RALPH B. GUY, JR., Circuit Judge.
Defendant Connecticut General Life Insurance Company (Connecticut General) appeals from the district court’s decision to enter judgment on the administrative record in favor of plaintiff Kurt Johnson, as beneficiary, on his claim for benefits under [461]*461an ERISA-governed group supplemental life insurance policy obtained by his wife, Kristen Johnson, less than two years before her death.1 Connecticut General asserts that the district court erred by: (1) considering and allowing discovery regarding Connecticut General’s alleged conflict of interest; (2) finding that the Ohio statute addressing false statements in applications for life insurance was not preempted by ERISA; and (3) concluding that Connecticut General’s denial of the claim was arbitrary and capricious. After review of the record and the applicable law, we affirm.
I.
On November 15, 2003, during an open enrollment period, Kristen Johnson applied for additional life insurance under a policy she already had through her employer. She completed a supplemental enrollment form at the time of application, and submitted to a paramedical exam and interview on December 26, 2003. Connecticut General approved the additional coverage on January 8, 2004, without conducting a review of her medical records. The policy provided that any increase in coverage would be incontestable after the increase has been in force for two years.
On October 4, 2005, 20 months after the increase was approved, Kristen Johnson was sent to the emergency room, went into a coma, and died the next day. The death certificate identified cardiac arrest and pulmonary embolism as causes of death, and noted that hypertension was a contributing condition. Kurt Johnson, Kristen’s husband, made a claim for benefits in November 2005. Connecticut General paid the $88,000 benefit due under the original insurance policy, but took the claim for the $174,000 in additional insurance benefits under review. Claim Specialist Linda Ca-navan requested three years of medical records from Drs. Coleflesh and Balias, and then asked plaintiff to identify any other doctors his wife had seen.2 Plaintiff called and provided the names of Drs. Morisetty and Parepally. With records from Drs. Coleflesh, Morisetty, and Pare-pally, the claim was sent to medical underwriting for review in January 2006.
In a March 20, 2006 letter, Connecticut General informed plaintiff of its decision to rescind the additional coverage on the grounds that the insured had made material misrepresentations in the supplemental enrollment form. Specifically, Connecticut General explained that although Kristen Johnson had disclosed a history of hypothyroidism diagnosed in April 2002,
[t]he medical records from Riverside Medical of Ohio indicated that she was diagnosed with hypertension in 2002; palpitations; asthma and shortness of breath documented in records from July 2, 2002 and January 16, 2003 and poly-cystic ovarian disease. Kristen Robinson-Johnson had the opportunity and obligation to disclose her full medical history at the time of application, in response to Questions B, C, E and G and a second opportunity in response to Questions 5, 9 and 10 at the time of her paramedical exam.
Connecticut General concluded that if it had been aware of her medical history, the [462]*462additional coverage would not have been approved. In May 2006, plaintiff, through counsel, appealed from the denial and disputed that the medical records referenced by Connecticut General reflected prior diagnoses of hypertension, palpitations, or cystic ovarian disease. The shortness of breath, plaintiff argues, was determined not to. be asthma but rather ventilatory muscle weakness due to the hypothyroidism.
Connecticut General denied plaintiffs appeal on June 26, 2006, reiterating the conditions that had not been disclosed and stating more precisely that Kristen Johnson had “advised her physician that she was hospitalized in 2002 for hypertension.” This time, Connecticut General specified that coverage would have been denied if the insured had disclosed “her full medical history of hypothyroidism, hypertension and palpitations.” Connecticut General would later concede that the insured had no history of palpitations or polycystic ovarian disease prior to the application for additional insurance. In addition, Connecticut General clarified that it was actually the combination of “hypothyroidism” and “hypertension” that would have resulted in an automatic denial of coverage.
This action, filed in January 2007, asserted state law claims for declaratory judgment and breach of contract. On Connecticut General’s motion to dismiss, the district court agreed that the state law claims were preempted by ERISA’s enforcement provisions; found that plaintiff should be permitted to amend to assert an ERISA claim for benefits; and concluded that an Ohio statute nonetheless provided the applicable “rule of decision” for plaintiffs ERISA benefit claim. After discovery concerning Connecticut General’s apparent conflict of interest and briefing on the merits, the district court concluded that Connecticut General’s decision to rescind coverage was arbitrary and capricious. The district court granted plaintiffs motion for judgment on the administrative record, and this appeal followed.
II.
Both the district court and this court review a decision denying ERISA benefits de novo, unless — as is the case here — the plan gives the administrator discretionary authority to determine eligibility for benefits or to construe the terms of the plan. Kalish v. Liberty Mut./Liberty Life Assur. Co., 419 F.3d 501, 506 (6th Cir.2005). When the administrator is given such discretion, the denial of benefits is reviewed under the arbitrary and capricious standard. Id. (citing Firestone Tire & Rubber Co. v. Bruch, 489 U.S. 101, 115, 109 S.Ct. 948, 103 L.Ed.2d 80 (1989)). Where there is an inherent conflict of interest, the standard is not altered but the conflict must be weighed as a factor in determining whether the decision was arbitrary and capricious. Id.; see also Metro. Life Ins. Co. v. Glenn, -U.S.-, 128 S.Ct. 2343, 171 L.Ed.2d 299 (2008).
A. Application of Ohio Rev.Code § 3911.06
Connecticut General argues that the district court erred in applying this state law as the “rule of decision” in its review of the ERISA claim for benefits. There is no dispute that Ohio Rev.Code § 3911.06 “relates to” an employee benefit plan and would be preempted by § 514(a) of ERISA unless exempted from preemption by § 514(b)(2)(A) of ERISA, which “saves” from preemption the “law of any State which regulates insurance, banking, or securities.” 29 U.S.C. § 1144(a) and (b)(2)(A). Whether ERISA preempts a particular state statute is a question of law that we review de novo. Crabbs v. Copper-[463]*463weld, Tubing Prods. Co., 114 F.3d 85, 89 (6th Cir.1997).
The ERISA saving clause has garnered the attention of the United States Supreme Court in UNUM Life Insurance Co. v. Ward, 526 U.S. 358, 119 S.Ct. 1380, 143 L.Ed.2d 462 (1999); Rush Prudential HMO, Inc. v. Moran, 536 U.S. 355, 122 S.Ct. 2151, 153 L.Ed.2d 375 (2002); and Kentucky Association of Health Plans, Inc. v. Miller, 538 U.S. 329, 123 S.Ct. 1471, 155 L.Ed.2d 468 (2003). In Miller, the Court made what it called a “clean break” from its previous reliance, albeit to varying degrees, on three factors used to determine whether a state law regulates the “business of insurance” for purposes of the McCarran-Ferguson Act. Miller, 538 U.S. at 341, 123 S.Ct. 1471. The Court explained that its reliance on the three McCarran-Ferguson factors in the ERISA context had “misdirected attention, failed to provide clear guidance to lower federal courts, and ... added little to the relevant analysis.” Id. at 339-40, 123 S.Ct. 1471. Articulating a new test, the Court in Miller held that to be deemed a law that “regulates insurance,” a state law must satisfy two requirements: (1) “the state law must be specifically directed toward entities engaged in insurance”; and (2) “the state law must substantially affect the risk pooling arrangement between the insurer and the insured.” Id. at 342, 123 S.Ct. 1471.
Applying the test from Miller, the district court found that both requirements were met in the case of Ohio Rev.Code § 3911.06, which provides that:
No answer to any interrogatory made by an applicant in his application for a policy shall bar the right to recover upon any policy issued thereon, ... unless it is clearly proved that such answer is willfully false, that it was fraudulently made, that it is material, and that it induced the company to issue the policy, that but for such answer the policy would not have been issued, and that the agent or company had no knowledge of the falsity or fraud of such answer.
In doing so, the district court acknowledged this court’s decision in Davies—which held that a similar Ohio statute fell outside the savings clause—but found that the reasoning in Davies had been implicitly overruled by the Supreme Court’s subsequent decisions in Ward and Miller.3 Davies v. Centennial Life Ins. Co., 128 F.3d 934 (6th Cir.1997). For the reasons that follow, we agree.4
1. Specifically Directed
The first requirement under Miller is that the state law be “specifically directed toward entities engaged in insurance.” 538 U.S. at 342, 123 S.Ct. 1471. In Davies, this court applied the previously articulated test that asked, first, whether the state law was “‘specifically directed’ toward the insurance industry.” Davies, 128 F.3d at 941 (quoting Pilot Life Ins. Co. v. Dedeaux, 481 U.S. 41, 50, 107 S.Ct. 1549, [464]*46495 L.Ed.2d 39 (1987)). This court noted that although § 3923.14 appeared, at first glance, to be specifically directed at the insurance industry, that was not necessarily the case since the statute had “its roots firmly planted in the general principles of Ohio contract law.” Id. Having said that, however, the court in Davies concluded only that this was a “close question.” As the district court recognized, this reasoning is undermined by the Supreme Court’s subsequent decision in Ward.
Specifically, the Court in Ward concluded that California’s notice-prejudice rule— which prevents an insurer from denying a claim as untimely unless it was prejudiced by the delay — was saved from preemption by § 514(b)(2)(A). Applying the first prong of the pr e-Miller test, the Court distinguished the Mississippi law that was found not to be specifically directed toward the insurance industry in Pilot Life and held that California’s notice-prejudice rule was specifically directed toward the insurance industry. “It is no doubt true that diverse California decisions bear out the maxim that ‘law abhors a forfeiture’ and that the notice-prejudice rule is an application of that maxim. But it is an application of a special order, a rule mandatory for insurance contracts, not a principle a court may pliably employ when the circumstances so warrant.” Ward, 526 U.S. at 370-71, 119 S.Ct. 1380.
Here, § 3911.06 is firmly applied to insurance contracts and prevents an insurer from denying coverage for an insured’s innocent misrepresentation. Plaintiff emphasizes that § 3911.06, previously designated Ohio Rev. St. § 3635, was enacted for the purpose of abrogating the general principle that an innocent misrepresentation would render an insurance policy void. See John Hancock Life Ins. Co. v. Warren, 59 Ohio St. 45, 51 N.E. 546, 547 (1898). We find that, like the notice-prejudice rule in Ward, § 3911.06 is an application of a special order, a rule mandatory for life insurance contracts, and not a principle that may be broadly applied. The district court did not err in finding, notwithstanding Davies, that § 3911.06 is a state law “specifically directed toward entities engaged in insurance.” Miller, 538 U.S. at 342, 123 S.Ct. 1471.
2. Substantially Affect Risk Pooling Arrangement
The second part of the pr e-Miller test asked whether the state law regulated the “business of insurance” as that term is defined for purposes of the McCarran-Ferguson Act. Davies, 128 F.3d at 940-41. The three McCarran-Ferguson criteria, in turn, ask whether the state law has the effect of transferring or spreading policyholder risk; whether the state law regulates an integral part of the policy relationship; and whether the state law is limited to entities within the insurance industry. Id. at 940. Finding that none of these factors were present, the court in Davies concluded in part that the statute did not “spread” policyholder risk because “[a]l-though it forces the insurer to bear the legal risks associated with innocent misrepresentations in an insurance application, § 3923.14 does not alter the risks for which the insurer and insured originally contracted — specific accident and medical costs.” Id. at 942.
Miller, however, made a “clean break” from the McCarran-Ferguson factors and requires only that “the state law must substantially affect the risk pooling arrangement between the insurer and the insured.” Miller, 538 U.S. at 342, 123 S.Ct. 1471 (emphasis added). The Court in Miller held that, by expanding the number of providers from whom an insured may receive health services, Kentucky’s “any willing provider” statute “alter[s] the [465]*465scope of permissible bargains between insurers and insureds in a manner similar to the mandated-benefit laws we upheld in Metropolitan Life, the notice-prejudice rule we sustained in [Ward ], and the independent-review provisions we approved in Rush Prudential.” 538 U.S. at 338-39, 123 S.Ct. 1471.
Particularly pertinent to this case is the Court’s clarification that this requirement “does not require that the state law actually spread risk.” Id. at 339 n. 3, 123 S.Ct. 1471 (emphasis added). Specifically, the Court added that “[tjhe notice-prejudice rule governs whether or not an insurance company must cover claims submitted late, which dictates to the insurance company the conditions under which it must pay for the risk that it has assumed. This certainly qualifies as a substantial effect on the risk pooling arrangement between the insurer and insured.” Id. Similarly, § 3911.06 alters the scope of permissible bargains by dictating the conditions under which the insurer may deny recovery for misrepresentations in the application for life insurance. The district court did not err in finding that the second Miller requirement was met in this case.5
3. Conflict Preemption
Finally, Connecticut General correctly argues that the saving clause must be interpreted in light of the preemptive force of the comprehensive enforcement scheme in § 502(a) of ERISA. 29 U.S.C. § 1132(a). As recognized in Pilot Life and Rush Prudential, “[u]nder ordinary principles of conflict pre-emption, ... even a state law that can arguably be characterized as ‘regulating insurance’ will be preempted if it provides a separate vehicle to assert a claim for benefits outside of, or in addition to, ERISA’s remedial scheme.” Aetna Health, Inc. v. Davila, 542 U.S. 200, 217-18, 124 S.Ct. 2488, 159 L.Ed.2d 312 (2004). There is no suggestion, however, that application of § 3911.06 as a rule of decision in any way duplicates, supplements, or supplants ERISA’s civil enforcement remedies. Conflict preemption under § 502(a) is not implicated when the state law simply supplies the relevant rule of decision in an ERISA claim for benefits under § 502(a)(1)(B). Ward, 526 U.S. at 376-77, 119 S.Ct. 1380.
B. Conflict of Interest
An inherent conflict of interest exists when, as in this case, an insurance company is both the administrator determining eligibility for benefits and the insurer responsible for paying the benefits out of its own pocket. Glenn, 128 S.Ct. at 2346. The Court in Glenn made clear that the significance of the conflict to that determination will depend on the circumstances of each case. Id. at 2346. “[A]ny one factor will act as a tiebreaker when the other factors are closely balanced, the degree of closeness necessary depending on the tie-breaking factor’s inherent or case-specific importance.” Id. at 2351. A conflict of interest “should prove more important (perhaps of great importance) where circumstances suggest a higher likelihood that it affected the benefits decision” and “less important (perhaps to the vanishing point) where the administrator has taken active steps to reduce potential bias and to promote accuracy.” Id. Conceding that such a conflict existed in this case, Connecticut General makes two distinct arguments: first, that it was error to allow [466]*466discovery concerning the conflict; and, second, that it was error to take the conflict into consideration absent proof establishing that the conflict affected the benefit decision.
The Court in Glenn rejected the latter argument and held that a reviewing court should consider the conflict as one of many factors in determining whether the benefit denial was an abuse of discretion. Id. at 2346; see also id. at 2353 (Roberts, C.J., dissenting) (“I would instead consider the conflict of interest on review only where there is evidence that the benefits denial was motivated or affected by the administrator’s conflict”). In addressing how the conflict ought to be taken into account, the Court provided the following admonition:
Neither do we believe it necessary or desirable for courts to create special burden-of-proof rules, or other special procedural or evidentiary rules, focused narrowly upon the evaluator/payor conflict. In principle, as we have said, conflicts are but one factor among many that a reviewing judge must take into account. Benefits decisions arise in too many contexts, concern too many circumstances, and can relate in too many different ways to conflicts — which themselves vary in kind and in degree of seriousness — for us to come up with a one-size-fits-all procedural system that is likely to promote fair and accurate review. Indeed, special procedural rules would create further complexity, adding time and expense to a process that may already be too costly for many of those who seek redress.
Id. at 2351. Accordingly, it was not error for the district court to consider the conflict as a factor in determining whether the denial of benefits was an abuse of discretion -without requiring proof that the conflict affected the benefit decision.
We review for abuse of discretion the district court’s decision to allow limited discovery concerning the conflict of interest. Green v. Nevers, 196 F.3d 627, 632 (6th Cir.1999). In reviewing the merits of a decision denying ERISA benefits, the court is limited to the administrative record available to the administrator at the time of final decision. Wilkins v. Baptist Healthcare. Sys., 150 F.3d 609, 615 (6th Cir.1998); Marks v. Newcourt Credit Group, Inc., 342 F.3d 444, 457 (6th Cir. 2003). An exception is recognized, however, when evidence outside the record “is offered in support of a procedural challenge to the administrator’s decision, such as an alleged lack of due process afforded by the administrator or alleged bias on its part.” Wilkins, 150 F.3d at 619 (Gilman, J., concurring). “This also means that any prehearing discovery at the district court level should be limited to such procedural challenges.” Id.
We have noted in a few ERISA cases that discovery might have been appropriate under the circumstances. See Kalish, 419 F.3d at 507; Calvert v. Firstar Fin., Inc., 409 F.3d 286, 293 n. 2 (6th Cir.2005). In other cases, we have affirmed the denial of discovery and explained that a “mere allegation of bias is not sufficient to permit discovery under Wilkins’ exception.” Putney v. Med. Mut. of Ohio, 111 Fed.Appx. 803, 807 (6th Cir.2004); see also Likas v. Life Ins. Co. of N. Am., 222 Fed.Appx. 481, 486 (6th Cir.2007); Huffaker v. Metro. Life Ins. Co., 271 Fed.Appx. 493, 504 (6th Cir.2008). Although Connecticut General argues that these cases should be interpreted to impose a threshold evidentiary showing of bias as a prerequisite to discovery under Wilkins, the Supreme Court’s admonition in Glenn discouraging the creation of special procedural or evidentiary rules for evaluating administrator/payor conflicts of interest counsels against it. [467]*467That does not mean, however, that discovery will automatically be available any time the defendant is both the administrator and the payor under an ERISA plan. The limitation on discovery recognized in Wilkins is a result of the determination that matters outside the administrative record are ordinarily not relevant to the court’s review of an ERISA benefit decision. District courts are well-equipped to evaluate and determine whether and to what extent limited discovery is appropriate in furtherance of a colorable procedural challenge under Wilkins. Plaintiff offered more than a mere allegation of bias, and the district court did not abuse its discretion by allowing plaintiff to conduct limited discovery concerning the conflict that existed in this case.6
C. Abuse of Discretion
Applying § 3911.06 as the substantive rule of decision, and considering the conflict in this case, we must determine whether the decision to rescind coverage for material misrepresentation was arbitrary and capricious. The decision will not be found to be arbitrary and capricious “so long as ‘it is possible to offer a reasoned explanation, based on the evidence, for a particular outcome.’ ” Kalish, 419 F.3d at 506.
[A]n insurer can satisfy the requirements of [§ ] 3911.06, so as to establish an answer to an interrogatory by an applicant as a bar to recovery upon a policy, by clearly proving that
(1) the applicant willfully gave a false answer
(2) such answer was made fraudulently
(3) but for such answer the policy would not have been issued and
(4) neither the insurer nor its agent had any knowledge of the falsity of such answer.
Jenkins v. Metro. Life Ins. Co., 171 Ohio St. 557, 173 N.E.2d 122, 125 (1961). When there is no evidence that the insured made an honest mistake, the insured’s false statement will be deemed to have been willfully false and fraudulently made as a matter of law. Id. at 126. Recent decisions interpreting § 3911.06 have held that an insurer is not required to prove fraudulent intent — only that the applicant knowingly provided a false answer. Spencer v. Minn. Life Ins. Co., 493 F.Supp.2d 1035, 1038 (S.D.Ohio 2007) (citing Blakely v. Security Dollar Bank, No.2000-T-105, 2001 WL 848581 at *3 (Ohio Ct.App.2001)).7
When the insured died within the two-year contestability period, Connecticut General conducted a misrepresentation review and rescinded the additional coverage for failure of the insured to disclose her full medical history. The initial denial was based on Kristen Johnson’s failure to disclose that she was diagnosed with hypertension in 2002 and that she had a history of palpitations, asthma and shortness of breath, and polycystic ovarian disease. The final denial in June 2006 indicated more specifically that coverage would have been denied if Kristen Johnson had dis[468]*468closed “her full medical history of hypothyroidism, hypertension and palpitations.” In the course of the litigation, Connecticut General conceded that Kristen Johnson had no prior history of palpitations or po-lycystic ovarian disease.8 Connecticut General also clarified that it was specifically the combination of hypothyroidism and hypertension that would have resulted in a complete denial of coverage if it had been disclosed. Since her hypothyroidism was disclosed, the critical question for us is whether it was arbitrary and capricious, taking into account the existing conflict of interest, for Connecticut General to conclude from the evidence before it that Kristen Johnson knowingly failed to disclose a prior diagnosis or treatment for “high blood pressure” in the application for additional coverage.
The enrollment form, signed by Kristen Johnson on November 15, 2003, asked, in pertinent part, whether: “During the last five years, has the proposed insured been diagnosed with or received treatment by/ from a member of the medical profession for any of the conditions listed in questions below?” Kristen Johnson marked “No” in response to Question B, which listed conditions including “high blood pressure,” “shortness of breath,” and “heart murmur.” She answered “Yes” to Question C, which asked about “thyroid disorder,” and other conditions like “asthma,” or “other disease or disorder of the respiratory tract.” In the space provided, she stated that she had hypothyroidism starting in April 2002, and was receiving ongoing treatment with Synthroid. In signing, the insured attested that her statements were true and correct to the best of her knowledge. The medical ques-tionaire completed on December 26, 2008, asked: “Have you ever had or been told you had high blood pressure, heart or lung disease?” The insured indicated “No,” and listed under details “Appendectomy— 1-82” and “Hypothyroidism — 4-02.” Coverage was approved.
To be sure, there is no dispute that in August 2004, after the application and before her death, Kristen Johnson was diagnosed with hypertension and began treatment with Diovan. Dr. Coleflesh made the diagnosis in August 2004, and Dr. Morisetty indicated hypertension as a current diagnosis in September 2004. These reports are part of the administrative record. Dr. Morisetty’s report also related that the insured had a history of “erratic blood pressure” in April 2002. Dr. Morisetty’s report specifically related that Kristen Johnson had seen “an endocrinologist because of erratic blood pressure and feeling dizzy with near syncope [fainting] after the [469]*469first pregnancy,” but said “[t]hey could not find a reason for this.” This is the history that Connecticut General claims the insured knowingly failed to disclose in response to the questions about high blood pressure.
Indeed, Connecticut General’s denial letter referenced a medical report dated July 2, 2002, in which Dr. Nashawati stated that “during her pregnancy, she had a lot of problems with blood pressure and had a lot of fainting and [a] workup for that including tilt table test and echocardio-gram.” The same report went on to state that: “Her labile blood pressure with fainting is now gone.” Neither this report, nor Dr. Morisetty’s report two years later, states that she had or was treated for hypertension in 2002. Bather, these reports reflect that Kristen Johnson experienced “problems with blood pressure,” experienced “erratic” or “labile” blood pressure, and had symptoms of near syncope or fainting.
Connecticut General’s underwriting materials define “hypertension” as “sustained elevation in blood pressure above that considered acceptable for the individual’s age and gender.” Also, normal blood pressure is defined as less than 140/85 mmHg. Connecticut General’s medical underwriting guidelines also expressly distinguish sustained hypertension from labile hypertension, stating that “[ljabile (or ‘white coat’) hypertension” is defined by “a fluctuating blood pressure with some levels clearly elevated,” and noting that the condition “may evolve into sustained hypertension.” The underwriter who recommended rescission testified in discovery that “labile” blood pressure refers to changing blood pressure, up or down, and acknowledged that having labile blood pressure is not the same as having hypertension. Further, as the district court observed, the administrative record included three normal blood pressure readings taken during the 18-month period preceding her application for additional coverage. None of them were recorded while she was prescribed medication for high blood pressure. Specifically, the administrative record reflects a blood pressure reading of 102/62 on July 2, 2002; a blood pressure reading of 114/78 on January 16, 2003; and a blood pressure reading of 124/78 at the time of the paramedical exam on December 26, 2003. It was not until after the insurance was in force that plaintiff was diagnosed with hypertension and began to take medication to control her high blood pressure.
It is apparent from the wording of the final denial — stating that the insured had “advised her physician that she was hospitalized in 2002 for hypertension” — that Connecticut General was relying on the records of Dr. Parepally. Specifically, a November 29, 2004 letter sent by Dr. Pa-repally to Dr. Morisetty reported that the insured had preeclampsia with her first child 2j£ years earlier, had been taking Diovan for high blood pressure for the last four months, and “was diagnosed with hypertension in 2002.” A handwritten notation on a medical intake form of the same date noted under hospitalizations: “1981 Apendeet” and “2002 Hypertension.” Another handwritten chart, however, noted more vaguely that she had a 2002 hospitalization for “Blood Pressure.”
• As the district court emphasized, these references were based on the history recorded in November 2004, after she had been diagnosed and was being treated for hypertension, and those brief notations in the history alternately referenced “hypertension” and “blood pressure.” Considering that the insured apparently explained more fully to Dr. Morisetty in September 2004, just a few months before seeing Dr. Parepally, that what she experienced were [470]*470problems with erratic blood pressure, dizziness, and near syncope, we find that the statement in Dr. Parepally’s report cannot provide a reasoned basis for concluding that the insured knowingly failed to disclose that she had been hospitalized for or diagnosed with hypertension or high blood pressure in 2002. We find that Connecticut General’s reliance on these records, to the exclusion of more specific histories provided regarding the same episode and three normal blood pressure readings over the 18-month period that preceded the application, was arbitrary and capricious.
AFFIRMED.