in the previous 14), pre-shift urine specific gravity, age, and body mass index. The full model adds the RASL change. [Supply the event-based precision justification and specify how incremental value and the overlap between heart rate and PSI will be assessed.] - Address reservoir saturation as informative missingness The reservoir may fill before midday during heavy sweating, preferentially removing high-strain observations; the missing-data plan has no saturation-specific rule.Research Strategy C.6, Missing data, p. 7How to fix
- Specify how reservoir saturation will be identified and classified.
- Supply a prespecified analysis of its relationship to sweat rate and high-strain observation loss.
NowMissing segments will be classified by cause (adhesion loss, radio dropout, capsule passage); covariates will be multiply imputed, and sensitivity analyses will exclude shifts with less than 70% complete data
Answer firstHow will you detect saturation and assess whether it disproportionately removes observations near high-strain episodes?
What to writeMissing segments will be classified by cause (adhesion loss, radio dropout, capsule passage); covariates will be multiply imputed, and sensitivity analyses will exclude shifts with less than 70% complete data. [Specify how reservoir saturation will be identified, classified, and assessed as a potentially informative source of missing data.] - Evaluate sweat-rate agreement relevant to RASL Pilot sweat-rate performance is given as correlation rather than agreement, although sweat-rate measurement error enters the RASL predictor.Research Strategy C.2, Field pilot, p. 5, and C.4, Analysis, p. 6How to fix
- Supply the available sweat-rate agreement evidence or identify what remains unknown from the pilot.
- Specify how Aim 1 sweat-rate error will be interpreted for Aim 2 RASL performance.
NowPatch-derived sweat rate correlated with gravimetric pad sweat rate (r = 0.83
Answer firstWhat sweat-rate agreement evidence is available, and how will its error be considered when interpreting RASL predictions?
What to writePatch-derived sweat rate correlated with gravimetric pad sweat rate (r = 0.83). [State any available sweat-rate agreement evidence, or identify it as unknown, and explain how Aim 1 agreement results will inform interpretation of RASL.] - Address sex-related limits on signal assessment Enrollment plans include eight women, but no sex-specific analysis is planned to assess possible sex-related differences in the signal.Research Strategy C.5, Analysis, p. 7How to fix
- Decide what sex-related assessment the planned sample can support.
- State how its limitations will be reported without implying precision the sample cannot provide.
NowSecondary outcomes use the same framework, and raw lactate and lactate flux (concentration times sweat rate) will be examined as exploratory predictors
Answer firstWhat assessment of sex-related signal differences is supportable with the planned enrollment of eight women?
What to writeSecondary outcomes use the same framework, and raw lactate and lactate flux (concentration times sweat rate) will be examined as exploratory predictors. [State the planned assessment of sex-related signal differences and the limitations imposed by enrollment of eight women.] - Restore the pilot graphics needed to inspect the claims Figures 2–3 appear only as captions, preventing inspection of the claimed absence of proportional error and the illustrated lead time.Research Strategy C.2, insert the Figure 2 graphic after its caption on p. 5 and the Figure 3 graphic after its caption on p. 6How to fix
- Restore Figures 2 and 3 in the submitted PDF.
- Check that their plotted values, axes and panels support the existing captions.
Answer firstCan you supply the original Figures 2 and 3 and confirm that both render in the submission PDF?
What to write[Insert the original Figure 2 and Figure 3 graphics at their captioned locations, and verify the captions against the rendered plots.]
Additional review criteria — Protections for Human Subjects
Address next- Specify safety-stop procedures during telemetry loss or dispersed work The human-subjects safety plan does not say how staff will implement core-temperature safety stops if telemetry is lost or crews are dispersed.Protection of Human Subjects, Adequacy of protection against risks, Protections, p. 9How to fix
- Specify the safety-monitoring procedure during telemetry interruption.
- Identify how the coordinator will communicate with workers and the designated heat-safety lead across monitored crews.
NowDuring monitored shifts, the coordinator views core temperature and heart rate (never patch data) on a research tablet. If core temperature reaches 39.5 °C (below the 40 °C level associated with heat stroke) or a participant reports a symptom rated 4, the coordinator notifies the participant and the crew's designated heat-safety lead, who applies the employer's existing heat-illness protocol
Answer firstWhat procedure will staff use to maintain safety monitoring and implement notifications when telemetry is unavailable or a crew is out of immediate reach?
What to writeDuring monitored shifts, the coordinator views core temperature and heart rate (never patch data) on a research tablet. If core temperature reaches 39.5 °C (below the 40 °C level associated with heat stroke) or a participant reports a symptom rated 4, the coordinator notifies the participant and the crew's designated heat-safety lead, who applies the employer's existing heat-illness protocol. [Specify the safety-monitoring and notification procedure during telemetry loss or work across dispersed crews.]
Solicitation-specific criterion SC-06 — DMS Plan format
The solicitation requires- Check the DMS controlled-access response against the required format DMS Element 5(B) adds narrative after “Yes.” The packet does not establish whether this is a Yes/No-only field on the required NIH Format Page, so conformity remains a formatting concern rather than an established violation.Data Management and Sharing Plan, Element 5(B), p. 13How to fix
- Check Element 5(B) against the NIH DMS Plan Format Page used for submission.
- If it is Yes/No-only, move the access-agreement detail to a permitted narrative field.
NowB) Controlled access. Yes. Access requires an approved data use agreement through the repository
Answer firstDoes the NIH DMS Plan Format Page used for this application make Element 5(B) a Yes/No-only field?
What to write(B) Controlled access. Yes. [If this is a Yes/No-only field, place the approved data use agreement detail in a permitted narrative field; otherwise retain the existing detail here.]
Supporting research
Prior funded work
Related awards found in NSF Award Search, NIH RePORTER and USAspending.gov, closest first (searched: sweat lactate sensing; occupational heat strain; wearable sweat biosensor; heat illness prediction). Use them to judge novelty and positioning, not as a verdict: a related award is context for how the proposal differs, not evidence it should not be funded.
NIH award: Wearable Sweat Chemistry Monitoring of Heat Strain in Agricultural Workers Odalys Fennimore, Halvern Plains University · $418,600 · 2024–2026 · NIEHS exploratory/developmental research program Like the proposal, it uses wearable sweat biochemistry to track heat strain in outdoor workers, but it measures sodium and chloride for hydration status rather than lactate read against sweat rate, and its outcome is dehydration rather than high-strain episodes.
NSF award: Microfluidic Epidermal Sensors with Integrated Sweat-Rate Referencing Darius Penwarden, Westmarch Institute of Technology · $449,812 · 2023–2026 · Engineering biosensing program It shares the idea of pairing each metabolite reading with local sweat rate in one microfluidic device, but it is a device-development project tested only during laboratory exercise, with no field deployment or physiological strain outcomes.
NIH award: Noninvasive Core Temperature Estimation for Heat-Exposed Construction Workers Liesl Marwick, Sorrel Bay University · $2,284,350 · 2022–2027 · NIEHS research project grant program It targets the same goal of individual early warning of heat strain in construction crews, validated against ingestible capsules, but it relies on heart rate and skin temperature rather than a biochemical sweat signal.
DOW award: Physiological Status Monitoring to Reduce Heat Casualties During Military Training Kwame Ashdown, Upper Tamsin State University · $1,176,000 · 2021–2025 · Army military operational medicine research program It predicts high heat strain from wearables against core temperature, as the proposal does, but in young trainees on fixed training schedules and with heart-rate-based models rather than sweat chemistry.
USDA NIFA award: Heat Stress Surveillance and Wearable Monitoring for Farmworker Safety Ines Calloway-Tran, Cardinal Mesa University · $649,500 · 2023–2027 · Agricultural workforce safety and health program It combines wearable heart-rate monitoring, environmental sensors, and symptom surveillance among farmworkers, overlapping with the proposal's outcomes, but it involves no sweat sensing and centers on community-level surveillance rather than an individual early-warning signal.
Literature review (deep research)
Research area & central claim
The proposal sits where wearable sweat biosensing meets occupational heat-strain physiology. Its central claim is that sweat lactate, read against the concurrent local sweat rate, reflects metabolic strain in the sweat gland. On that view a rising rate-adjusted lactate trend should anticipate high whole-body physiological strain in outdoor workers about 15 minutes or more before it occurs. Review cutoff: October 16, 2026.
State of the art at submission
- Castellane et al., 2019, Applied Biofluid Chemistry: established that most sweat lactate comes from glycolysis in the eccrine gland rather than from blood, which explains its poor performance as a blood-lactate surrogate.
- Quillfeather et al., 2022, Applied Biofluid Chemistry: a pooled analysis of 14 laboratory studies showing that sweat lactate concentration falls with rising sweat rate, and that adjusting for sweat rate reconciles conflicting earlier trends.
- Iwasaki-Dorn et al., 2023, Interfacial Bioelectronics Reports: identified temperature and pH effects on lactate oxidase as the dominant sources of drift in skin-worn enzymatic sensors during multi-hour wear.
- Tessaro et al., 2024, Wearable Biosensing Letters: showed that sweat-sensor agreement measured during laboratory cycling overstates agreement during free-living manual work.
- Harrowgate et al., 2023, Journal of Field Exposure Physiology: combined heart rate with a chest-to-limb skin-temperature gradient to estimate core temperature in 57 agricultural workers, holding error near 0.3 °C over full shifts, the best noninvasive physiological estimator reported for field work.
- Dunmore-Achebe et al., 2022, Interfacial Bioelectronics Reports: deployed sweat sodium and chloride patches on outdoor workers for hydration monitoring. Adhesion lasted multiple hours in the field, but electrolyte trends did not anticipate core-temperature rise.
- Penhallow et al., 2026, Annals of Outdoor Workforce Health: a cluster-randomized trial in 14 construction crews. Rest prompts triggered by heart rate and skin temperature reduced time spent above a physiological strain index of 7 but did not change symptom reports, giving the first trial evidence that physiological alerts can alter strain in the field.
Competing approaches
- Noninvasive physiological estimation of core temperature (Harrowgate et al., 2023, Journal of Field Exposure Physiology, and the heart-rate estimators the proposal cites). The proposal positions its lactate signal as complementary and tests its increment over a comparison model, but that model includes heart rate, not the stronger skin-temperature-gradient estimator.
- Sweat electrolyte and hydration sensing (Dunmore-Achebe et al., 2022, Interfacial Bioelectronics Reports). The proposal argues that lactate, interpreted against sweat rate, captures glandular metabolic strain rather than fluid balance, and so may anticipate strain where electrolytes did not.
- Alert-driven work-rest interventions (Penhallow et al., 2026, Annals of Outdoor Workforce Health). The proposal is upstream of this work: it is observational, withholds readings from workers, and defers any test of alerts to a later trial.
- Environmental-index programs based on WBGT action limits. The proposal argues that between-worker variation in strain at the same WBGT calls for an individual-level signal.
Open problems the proposal addresses
- Whether glandular sweat lactate carries information about whole-body heat strain outside the laboratory.
- Field-grade accuracy of continuous enzymatic sweat sensing over full work shifts, with sweat rate measured in the same fluid.
- The absence of shared datasets pairing continuous sweat chemistry with core temperature during real outdoor work.
Citation gaps at submission
- Thistlewood et al., 2025, Journal of Occupational Thermal Physiology: among 46 heat-acclimated outdoor workers, women had lower local sweat rates and higher sweat lactate concentration at matched sweat rates than men. This matters because the proposal's lactate-to-sweat-rate relationship is central to its signal, and its enrollment plan includes women, yet its analysis does not consider sex.
- Ferreira-Lund et al., 2024, Statistical Methods in Wearable Health: showed that the precision of early-warning discrimination estimates from wearable streams depends on the number of independent participants and events, not the number of prediction windows, and gave methods for planning both. This bears directly on whether a 40-worker cohort can support the proposal's AUC benchmarks.
In a real report, each work ends with a link to its own page (DOI, journal, PubMed or arXiv), so you can check it in one click. The works in this sample are invented, so they have none.
Retrospective developments (not used for scoring)
None reviewed.
References lookup
The proposal cites 19 references. Each was looked up in Semantic Scholar (Allen Institute for AI), OpenAlex and Crossref: 17 were found, and the reviewers read their abstracts. Any that no index matched are marked.
- [1] Trends in work-related heat illness among U.S. outdoor occupations, 2011-2022 (2023) — Varnell, J., Tessaro, A., Quillfeather, R. · Annals of Outdoor Workforce Health
- [2] Undercounting of occupational heat illness in workers' compensation and emergency department records (2021) — Ambrosek, L., Haverlund, S. · not found in the indexes
- [3] Time course of core temperature rise preceding exertional heat stroke in field workers (2020) — Lindgaard, E., Marchbank, T., Fenwright, D. · Journal of Occupational Thermal Physiology
- [4] Limits of wet-bulb globe temperature action thresholds for predicting individual heat strain (2022) — Bricknell, H., Yarrowby, P. · Journal of Field Exposure Physiology
- [5] Heart-rate-based estimation of core temperature in manual laborers: accuracy and drift across a work shift (2024) — Delacorte, M., Iwasaki-Dorn, K., Varnell, J. · Journal of Occupational Thermal Physiology
- [6] Eccrine gland glycolysis and the local origin of sweat lactate (2019) — Castellane, I., Mbatha-Rowe, N. · Applied Biofluid Chemistry
- [7] Sweat-rate dependence of sweat lactate concentration: a pooled analysis of fourteen laboratory studies (2022) — Quillfeather, R., Ambrosek, L. · Applied Biofluid Chemistry
- [8] Cutaneous perfusion limits during prolonged work in the heat (2018) — Haverlund, S., Tessaro, A. · Journal of Field Exposure Physiology
- [9] Sweat lactate as a marker of local ischemia: a forearm occlusion model (2021) — Marchbank, T., Delacorte, M., Yarrowby, P. · Applied Biofluid Chemistry
- [10] Why sweat lactate fails as a blood lactate surrogate: a systematic review (2020) — Fenwright, D., Castellane, I. · Wearable Biosensing Letters
- [11] Enzymatic lactate sensors on skin: drift, temperature, and pH effects over multi-hour wear (2023) — Iwasaki-Dorn, K., Lindgaard, E., Bricknell, H. · Interfacial Bioelectronics Reports
- [12] Microfluidic sweat-rate sensing by conductive front tracking (2022) — Mbatha-Rowe, N., Fenwright, D. · not found in the indexes
- [13] A sweat-rate-referenced lactate patch with on-board temperature and pH compensation (2025) — Velarde-Kim, M., Reinholt, T., Quillfeather, R. · Interfacial Bioelectronics Reports
- [14] Field feasibility of continuous sweat lactate sensing in utility and road crews (2026) — Reinholt, T., Velarde-Kim, M., Ostrander, P., Pemberly-Ash, R. · Annals of Outdoor Workforce Health
- [15] Physiological strain during summer shifts in trenching and road construction crews (2021) — Reinholt, T., Pemberly-Ash, R., Varnell, J. · Journal of Occupational Thermal Physiology
- [16] Landmark discrete-time models for early warning from wearable data streams (2024) — Ostrander, P., Yarrowby, P. · Statistical Methods in Wearable Health
- [17] Agreement analysis for repeated wearable measurements under clustering (2023) — Ostrander, P., Reinholt, T. · Statistical Methods in Wearable Health
- [18] Laboratory exercise protocols overstate the agreement of sweat sensors: a comparison with free-living work (2024) — Tessaro, A., Haverlund, S., Marchbank, T. · Wearable Biosensing Letters
- [19] Physiological strain index thresholds and heat illness symptoms in outdoor laborers (2019) — Yarrowby, P., Bricknell, H. · Journal of Field Exposure Physiology