{"id":"79b42bd6-1d88-4800-8974-8827da715839","arxiv_id":"1908.09371","paper_version":3,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"In a reduced two-state baroreflex model with a sympathetic delay, external forcing stabilizes the sink region and leaves the Hopf and unstable boundaries unchanged, with a POTS patient classified as a stable focus.","lead":"This paper analyzes a delay differential equation model of the body's blood pressure and heart rate control during the Valsalva maneuver, and maps where the model becomes oscillatory or unstable. It reports that the external blood-pressure forcing shrinks the oscillatory region and that a POTS patient falls in a damped-oscillation region, supporting the idea of overactive sympathetic signaling.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed sink-region expansion in Fig. 9b may be an artifact of the artificially extended SBP polynomial used to build the forcing, so the central nonhomogeneous stability result is not established.","rationale":"The homogeneous analysis is mostly sound: the Lambert-W derivation of tau_s = e D_s as the real-root threshold and D_s = (pi/2) tau_s as the imaginary-axis crossing is standard, and the numerical solver choice is reasonable. The load-bearing part of the paper, however, is the claim that the forcing function qualitatively enlarges the sink region while preserving the Hopf and unstable locations. That claim is entirely supported by a numerical classifier applied to solutions driven by an artificially extended, polynomial-fitted forcing signal. The paper explicitly discloses the artificial extension, and that disclosure is exactly where the argument is weakest. Without a re-run using forcing built from unextended data, or at least a sensitivity analysis of the polynomial degree and classifier thresholds, the reported boundary shift cannot be distinguished from a numerical artifact. There are additional issues: the 'transcritical bifurcation' at tau_s = e D_s is not a bifurcation at all, since the zero root is simple and the origin remains stable on both sides, and the term 'Hopf bifurcation' is applied to a nonautonomous linear system without a well-defined equilibrium. These issues reinforce the need for a conditional verdict, but the artificial forcing extension is the first-order concern because it bears directly on the central numerical result and on the patient classification that follows from it.","tokens_in":17230,"tokens_out":8070,"duration_ms":93958,"concrete_test":"Recompute the nonhomogeneous bifurcation map (Fig. 9b) with three forcing variants: (i) the fitted polynomial truncated at the end of VM phase IV, with no artificial post-maneuver extension; (ii) the one-second movmean SBP used directly, without the polynomial fit; and (iii) a simple boxcar Pth step over constant baseline SBP. If the gray region no longer extends beyond tau_s = e D_s, or if the orange/blue lines move substantially, the reported stabilization is an artifact of the artificial extension. Also recompute the map with at least one higher and one lower polynomial degree and with eta1 perturbed by +/-0.1, reporting how the sink/focus and limit-cycle boundaries shift.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central numerical claim (Fig. 9b and Sec. 4) is that the forcing vector f(t) moves the sink/stable-focus boundary beyond the homogeneous line tau_s = e D_s while leaving the limit-cycle and unstable lines at the same locations. That claim is not a stability theorem: the homogeneous boundary comes from characteristic roots of Eq. (29), whereas the nonhomogeneous map is produced by Algorithm 1, which counts local extrema of Ts after the VM and regresses their amplitudes. The forcing in Sec. 2.4 is manufactured: SBP is smoothed with a one-second movmean, fit by a 10th-degree polynomial, and then 'artificially extended ... after to accentuate the oscillatory behavior of the signal if it arises' (paragraph after Fig. 5). A 10th-degree polynomial extrapolated to 120 s is not physiological data, and because G_p and G_s saturate, f(t) and g(t) are bounded, non-decaying, non-periodic waveforms that can force oscillations even in a linearly stable system. The reported gray-region expansion and the claimed coincidence of the orange/blue lines are therefore indistinguishable from properties of the polynomial tail, the artificial extension, and the hand-picked thresholds (eta1 = 0.5, eta2 = -1e-2, mu = 0.8), rather than intrinsic DDE dynamics.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies a two-state reduction of a baroreflex model of the Valsalva maneuver, consisting of a linear delayed differential equation for sympathetic outflow Ts coupled to heart rate H, together with nonhomogeneous forcing terms f(t) and g(t) derived from filtered and polynomial-fitted systolic blood pressure data. For the homogeneous system the authors derive analytical stability boundaries via the Lambert W function, obtaining a real-to-complex root transition at τs = e Ds and an imaginary-axis crossing at Ds = πτs/2. For the nonhomogeneous system they use an algorithmic classification of solution waveforms to claim that the forcing expands the sink region beyond the τs = e Ds line while leaving the limit-cycle and unstable boundaries at the same locations. They then classify two control subjects and one POTS patient as belonging to the sink or stable-focus regions and argue that the model supports the hypothesis of altered sympathetic activity in POTS.","tokens_in":17516,"tokens_out":4385,"duration_ms":43784,"significance":"If the nonhomogeneous stability map were validated, the paper would provide an interesting demonstration that bounded, nonperiodic forcing can shift apparent stability boundaries in a physiological delay differential equation model, and the homogeneous boundary in the (Ds, τs) plane is a clean, correct analytical result. The paper has several strengths: the Lambert-W-based derivation of the homogeneous stability boundaries is careful and reproducible; the reduction from five states to two is clearly documented; and Algorithm 1 is explicit enough to be checked. However, the central claim about the nonhomogeneous system rests on forcing functions constructed with an artificial extension of the data and on an ad-hoc waveform-classification algorithm. Without robustness checks, the claimed stabilization by forcing is not established, and the clinical interpretation for the POTS patient depends on that unvalidated numerical map.","major_comments":[{"comment":"The transition at τs = e Ds is repeatedly called a 'transcritical bifurcation' (Abstract, Section 3.1, and the caption of Fig. 9). However, for the homogeneous system (21) the origin is the unique equilibrium and remains stable on both sides of this line; what changes is that the two real characteristic roots coalesce and become complex. There is no second equilibrium crossing and no exchange of stability, so this is not a transcritical bifurcation. This mischaracterization propagates to the abstract and conclusions and should be corrected, e.g., by describing it as a node-to-focus transition or a change in the nature of the stable equilibrium.","section":"Section 3.1, Eq. (37) and Fig. 9; also Abstract"},{"comment":"The central nonhomogeneous result—that the forcing expands the sink region beyond τs = e Ds while preserving the Hopf and unstable boundaries—is produced by Algorithm 1 applied to solutions driven by forcing functions built from a 10th-degree polynomial fit to a one-second moving average of SBP and artificially extended after the maneuver 'to accentuate the oscillatory behavior of the signal if it arises.' Because these forcing functions are bounded, non-decaying, and non-periodic, they can generate oscillatory or decaying-oscillatory transients that Algorithm 1 may attribute to intrinsic DDE stability. As written, the reported boundary shift is indistinguishable from an artifact of the polynomial tail and extension procedure. The authors should repeat the classification with forcing built from the unextended data (or with truncation at several different times), or otherwise demonstrate that the claimed boundary shift is invariant to the extension, or refrain from claiming a stabilizing effect of the forcing.","section":"Section 2.4, paragraph after Fig. 5; Section 3.2; Fig. 9b"},{"comment":"The numerical 'stability regions' for the nonhomogeneous system are defined by counting local extrema of Ts after the maneuver and regressing their amplitudes with ad-hoc thresholds η1 = 0.5, η2 = −10⁻², and μ = 0.8. This classifies transient waveform shapes, not stability in the sense of convergence to an equilibrium; a nonautonomous system with persistent bounded forcing does not possess an equilibrium whose stability can be read from the homogeneous characteristic roots. The paper should justify that the Algorithm-1 boundary corresponds to a dynamically meaningful invariant (e.g., exponential contraction to a bounded attractor) or at least validate the thresholds and the post-maneuver window against problems with known decay rates. Without such validation, the 'stability regions' in Fig. 9b are not tied to the analytical stability theory of Section 3.1.","section":"Section 3.2, Algorithm 1"}],"minor_comments":[{"comment":"The sentence 'Substituting λ* into equation (29) and setting λ* ≤ 0 yields' appears to be a typo: the condition is φ(λ*) ≤ 0, not λ* ≤ 0. Please fix the wording.","section":"Section 3.1, text near Eq. (36)"},{"comment":"There are two unresolved 'Figure ??' placeholders: one after the description of the SBP polynomial fit and one in the discussion of the Lambert W branches. These should be replaced with the correct figure references.","section":"Section 2.4 and Section 3.1"},{"comment":"The filtering rule based on |Mi − mi| < 10⁻⁸ is described in prose but the formatting of the vector definitions and loop makes it easy to misread; a numbered or displayed procedure would improve clarity.","section":"Algorithm 1, step 4"},{"comment":"The estimated Ds and τs values for the three subjects are presented without uncertainties or details of how they were estimated; a brief explanation of the estimation procedure would help readers assess the classification claims.","section":"Table 4"}],"recommendation":"major_revision","confidential_remarks":"The homogeneous analysis is sound and the model reduction is attractive, but the nonhomogeneous stabilization claim is the paper's headline result and it depends on an artificial extension of the forcing data and on an unvalidated classification algorithm. The 'transcritical bifurcation' mislabeling should also be corrected. I recommend major revision: the authors need to redo or robustify the nonhomogeneous analysis before the central claim can be accepted."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague — the honest reading: the homogeneous half is a correct, if textbook, exercise; the nonhomogeneous half is where the load-bearing weakness sits, and I don't think the central claim survives contact with the forcing construction.\n\nWhat's genuinely new: applying Lambert W stability analysis to this two-state baroreflex model, deriving the τs = e Ds and Ds = (π/2)τs boundaries, and then attempting a numerical two-parameter classification for the forced system. The model reduction from five to two states is useful and the comparison against the full model looks fair. The algorithm that counts local extrema and regresses amplitudes is a reasonable heuristic for classifying post-VM dynamics, though it is just that.\n\nThe soft spots, in order of severity. First, the nonhomogeneous result in Figure 9b is not an established stability map. The forcing vector comes from a 10th-degree polynomial fit to a one-second moving average of SBP, and that polynomial is artificially extended after the data window \"to accentuate the oscillatory behavior of the signal if it arises.\" That is not physiology; it's a way of injecting oscillations. A bounded, non-decaying, non-periodic forcing can produce persistent oscillations in a linearly stable DDE, so the expanded gray sink region and the claimed shift of the red line are at least partly an artifact of the polynomial tail. The stress-test note is right about this.\n\nSecond, the bifurcation language is wrong. The line τs = e Ds is where two real characteristic roots become complex, with the origin stable on both sides. That is not a transcritical bifurcation; there is only one equilibrium and no exchange of stability. Likewise, in a linear/nonautonomous system, calling the imaginary-axis crossing a Hopf bifurcation is a misuse of the term. These are not cosmetic; they affect how a reader interprets the stability diagram.\n\nThird, the subject-specific parameters in Table 4 (Ds, τs) are reported without a fitting procedure or uncertainty; readers can't reproduce the classification. And the clinical conclusion rests on a single POTS patient.\n\nWhat holds up: the homogeneous derivation is correct, the reduction is reasonable, and the qualitative comparison with the full model is credible. If the forcing were cleaned up — remove the artificial extension or justify it, run a sensitivity analysis on the thresholds — the numerical map might become a useful tool for categorizing Valsalva responses.\n\nWho should read it: people working on DDE stability in cardiovascular modeling, and anyone who wants a cautionary example of how synthetic forcing can contaminate numerical stability classification. It deserves a serious referee but not acceptance in this form. The revision should fix the nomenclature, address the forcing, and provide reproducible parameter estimation.","headline":"A correct but standard homogeneous DDE stability analysis attached to a nonhomogeneous numerical map whose central claim is likely contaminated by the artificially extended forcing function.","tokens_in":18022,"tokens_out":3607,"would_cite":false,"duration_ms":37074,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["34K18","34K20","37G10","92C30"],"pacs":[],"model":"deepseek-v4-flash","headline":"Blood-pressure-derived forcing widens the stable sink region of a delay-equation Valsalva model while leaving the Hopf threshold unchanged.","keywords":["delay differential equations","Valsalva maneuver","Hopf bifurcation","transcritical bifurcation","postural orthostatic tachycardia syndrome","baroreflex model","Lambert W function","stability analysis"],"falsifier":"Recompute the nonhomogeneous boundary with the forcing functions truncated at the end of the Valsalva maneuver, removing the artificial post-maneuver extension; if the overdamped-sink region no longer extends beyond $\\tau_s = eD_s$, or if the POTS subject moves to a different region, the claimed stabilization is an artifact of the constructed input rather than a property of the DDE.","tokens_in":16978,"feed_emoji":"🫀","tokens_out":5450,"duration_ms":50375,"temperature":0.7,"pith_summary":"The paper reduces a five-state model of autonomic response to the Valsalva maneuver to two states, sympathetic tone and heart rate, and asks where oscillations and instability arise as the sympathetic delay $D_s$ and time-scale $\\tau_s$ vary. For the homogeneous unforced system, the authors derive exact boundary lines from the characteristic equation: $\\tau_s = eD_s$ separates real from complex roots, and $D_s = \\tau_s\\pi/2$ is the Hopf line where the origin loses stability. For the nonhomogeneous system driven by blood-pressure-derived forcing, numerical classification shows that the forcing widens the overdamped-sink region beyond $\\tau_s = eD_s$ while leaving the Hopf and unstable regions essentially unchanged. Placing three subjects in this diagram, two controls and one patient with postural orthostatic tachycardia syndrome (POTS), the stable-focus location of the POTS patient is read as support for the clinical hypothesis of overactive sympathetic activity.","feed_headline":"Blood-pressure forcing widens stable region in Valsalva heart model","feed_subtitle":"Patient blood-pressure inputs stabilize the model's sink region; the Hopf instability line does not move.","key_machinery":"The carrying object is the characteristic equation $\\varphi(\\lambda) = \\tau_s\\lambda + e^{-\\lambda D_s} = 0$ for the delayed sympathetic-tone equation, solved through the Lambert W function; its roots determine the homogeneous stability boundaries. The companion mechanism is the forcing pair $f(t)$ and $g(t)$ built from a tenth-degree polynomial fit to one-second moving-average systolic blood-pressure data, which turns the homogeneous system into the nonhomogeneous system $\\dot{x} = Ax + Bx_{D_s} + \\mathbf{f}$. Numerical solution on a $D_s \\times \\tau_s$ mesh with a qualitative amplitude-trend classifier then locates the nonhomogeneous boundaries.","core_discovery":"The central claim is that in the nonhomogeneous system the forcing function acts as a stabilizer: the gray overdamped-sink region extends beyond the analytically derived line $\\tau_s = eD_s$, the transcritical boundary becomes a curve rather than a line, and the stable-focus region shrinks accordingly, while the Hopf bifurcation line $D_s = \\tau_s\\pi/2$ and the unstable region occur at the same parameter locations as in the homogeneous system. The authors further claim that the model's classification of the POTS patient in the stable-focus region supports the clinical hypothesis that POTS involves altered, overactive sympathetic nervous system activity.","pith_inferences":["If the polynomial fit's artificial extension after the maneuver injects a decaying oscillation, the claimed widening of the sink region may be partly an input artifact; a natural test is to rerun the grid with forcing truncated at the end of the maneuver.","The same stability analysis could be applied to other transient autonomic challenges, such as tilt-table tests, by replacing the Valsalva forcing while keeping the homogeneous characteristic equation fixed.","Parameter pairs just inside the stable-focus region near the Hopf line predict slowly damped oscillations, so fitting individual patient data there could yield a quantitative metric of sympathetic damping time.","The boundary shift implies that the effective stabilization depends on the amplitude and shape of the forcing, so constructing forcing from real-time beat-to-beat data rather than a polynomial fit would make the predicted patient classifications testable in clinical recordings."],"forward_implications":["In the two-parameter plane, the stable sink region for the forced system is larger than the homogeneous prediction $\\tau_s = eD_s$, meaning some parameter pairs that would oscillate without forcing are stabilized by the blood-pressure input.","The Hopf bifurcation line $D_s = \\tau_s\\pi/2$ is unchanged by forcing, so the onset of limit cycles and divergence is governed by the same delay-to-time-scale ratio in both systems.","A control subject can sit in either the sink or stable-focus region, so stable-focus behavior by itself is not a disease marker; the POTS patient's stable-focus placement is consistent with overactive sympathetic signaling rather than with a limit cycle.","To stay physiologically relevant, model parameters should be restricted to the sink and stable-focus regions, since limit-cycle and unstable regions are not observed in practice.","The reduced two-state model preserves the qualitative stability classification of the full five-state model, so the bifurcation analysis can be carried out on the simpler system."],"supporting_citations":[{"why":"Supplies the Lambert W root analysis of the characteristic equation for linear delay systems.","marker":"[1]"},{"why":"Standard reference for stability of differential-difference equations used for the exponential-solution ansatz.","marker":"[3]"},{"why":"Provides the stiff delay-differential solver whose numerical solutions underlie the nonhomogeneous classification algorithm.","marker":"[9]"},{"why":"Background on stability boundaries for delay differential equations and the Hopf-type threshold.","marker":"[14]"},{"why":"Prior two-parameter bifurcation analysis of baroreflex delay, the direct predecessor this study extends to patient data.","marker":"[22]"},{"why":"Defines the M blood-pressure response used to classify the POTS patient and the overactive-sympathetic hypothesis.","marker":"[23]"},{"why":"The full five-state baroreflex model and nominal parameters from which the reduced two-state system is derived.","marker":"[24]"},{"why":"General results on how time delays can stabilize or destabilize systems, motivating the forcing analysis.","marker":"[29]"}],"fun_headline_variants":["Forcing stabilizes Valsalva model, Hopf line unchanged","Blood-pressure inputs widen stable region in Valsalva DDE","POTS model: forcing expands stability, Hopf bifurcation fixed","Valsalva DDE: nonhomogeneous terms shift stability regions"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The stability-boundary shift depends on a 10th-degree polynomial fit to one-second averaged blood-pressure data that was artificially extended after the maneuver; if that extension manufactures oscillations, the widened sink region and the patient classifications are artifacts of the input construction rather than intrinsic dynamics of the delay system.","fun_headline_variants_meta":{"raw":{"variants":["Forcing stabilizes Valsalva model, Hopf line unchanged","Blood-pressure inputs widen stable region in Valsalva DDE","POTS model: forcing expands stability, Hopf bifurcation fixed","Valsalva DDE: nonhomogeneous terms shift stability regions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000916,"raw_usage":{"total_tokens":3913,"prompt_tokens":904,"completion_tokens":3009,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":520,"completion_tokens_details":{"reasoning_tokens":2932}},"tokens_in":520,"tokens_out":3009,"duration_ms":22744,"temperature":1.0,"reasoning_tokens":2932,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:14:18.526790+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the nonhomogeneous boundary with the forcing functions truncated at the end of the Valsalva maneuver, removing the artificial post-maneuver extension; if the overdamped-sink region no longer extends beyond $\\tau_s = eD_s$, or if the POTS subject moves to a different region, the claimed stabilization is an artifact of the constructed input rather than a property of the DDE.","supporting_citations":[{"cited_title":"Analysis of a sys tem of linear delay differential equations","cited_arxiv_id":null,"evidence_quote":"Supplies the Lambert W root analysis of the characteristic equation for linear delay systems."},{"cited_title":"Differential-Difference Equa- tions","cited_arxiv_id":null,"evidence_quote":"Standard reference for stability of differential-difference equations used for the exponential-solution ansatz."},{"cited_title":"Implementing Radau IIA metho ds for stiff delay differential equations","cited_arxiv_id":null,"evidence_quote":"Provides the stiff delay-differential solver whose numerical solutions underlie the nonhomogeneous classification algorithm."},{"cited_title":"Delay Differential Equations with Applications in Popu- lation Dynamics , volume 191","cited_arxiv_id":null,"evidence_quote":"Background on stability boundaries for delay differential equations and the Hopf-type threshold."},{"cited_title":"Modelling of the baroreflex-feedbac k mecha- nism with time-delay","cited_arxiv_id":null,"evidence_quote":"Prior two-parameter bifurcation analysis of baroreflex delay, the direct predecessor this study extends to patient data."},{"cited_title":"The utility of Valsalva maneuver in the diagnoses of orthostatic disord ers","cited_arxiv_id":null,"evidence_quote":"Defines the M blood-pressure response used to classify the POTS patient and the overactive-sympathetic hypothesis."},{"cited_title":"A model-based analysis of auto nomic nervous function in response to the Valsalva maneuver","cited_arxiv_id":null,"evidence_quote":"The full five-state baroreflex model and nominal parameters from which the reduced two-state system is derived."},{"cited_title":"Stability and stabilization of systems w ith time delay","cited_arxiv_id":null,"evidence_quote":"General results on how time delays can stabilize or destabilize systems, motivating the forcing analysis."}],"review_version":1}