OPINION
THOMAS L. LUDINGTON, District Judge.
Appellant John Schwalm (“Schwalm”) contends that Appellee Guardian Life Insurance Company of America’s (“Guardian”) decision to terminate his long-term disability benefits was arbitrary and capricious.
See
Employee Retirement Income Security Act of 1974 (“ERISA”) 29 U.S.C. §§ 1001-1461 (2006). The district court reviewed the administrative record, determined that Guardian’s decision was supported by substantial evidence, and dismissed Schwalm’s complaint. We AFFIRM.
I.
Schwalm injured his back on July 2, 1999. The initial “twinge,” felt while lifting luggage on a business trip, turned out to be a herniated disk that required surgery. The first surgery in November 1999 made matters worse, and Schwalm returned for a second surgery on December 31 of that year. The second surgery initially provided relief, but Schwalm’s back pain steadily increased. In 2002, he sought a second opinion from Dr. Louis Keppler, an orthopedic surgeon at the Cleveland Clinic. On December 5, 2002, Dr. Keppler performed a spinal fusion surgery on Schwalm’s L-4 and L-5 vertebrae. Recovery was difficult because of complications during the surgery and continued back pain, but Schwalm attempted to return to work on a limited basis in February of 2003. At the time, Schwalm was employed as the chief administrative officer of Acero, Inc., a technology company based in Cleveland, Ohio. He had previously served as chief executive officer of Acero, and as an executive with other companies. Schwalm earned a base salary of $140,000 per year.
On March 13, 2003, Schwalm completed an application for long-term disability benefits from Guardian. Schwalm was covered by a group insurance policy issued to Acero employees. Guardian provided the employees with life insurance and long-term disability insurance pursuant to an ERISA-qualifying plan. The plan provides two definitions of disability. The first definition is applicable during the two years immediately following the date the insured applies for benefits. In this case, that period began on March 13, 2003 and concluded on March 13, 2005. During this period, which Guardian refers to as “the own occupation period,” the insured is “disabled” if “physical, mental or emotional limits caused by a current
sickness
or
injury”
leave the insured person “not able to perform, on a full-time basis, the major duties of [his]
own occupation.”
After March 13, 2005, the second definition applies. During this period, the insured is disabled if he remains “not able to perform, on a full-time basis, the major duties of any
gainful work.”
The plan also defines the term “gainful work” as “[w]ork for which you are or may become qualified
by: (a) training; (b) education; or (c) experience. Such woi'k must also be consistent with the level of your
insured earnings.”
Guardian reviewed the application, determined Schwalm was disabled pursuant to the “own occupation” definition, and mailed the first benefit payment on May 1, 2003. Schwalm received $7,000 per month — sixty percent of his base salary.
Initially, Schwalm and Guardian both believed that he would be able to return to work in July 2003, but persistent back pain and concentration problems associated with the analgesic pain medications he had been prescribed delayed his return. Although doctors were unable to identify the cause of the pain in an MRI, Schwalm and Dr. Keppler scheduled a procedure for March 18, 2004 to remove the fusion hardware from his back. Dr. Keppler believed that the presence of the hardware, as well as a substantial amount of scar tissue around it, may have been causing the continued pain. The surgery, however, did not significantly improve Schwalm’s condition and limitations due to back pain continued. Guardian’s records show that Schwalm’s anticipated return to work was repeatedly delayed because of the pain and cognitive side effects of the medication.
Following the surgery, Dr. Keppler also performed a nerve block, but the procedure did not significantly reduce Schwalm’s pain. Dr. Keppler eventually referred Schwalm to Dr. Edwin Covington in the Cleveland Clinic’s pain management division for an initial appointment on February 8, 2005. Dr. Covington diagnosed Schwalm with post-laminectomy syndrome with moderate functional impairment and prescribed Maprotiline for pain and quinine for leg cramps. At various times, Schwalm’s doctors also prescribed Kadian, Vicodin, Effexor, methadone, Lyrica, and Oxycontin to help manage Schwalm’s pain, as well as other medications to aid sleep and control side effects. After consulting with Dr. Covington for several months, Schwalm remained unable to return to work because of the pain and the side effects associated with the medications.
In April 2005, Guardian scheduled an independent medical exam (“IME”) with Dr. Michael Harris for May 16, 2005. Guardian also arranged for surveillance of Schwalm’s activities for two days before and two days after the IME. Guardian was apparently concerned Schwalm might not be disabled because he was difficult to reach on the telephone. Schwalm also reported to Guardian that he would occasionally go to the office even though he was not working.
In the IME report, Dr. Harris reviewed the notes and records from Schwalm’s visits with Dr. Keppler, a questionnaire concerning Schwalm’s daily activities, X-rays from September 2003, and an MRI from October 2003. Dr. Harris also discussed Schwalm’s reported symptoms, which included a pain-level range between four and eight out of ten. Schwalm described the pain as “aching,” and sometimes “sharp” or “stabbing.” The pain began in his lower back and “radiat[ed]” down his right buttock, thigh, lower leg, and foot. Schwalm indicated he could sit for sixty to ninety minutes, stand in one position for ten to fifteen minutes, or walk for forty-five minutes. Schwalm told Dr. Harris he needed to lie down once or twice per day to rest.
Dr. Harris’s physical examination noted “no obvious distress” or “exaggerated pain behaviors or magnification.” He observed some tenderness and reduced sensation in Schwalm’s lower back. He described Schwalm’s gait as “independent” and his strength as “normal,” but he also noted limits to his range of motion. Dr. Harris indicated that Schwalm had “significant
physical limitations and loss of function” due to the injury and subsequent surgeries. He noted that Schwalm could “function at a sedentary level” with “significant aceom[m]odations,” including the ability to lie down several times per day to rest his back. Dr. Harris also observed that “the opioids do interfere with his level of cognition and although he functions independently, I am not convinced that he can make the appropriate decisions that need to be made at the level of CEO of a technology company.” Dr. Harris concluded, “I do not believe that Mr. Schwalm is capable of working as a CEO at full capacity.”
Two months after the May 16, 2005 exam, Dr. Harris also reviewed the surveillance video and report that was prepared on May 11, 16, and 17, 2005. The video depicted Schwalm
moving around fairly comfortably; bending over into a car and carrying what appeared to be a black case with a lap top over his shoulder without any significant discomfort. He was seen carrying a small child without any significant discomfort. He was also seen sitting in a coffee house working at his lap top for a 12 minute stretch at one point and then another 25 minutes, after that.... There were no signs of discomfort. He had almost no position change except for occasionally bending his knees or extending his knees.
Dr. Harris was particularly surprised by Schwalm’s activities at the coffee house, which also included a twenty-one minute period there on a different date. Dr. Harris noted that it “did strike” him that with his reported pain Schwalm “was able to [sit] in a hard wooden chair at a coffee house” on both May 11 and May 16, 2005.
Nevertheless, Dr. Harris affirmed his conclusion that Schwalm would be unable to resume his position as chief executive officer of a technology company. Although Schwalm was “functioning at a level higher than one would expect” given the reported pain level of four to eight, he did not alter his conclusion that Schwalm’s cognitive limitations would prevent him from resuming his chief executive duties full time.
Clearly, he can easily work at a sedentary level doing some type of management job especially if he is allowed to sit in a comfortable leather chair with adjustable seat height and arm rest with a high back chair and as long as he is allowed to change positions frequently. However, the fact is he is s/p multiple spine surger[i]es, has significant scarring, and is on significant doses of opioid medication which can alter cognitive abilities. I still question his capability of functioning at the level of a chief executive officer, 8 hrs a day, which requires quick decision making, the results of which can have drastic financial consequences.
On June 22, 2005, a Guardian medical specialist reviewed the results of the IME, and updated the review on August 11, 2005, after receiving Dr. Harris’s supplementary surveillance report. The medical specialist concluded that Schwalm was still disabled, even though he could “perform some type of managerial position,” because the “complexity” of the duties associated with high-level business management may be impossible while on high doses of opioid medications. Guardian continued to pay Schwalm long-term disability benefits.
On June 1, 2005, Schwalm’s application for disability benefits from the Social Security Administration was denied. The Social Security Administration determined that Schwalm’s “condition is not severe enough” to keep him from working. The denial letter noted that Schwalm “re-tainted] the ability to stand, sit, and walk for a normal work day.” Despite the pain,
the Social Security Administration observed that Schwalm was “able to move about and use [his] arms, legs, hands, and back to perform some limited types of activities.” Social security benefits were denied again on September 11, 2006.
Following the IME and the Social Security Administration’s initial denial, Guardian continued to pay Schwalm’s disability benefits for another year without raising new questions about his continued eligibility. Then, on August 30, 2006, Guardian received treatment records compiled by Dr. Covington’s office from February 2005 through August 2006. The records indicated that Schwalm was spending a significant amount of his time working in an executive management and finance role at a technology start-up called Peritus Technologies, LLC (“Peritus”). Schwalm was spending up to twelve hours per day at the company’s headquarters. He did, however, leave early about three times a week to exercise and he would lie down to rest on a cot in the offices three to five times per day for up to one hour at a time.
Guardian also received a “Cooperation Agreement” between Schwalm and Peritus with an effective date of July 1, 2005. Peritus was a new company that hoped to profit by connecting technology companies with qualified technology-industry employees. The Cooperation Agreement provided that Schwalm was responsible for all the company’s “executive management and financing functions including, business planning, bookkeeping, governance, legal administration, marketing and all other non-Business Development functions.” Schwalm’s partner in the new venture, another former Acero employee named Kevin Marquart (“Marquart”), was responsible for all business development, including “technology resources, establishing new clients and developing strategic partners.” The Cooperation Agreement acknowledged Schwalm’s injury and the side effects arising from the pain medications. It emphasized that Schwalm would require a very flexible schedule and significant accommodations, including frequent rest breaks of up to an hour and a cot to lie down on. Still, Schwalm agreed to spend at least 160 hours per month on his “assigned business responsibilities.” The Cooperation Agreement provided that Schwalm’s normal business hours would be six in the morning until six in the evening, “which allows four hours for rest and rehabilitation activities each day.”
Schwalm and Marquart agreed to equally divide the company’s equity, but with different salary arrangements. Neither partner would receive any salary until the company’s earnings before interest, taxes, and debt service exceeded the company’s other operation expenses. Then, Marquart would receive his entire “targeted” yearly salary of $100,000 before Schwalm would begin receiving his “targeted” yearly salary of $115,000. The partners acknowledged that their salaries depended on revenue and it “could be six months or longer” before they were paid at all. The Cooperation Agreement provided for a 270-day probationary period for Schwalm, and three “milestone assessment” meetings to evaluate Schwalm’s performance.
The records Guardian received in August 2006 also contained repeated references to Schwalm’s alleged “denial” of cognition and alertness problems related to the pain medications. Based on this new information, Guardian began another review of Schwalm’s eligibility for benefits and made the initial arrangements for additional medical tests, including a functional capacity evaluation (“FCE”).
On September 5, 2006, Schwalm visited Dr. Anantha Reddy, also of the Cleveland Clinic, for a disability assessment. Schwalm reported to Dr. Reddy that he
was unable to return to his position as a chief executive officer because of chronic back pain and “his mental condition due to medications” used to treat the pain. During the visit, Dr. Reddy diagnosed “[m]eehanical back pain, failed back syndrome, chronic pain, intact motor strength, [and] chronic right leg pain.” Dr. Reddy suggested an MRI, X-ray, and EMG, but Schwalm declined due to the expense. Dr. Reddy also recommended an FCE, which Schwalm indicated would depend on insurance coverage.
On November 29, 2006, Schwalm underwent a FCE performed by Eric O’Brien, a physical therapist. O’Brien concluded that Schwalm
was able to demonstrate physical tolerances within the medium category of work on both the occasional and frequent basis. At minimum[,] he can perform at a light physical demand level for a typical 8 hour day according to the Standards of the US Department of Labor. A job description was provided with the referral and indicated that the patient’s job was in the sedentary classification. The patient is able to meet the demands of a sedentary job. He was consistent for 15 of 15 measures of isometric consistency.
O’Brien noted in the section of his report titled “subjective history” that mental side effects from the prescribed pain medications limited Schwalm’s ability to work. He also noted that Schwalm reported “difficulty thinking and communicating clearly.” O’Brien’s objective testing, however, focused solely on Schwalm’s physical abilities. O’Brien concluded, unequivocally, that Schwalm was physically able to perform the job functions of a chief executive officer.
On April 18, 2007, Dr. Covington reviewed the FCE report at Guardian’s request and concluded, “to my eye it seems reasonably correct.” He also observed that “Schwalm scored at the medium level. However, they certified him as fit to function at the light level of physical capacity given the fact that he works many hours a day but spendfs] much of it lying down because of intractable pain. This seems a reasonably appropriate determination.”
On February 26, 2007, Guardian received a vocational assessment (“VA”) performed by Jacqueline M. Pickering, a certified rehabilitation counselor and licensed professional counselor at The Sierra Group. Pickering reviewed Schwalm’s employment history, the results of the FCE, a letter Schwalm wrote to Guardian on September 4, 2006, and the Peritus Cooperation Agreement. Pickering observed, based on the FCE and Schwalm’s reported work activities, that Schwalm was capable of working an eight hour day and that his principal complaint, cognitive difficulties, was not supported by the objective evidence in the record. Pickering’s report emphasized that Schwalm had drafted the Cooperation Agreement and the lengthy September 4, 2006 letter explaining his condition, which were activities that seemed to require a relatively high-level of mental capacity. Pickering also noted in the VA that the terms of the Cooperation Agreement were unusual, including the lengthy probationary period, the absence of records concerning the “milestone assessment” meetings, and the provision for full compensation for Marquart before any compensation for Schwalm. The VA concluded: “There also does not appear to be any
objective evidence
of cognitive difficulties, nor any mention of basic coping techniques . that were attempted to lighten the cognitive load.”
On September 11, 2007, Guardian completed its review of Schwalm’s records and determined that he was no longer “disabled” within the meaning of the long-term
disability insurance policy. Guardian discontinued his benefits as of that date. In a letter explaining its decision, Guardian reviewed Schwalm’s medical records from February 8, 2005 through August 9, 2006, the FCE, and the VA. Guardian noted that Schwalm had been working as many as twelve hours per day, five days per week, since July 1, 2005. Guardian observed that Schwalm’s medical records reflected mild to moderate impairment due to pain, but that the pain was manageable with medications. Moreover, the records reflected consistent denials by Schwalm of cognitive side effects from the medications. Guardian noted that the FCE concluded, and Dr. Covington confirmed, that Schwalm could perform at the “light level” of physical activity. It emphasized that there was no objective evidence of cognitive difficulties in the record, and that Schwalm had been working for more than a year, albeit without compensation, in an occupation that was very similar to the one he left in 2002. It also acknowledged that Schwalm’s voluntary agreement to work uncompensated until his partner was fully compensated was “not standard business practice,” even in a probationary setting.
Schwalm promptly appealed on October 8, 2007, delivering a series of letters and emails to Guardian protesting the decision to terminate his benefits. The appeal was referred to Kelly Holmes, a Guardian “Adjudication Procedural Specialist.” Schwalm emphasized that he continued to suffer from debilitating pain and severe side effects arising-from the pain management medications. He argued that the Cooperation Agreement evidenced a “vocational rehabilitation” project that he embarked on with a friend in an effort to preserve “professional capital,” and not any ability to engage in gainful work at a salary comparable to his former salary. Schwalm emphasized that his performance at Peritus did not meet even “minimum expectations” and that the accommodations provided by the company would not be available in a competitive work environment.
Schwalm also provided “supporting documentation” to Holmes, including a “Pain Disability Questionnaire,” additional medical records, his own list of reported disabilities, and printouts from several websites with information about his conditions and the reported side effects of his medications. Schwalm also stated that he was “submitting a copy of Dr. Covington’s letter to Kelly Holmes which provides a number of affirmative statements to put proper context around statements and inferences in the Medical Records which were used inappropriately by Guardian in the initial review.” The letter was actually prepared by Schwalm himself, and the copy included in the record was never signed by Dr. Covington. Indeed, when Guardian asked Dr. Covington about the letter, Dr. Covington emphasized that Schwalm had prepared it.
Schwalm also retained counsel to assist in his appeal, and in early 2008, he underwent a vocational evaluation (“VE”) at his attorney’s request. The VE was performed by Kathleen L. Reis, a certified vocational rehabilitation counselor at All-work, Inc., in Cleveland. Reis reviewed the FCE, information from Dr. Covington, information about Schwalm’s medications, the VA, the IME from Dr. Harris, Guardian’s September 11, 2007 termination letter, Schwalm’s resume, and writing samples from before and after the disability onset date. Reis also interviewed Schwalm, his wife, and Marquart, and met with Schwalm a second time at the Peritus offices.
Reis observed that Schwalm “appeared to have word-finding difficulty,” and frequently provided rambling or off-topic an
swers to her questions. Reis also noted that he “exhibited pain behaviors,” including “weight shifting, grimacing, and unannounced changes of posture.” She reported that his post-disability writing samples were “full of effort” but “hard to follow” because they “mixed first and third person, were full of incomplete sentences, and awkward constructions.” His pre-disability writing samples, by contrast, were “concise” and “on point.”
Reis concluded that Schwalm’s position at Peritus was not “competitive employment.” She acknowledged that he may be able to meet the physical demands of his “own occupation” or other “gainful work,” but she emphasized that his cognitive limitations left him unable to perform any job with a salary similar to what he earned at Acero. She noted that the policy definition of “gainful work” requires, according to Guardian’s termination letter, that the insured be able to earn at least eighty percent of his pre-disability salary before benefits are terminated. Reis concluded that Schwalm was not capable of performing any job paying $112,000 annually, and as a result, he remains disabled within the meaning of the policy.
On January 31, 2008, Schwalm visited Dr. Nagy A. Mekhail, head of the Cleveland Clinic’s pain management division, who noted that Schwalm was no longer working and was dependent on his friends and family for income. Dr. Mekhail noted that Schwalm was “interested in exploring his interventional options” because he hoped to return to work and he believed the medications negatively impacted his concentration. Dr. Mekhail suggested that Schwalm was a “good candidate for neuromodulation,” a medical procedure involving electrical stimulation of the spinal cord to relieve pain. Dr. Mekhail also reviewed the results of an MRI performed on January 9, 2008 — Schwalm’s first MRI since 2004. He observed that the MRI showed damage at the L3/L4 level, in addition to the existing L4/L5 damage. Dr. Mekhail diagnosed Schwalm with post-laminectomy syndrome-lumbar, lumbrosacral spondylosis, and lumbrosacral neuritis. He recommended that Schwalm continue with the course of treatment prescribed by Dr. Covington.
On May 20, 2008, Schwalm’s attorney delivered an eleven-page letter to Guardian, which reviewed the record and emphasized favorable evidence. On June 12, 2008, Guardian issued its final denial of Schwalm’s long-term disability claim. In a letter signed by Holmes, Guardian noted that it had reviewed all of the additional information received during the appeals process, and that it determined the decision to stop benefits on September 12, 2007 was correct. Guardian emphasized that Schwalm is responsible for furnishing medical evidence supporting the existence of a disability, and that he had not met that burden.
On June 16, 2008, Schwalm filed a complaint in the U.S. District Court for the Northern District of Ohio, challenging Guardian’s decision to terminate his long-term disability benefits. On September 4, 2009, the Honorable Christopher A. Boyko awarded judgment to Guardian based on his review of the administrative record. Judge Boyko determined that there was substantial evidence supporting Guardian’s decision to terminate benefits, and that it was not arbitrary and capricious to give more weight to Dr. Covington’s records, the FCE, and The Sierra Group VA, than to the Allwork VE and the IME performed by Dr. Harris.
II.
The district court’s findings of fact are reviewed for clear error.
Sanford v. Harvard Indus., Inc.,
262 F.3d 590,
594-95 (6th Cir.2001). Conclusions of law, including the determination of the appropriate standard by which the plan administrator’s decisions are considered, are reviewed de novo.
Id.
Section 502(a)(1)(B) of ERISA authorizes an individual to bring an action “to recover benefits due to him under the terms of his plan, to enforce his rights under the terms of the plan, or to clarify his rights to future benefits under the terms of the plan.” 29 U.S.C. § 1132(a)(1)(B). Generally, federal courts review a plan administrator’s decision to deny benefits de novo.
Sanford,
262 F.3d at 595 (citations omitted). But, where the plan administrator reserves discretionary authority to determine eligibility and construe policy terms, the more deferential arbitrary and capricious standard of review applies.
Id.
As the parties acknowledge, the arbitrary and capricious standard applies in this case because the policy grants Guardian “discretionary authority to determine eligibility for benefits and to construe the terms of the
plan
with respect to claims.”
The arbitrary and capricious standard is “the least demanding form of judicial review of administrative action. When it is possible to offer a reasoned explanation, based on the evidence, for a particular outcome, that outcome is not arbitrary or capricious.”
Shields v. Reader’s Digest Ass’n, Inc.,
331 F.3d 536, 541 (6th Cir.2003) (quotation marks and citation omitted). The arbitrary and capricious standard requires courts to review the plan provisions and the record evidence and determine if the administrator’s decision was “rational.”
Id.
Although the evidence may be sufficient to support a finding of disability, if there is a reasonable explanation for the administrator’s decision denying benefits in light of the plan’s provisions, then the decision is neither arbitrary nor capricious.
Williams v. Int’l Paper Co.,
227 F.3d 706, 712 (6th Cir.2000). Yet the deferential standard of review does not mean courts should “rubber stamp[]” a plan administrator’s decision — a court must review the quantity and quality of the medical evidence on each side.
Evans v. UnumProvident Corp.,
434 F.3d 866, 876 (6th Cir.2006). A decision reviewed according to the arbitrary and capricious standard must be upheld if it results from “a deliberate principled reasoning process” and is supported by “substantial evidence.”
Baker v. United Mine Workers of Am. Health & Ret. Funds,
929 F.2d 1140, 1144 (6th Cir.1991).
A court may consider only that evidence presented to the plan administrator at the time he or she determined the employee’s eligibility in accordance with the plan’s terms. The court’s review is thus limited to the administrative record.
See Wilkins v. Baptist Healthcare Sys., Inc.,
150 F.3d 609, 618 (6th Cir.1998).
III.
Although Schwalm raises numerous objections to Guardian’s decision to terminate his benefits, his arguments can be divided into two categories. First, Schwalm contends that Guardian’s decision was arbitrary and capricious because Guardian did not consider all of the evidence in the record and misinterpreted other evidence, including its own policy. Second, Schwalm contends that Guardian’s decision is not supported by substantial evidence. The first category of objections focuses on the way Guardian made the decision; the second category focuses on whether that decision was correct.
A.
Schwalm contends that Guardian’s termination of his benefits was arbitrary and
capricious because the company ignored the “insured earnings” requirement in the plan’s definition of disability; did not properly consider cognitive limitations, as reported by Dr. Harris and Reis; “cherry picked” the evidence which was provided to independent evaluators, like The Sierra Group; and overemphasized the Cooperation Agreement even though Schwalm explained his employment at Peritas was not competitive. Schwalm also contends that the district court was too deferential to Guardian in light of the inherent conflict of interest created when the plan administrator is also required to pay the benefits due under the plan.
Schwalm first contends that Guardian’s final denial letter misinterpreted the plan’s second definition of disability because it did not acknowledge the salary requirement in the definition of gainful work. Under the plan, Schwalm continued to be disabled unless he could perform “gainful work” at a salary “consistent with” his insured earnings of $140,000. Schwalm acknowledges that he may be able to perform sedentary work, but not sedentary work for compensation “consistent with” $140,000 per year.
Although the policy does not define “consistent with,” it does provide a limit to “the amount of income [the insured person] may earn, or may be able to earn, and still be considered
disabled.”
It continues: “Payments from this plan will also end if [the insured person is] able to earn more than 80% of [his]
insured earnings.”
Accordingly, regardless of what “consistent with” means as it is used in the definition of “gainful work,” if Schwalm was “able to earn” $112,000 per year, he no longer qualified for long-term disability benefits under the plan.
Moreover, in its June 12, 2008, letter concluding that Schwalm was no longer disabled, Guardian gives substantial attention to Schwalm’s physical and cognitive capabilities and his ability to return to work at a salary consistent with his insured earnings. Guardian also provides a definition of gainful work, noting that it must provide a salary that is consistent with insured earnings. Guardian further acknowledges that Schwalm reported to his doctors and to Allwork’s Kathleen Reis that he had “lost the ability to earn a living” and that he was working as many as fifty hours per week for no salary. Guardian dismissed Schwalm’s contention that he could no longer earn a comparable salary, concluding that Reis’ statement that he “lost his ability to earn a living is not supported by medical information we have in [the] claim file.”
Notably, the Cooperation Agreement provided for a salary target at Peritas of $115,000 per year. Although Schwalm had not yet begun to draw a salary, the agreement provides some indication of Schwalm’s own salary expectations and his own perceived ability to perform effectively at a high level. The definition of disability provided by the plan does not require that the insured person actually earn a salary consistent with his pre-disability salary, only that he “is able to perform ... the major duties” of such an occupation. Actual earnings are relevant elsewhere in the policy, but not to the determination of whether the insured person is disabled.
Guardian explained in the letter that the decision to stop benefit payments was not based on any actual earnings, but based on the decision that Schwalm was no longer disabled within the meaning of the plan. The decision could not be based on actual earnings because Schwalm did not provide tax returns or other documentation to confirm his earnings during the disability period, despite repeated attempts by Guardian to gain access to the information and Schwalm’s ongoing obligation under the
plan to provide the information. As Guardian explained in the letter, “our review of Mr. Schwalm’s appeal does not include a final decision regarding any income earned by him during the period of disability claimed and how such earnings affect his claimed benefits under this plan.” Guardian specifically explained that it was discontinuing benefits because the medical evidence demonstrated he was able to return to gainful employment at a salary consistent with his insured earnings.
Schwalm relatedly contends that Guardian ignored evidence of cognitive limitations that would have prevented him from returning to an executive level job. Schwalm argues that even if, as Dr. Harris concluded after the IME, Schwalm could return to some type of management position, the cognitive limitations prevented the type of high-level thinking required of chief executives or any other position paying $140,000 per year. Schwalm relies primarily on Dr. Harris’ IME report, his own affidavit, the affidavit of his partner, and the VE performed by Reis.
Guardian emphasizes in response that Schwalm has the burden of proving continued disability, and that the portions of the record that address cognitive impairments are from 2005 and outdated. Perhaps the single most important piece of evidence, from Schwalm’s point of view, is the All-work VE that was completed by Reis on March 13, 2008. Reis interviewed Schwalm, his wife, and his partner, but she performed no objective testing. She is a certified rehabilitation counselor who is qualified to evaluate Schwalm’s ability to perform as a chief executive based on a given set of physical and mental limitations, but she is not qualified to diagnose physical or mental limitations. She is not a medical professional. Moreover, Guardian did not ignore Reis’ assessment of Schwalm’s ability to return to work. Rather, it concluded that it was unsupported by objective evidence and rejected it in favor of other record evidence. The conclusion was supported by substantial evidence, including the notes prepared by Dr. Covington’s nurse repeatedly noting Schwalm’s long hours and denial of cognition problems, The Sierra Group VA noting Schwalm’s physical ability to work and lack of objective signs of cognitive problems, and Schwalm’s own work product.
Schwalm next contends that Guardian “cherry picked” among the available evidence, emphasizing those parts of the administrative record that were favorable to Guardian and ignoring the parts of the record that were favorable to Schwalm. Specifically, Schwalm argues that The Sierra Group VA should be disregarded because Guardian “cherry picked” among the evidence it provided to The Sierra Group in an attempt to manipulate the outcome of the report.
Cf. Spangler v. Lockheed Martin Energy Sys., Inc.,
313 F.3d 356, 361-62 (6th Cir.2002) (concluding it is arbitrary and capricious for a insurance provider to rely on a vocational consultant’s report in making a disability determination where the provider “cherry picked” the insured’s file “in hopes of obtaining a favorable report”).
Although Pickering, The Sierra Group counselor who completed the report, never examined Schwalm, Pickering did examine the medical records from Dr. Covington’s office, the results of the FCE, the Cooperation Agreement, and Schwalm’s September 4, 2006 letter to Guardian. Pickering also reviewed a newspaper article concerning Schwalm’s former employer, Acero, noting that the company had fallen behind on servicing its debts as early as mid-2002 and that it had laid off sixty of its one hundred employees in September 2002, while Schwalm was still chief executive officer. Pickering’s review accurately
summarized the record at the time and Guardian was not arbitrary or capricious in its evaluation of the report.
Moreover, three of the four items that Schwalm contends The Sierra Group should have been provided with were either irrelevant or unavailable at the time of VA. There was no reason to review the plan’s definition of disability because the VA was not intended to evaluate whether Schwalm was disabled within the meaning of the plan. Rather, it was intended to evaluate his “status and occupation.” Similarly, the affidavits prepared by Marquart and Schwalm, although perhaps relevant to The Sierra Group’s inquiry, were completed in March 2008, more than one year after the VA was conducted. This is not a situation, as in
Spangler,
where the plan administrator picked one aberrant medical evaluation from the administrative record and provided it to a “neutral” reviewer for evaluation.
See
313 F.3d at 361-62.
Schwalm also contends that Guardian placed too much weight on the Peritus Cooperation Agreement, and not enough on other evidence indicating the Cooperation Agreement did not accurately summarize his activities at Peritus. He contends that Peritus was less successful than its partners had hoped, due in part to Schwalm’s deficient performance, and that Schwalm never received any compensation beyond medical benefits and reimbursement of business expenses. Schwalm asserts that his duties consisted primarily of bookkeeping, and that the nature of the business — recruiting qualified technology industry employees for client companies— meant that Marquart’s business development role shouldered most of the responsibilities.
First, Guardian does not contend that Schwalm’s activities at Peritus constituted “gainful work,” but that those activities are some evidence of Schwalm’s ability to perform gainful work. Second, although Schwalm consistently characterizes his Peritus activities as “vocational rehabilitation” rather than “work,” it was not unreasonable to conclude that a person who devotes nearly twelve hours per day to an office, managing the administration and finance of a company, has the ability to “work.” Many of Schwalm’s hours at Peritus may have been dedicated to rest and personal obligations, but some were devoted to business. However those hours doing business are characterized, Schwalm’s activities at Peritus are some evidence of his ability to perform “gainful work” and Guardian’s consideration of those activities was not arbitrary or capricious.
Schwalm next contends that the district court did not adequately consider the inherent conflict of interest in Guardian’s review of Schwalm’s disability claim. Guardian, after all, is both the payor of any long-term disability benefits and the administrator vested with discretion to determine his eligibility for those benefits. Indeed, such an inherent conflict of interest is “one factor” that must be considered when evaluating a plan administrator’s decision to deny benefits under ERISA.
Metro. Life Ins. Co. v. Glenn,
554 U.S. 105, 128 S.Ct. 2343, 2351, 171 L.Ed.2d 299 (2008);
Firestone Tire & Rubber Co. v. Bruch,
489 U.S. 101, 115, 109 S.Ct. 948, 103 L.Ed.2d 80 (1989). But there are many other factors a reviewing judge must consider as well.
Glenn,
128 S.Ct. at 2351. In its opinion below, the district court acknowledged Guardian’s conflict of interest, but concluded, based on the record as a whole, that its decision to terminate Schwalm’s benefits was nevertheless supported by substantial evidence. The district court’s consideration of the inherent conflict of interest was proper.
The Supreme Court made clear in
Glenn
that such a conflict is a red flag that may
trigger a somewhat more searching review of a plan administrator’s decision, but the arbitrary and capricious standard remains in place.
Glenn,
128 S.Ct. at 2350. Guardian did not refuse to consider Schwalm’s evidence of disability.
See Killian v. Healthsource Provident Adm’rs, Inc.,
152 F.3d 514, 521 (6th Cir.1998). It accurately summarized and considered the evidence Schwalm submitted after the initial denial of benefits, including the All-work VE and Dr. Reddy’s analysis. Then, it determined that Schwalm retained the physical capability to perform a sedentary job, like a chief executive officer, and that there was insufficient medical evidence of mental limitations to prove disability. Guardian provided a thorough review of the record and there is no indication that the review was improperly influenced by the inherent conflict of interest.
B.
After Schwalm’s procedural objections to Guardian’s decision are eliminated — i.e. objections to
how
the decision was made' — Sehwalm’s principal contention is that Guardian simply made the wrong decision. On that score, the administrative record does provide some contradictory evidence. Dr. Covington’s records alone contain numerous statements suggesting physical and cognitive impairment, as well as a substantial number of statements suggesting exactly the opposite. Indeed, as Schwalm emphasizes, Dr. Covington’s records often contain contradictory evidence from the same visit.
Still, the evidence suggesting that Schwalm suffered from a continuing disability is not so one-sided that the decision to deny benefits can be considered arbitrary or capricious.
The FCE, The Sierra Group VA, Dr. Covington’s records and his comments on the FCE, Dr. Harris’ IME, the denial of benefits by the Social Security Administration, and even Schwalm’s own VE from Allwork suggest that Schwalm was physically capable of returning to a sedentary job. There is some evidence that lingering pain and the side effects from the pain medications would prevent the high-level work required of a chief executive officer, particularly the Allwork VE, the affidavits from Schwalm and his partner, and the 2005 IME. That evidence, however, was based on Schwalm’s subjective representations concerning the side effects of his medications and Guardian was entitled to discount it based on the other evidence in the record. For example, Schwalm drafted the Cooperation Agreement, which is a brief but comprehensive agreement between the partners concerning their respective obligations and entitlements with respect to Peritus. Moreover, the agreement provided a rather extensive list of duties for Schwalm, including performing all of the company’s executive and administrative tasks and working at least 160 hours per month. The accommodations provided to Schwalm were significant, but acceptable to the company. Schwalm’s later representations to the contrary were self-serving and contradicted many of
Schwalm’s statements to his medical providers.
Schwalm’s plan required him to provide objective evidence to support a continued finding of disability.
Cf. Cooper v. Life Ins. Co. of N. Am.,
486 F.3d 157, 166 (6th Cir.2007) (noting it is reasonable for a plan administrator to require “a claimant to provide objective medical evidence of disability”). Schwalm did not provide the necessary information. Indeed, until the appeals process began, Schwalm focused on his physical limitations and began emphasizing his cognitive limitations only after his benefits had already been terminated.
In response to Guardian’s concerns about the lack of objective evidence of cognitive limitations, Schwalm provided Guardian with a list of medications and printouts from several Internet sites outlining their potential side effects. The printouts, while certainly some evidence that opioid pain medications can lead to cognitive side effects, see
Smith v. Cont’l Cas. Co.,
450 F.3d 253, 264-65 (6th Cir. 2006), do not demonstrate that Schwalm was disabled by those side effects in this case. Schwalm’s medications were carefully managed by Dr. Covington and Jersan, and often changed in response to Schwalm’s concerns about mental sharpness and memory. On one occasion, Dr. Covington noted that he could “not detect” any evidence of cognitive limitations despite Sehwalm’s concerns. Schwalm repeatedly asked Dr. Covington for an opinion on whether Schwalm was able to return to work, but Dr. Covington declined. Unlike the treating physician in
Smith,
Dr. Covington never suggested that the medications would leave “anyone” unable to function.
See
450 F.3d at 264-65.
IY.
Substantial evidence supports Guardian’s conclusion that Schwalm was no longer disabled within the meaning of the plan. Accordingly, the district court’s decision is AFFIRMED.