concurring in part and dissenting in part.

I concur in the majority’s affirmance of the denial of Merklin’s motion for summary judgment. I respectfully dissent, however, from the majority’s affirmance of the grant of Liberty’s motion for summary judgment in this case. Regardless whether the district court should have reviewed the case under a de novo or an abuse of discretion standard, summary judgment for the defendants was improperly granted.

I. Coverage

Merklin met her burden to prove she was disabled as defined under the policy, thus invoking coverage. Sabatino v. Liberty Life Assur. Co. of Boston, 286 F.Supp.2d 1222, 1223 (N.D.Cal. 2003). The policy defines the grant of coverage as whether the insured is disabled from performing stated functions, rather than as requiring specific medical findings as to the cause of that disability. For the first 24 months, Merklin is “disabled” if she cannot function in her occupation because of any “Injury or Sickness.”

Thereafter, for Merklin to continue to be disabled, she must be unable to function in any occupation for which she is or becomes reasonably fitted by training, education, experience, and her personal attributes.

Merklin submitted reports from several treating and examining physicians establishing that she was not capable of performing all of the material and substantial duties of her own or any other occupation for which she is or could become reasonably fitted by training, education, experience, age and physical or mental capacity.

The burden then shifted to Liberty to prove an exclusion under the policy. Id. Liberty failed to prove that Merklin’s disability was caused solely by mental illness as defined in the policy. Importantly, the policy provides coverage beyond 24 months for a disability caused by both a physical impairment and a mental illness. Mongeluzo v. Baxter Travenol Long Term Disability Benefit Plan, 46 F.3d 938, 943 (9th Cir. 1995). Although Dr. Leet, Merklin’s psychiatrist, stated that Merklin suffered from severe and recurrent major depression and generalized anxiety disorder, he did not say her mental disorders *33caused her inability to work. Further, Dr. Monka (Merklin’s treating physician in New Jersey) wrote to Liberty to clarify that it was Merklin’s physical disabilities that prevented her from working, and it was the fact that she could not work which in turn caused Merklin to become depressed. There is no evidence in the record to dispute this. Depression caused by physical ailments does not constitute a mental disorder such that the 24-month limitation would apply. Lang v. Long-Term Disability Plan of Sponsor Applied Remote Tech., Inc., 125 F.3d 794, 799 (9th Cir. 1997); Ellis v. Egghead Software Short-Term and Long-Term Disability Plans, 64 F.Supp.2d 986 (E.D.Wash. 1999).

There is no evidence in the record— certainly none that establishes this causation so as to remove any triable issue of fact — that, after 24 months of benefits, Merklin’s disability was caused exclusively by a mental illness as defined in the policy, rather than her physical impairments.

Thus, Merklin’s evidence raises genuine issues of material fact as to whether Merklin is cover under the policy.

II. The District Court’s Standard of Review

Additionally, the district court should have reviewed this case de novo instead of under an abuse of discretion standard because Merklin submitted “material, probative evidence, beyond the mere fact of the apparent conflict, tending to show that [Liberty’s] self-interest caused a breach of the administrator’s fiduciary obligations” and “may have influenced the decision.” Alford v. DCH Group Long Term Disability Plan, 311 F.3d 955, 957 (9th Cir. 2002). Liberty failed to rebut this evidence.

Merklin provided compelling evidence that Liberty was influenced by an actual conflict of interest and acted as an adversary against Merklin by offering the following evidence:

First, Liberty claims on appeal it denied the claim because Merklin did not produce what Liberty deemed to be “objective medical evidence” even though (1) there is no requirement in the policy that Merklin provide “objective medical evidence” other than physicians’ reports to prove her disability, and Liberty may not read such a requirement into the policy, see Sabatino v. Liberty Life Assur. Co. of Boston, 286 F.Supp.2d 1222 (N.D.Cal. 2003); (2) Liberty failed to give Merklin an opportunity to present such proof as it was required to do under the policy; neither did it tell her what “objective medical evidence” it required; (3) the failure to provide what Liberty deemed “objective medical evidence” was not the basis of Liberty’s denial — this is simply an argument Liberty invented for the first time after Merklin filed suit and; (4) Merklin’s treating physicians sent reports to Liberty diagnosing physical conditions such as asthma, diabetes mellitus, hypertension, medical meniscus tears, degenerative arthritis, and allergies, all of which generally are observable by a clinician in an examination and do not require the clinician to rely solely on the descriptions by the patient. In other words, such diagnoses are based on objective medical evidence. These same doctors offered to provide further information if requested, but Liberty failed to request further evidence. Merklin also offered to be examined by Liberty’s doctors, but again, Liberty failed to request-such an examination.

Second, Liberty relied upon an improper definition of disability in processing Merklin’s claim by trying to make Merklin prove her physical disability was the primary cause of her disability, when that is not how the policy defines disability. The majority makes the same mistake.

Third, Liberty acted as an adversary determined to deny the claim. This is par*34tieularly evidence Liberty’s cherry-picking among the various doctor’s reports to cite only the evidence favoring a denial.

Each of these actions is enough to prove that the apparent conflict of interest may have influenced Liberty’s decision to deny Merklin’s claim and, as such, to require that the district court conduct a de novo review. Friedrich v. Intel Corp., 181 F.3d 1105, 1110 (9th Cir. 1999); Siebert v. Standard Ins. Co. Group Long-Term, Disability Policy, 220 F.Supp.2d 1128, 1135 (C.D.Cal. 2002).

III. This Court’s Standard of Review

Finally, the majority opinion fails to analyze this case under the proper standard of review for a summary judgment. We must determine whether, viewing the evidence in the light most favorable to Merklin, there are any genuine issues of material fact and whether the district court correctly applied the relevant substantive law. Fed. R. Civ. P. 56; Celotex Corp. v. Cattrett, 477 U.S. 317, 322, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986). Throughout its analysis, the majority improperly weighs the conflicting evidence in this case, and construes all relevant evidence in favor of the movant, Liberty, instead of the nonmovant, Merklin.

IV. Conclusion

When viewed in the light most favorable to Merklin, the evidence presented is sufficient to raise genuine issues of material fact as to whether Merklin was disabled within the meaning of the policy’s grant of coverage. Accordingly, I dissent from that portion of the majority’s memorandum disposition affirming the district court’s order granting defendants’ summary judgment motion.

*31MEMORANDUM *

Mary A. Merklin and Liberty Life Assurance Company of Boston filed cross-motions for summary judgment in her action for disability benefits, and she appeals the judgment in Liberty’s favor. We affirm.

I

We agree with the district court that review is for abuse of discretion because Merklin did not show material, probative evidence beyond the fact of mere apparent conflict of interest tending to show that self-interest caused a breach of Liberty’s fiduciary obligations. Friedrich v. Intel Corp., 181 F.3d 1105, 1109 (9th Cir. 1999); Atwood v. Newmont Gold Co., 45 F.3d 1317, 1322 (9th Cir. 1995). While Liberty was uncertain about Merklin’s diagnosis for a brief period of time, it continuously relied on the policy’s definition of disability and indicated to Merklin that she could only receive benefits for mental illness for 24 months. Once Liberty concluded that her primary disabling condition was psychological and that there was no objective medical evidence that she was unable physically to perform sedentary work as the policy required, it so informed Merklin.. This was months before her benefits were terminated. It was not “manipulative” for Liberty to have Merklin’s file reviewed independently by the same kind of specialists whom she was consulting in order to determine whether there was objective medical evidence of inability to do sedentary work. Finally, no basis appears in Liberty’s final denial letter of June 14, 2000, to find deception or inconsistency. Dr. Massie reported that Merklin was unable to work from a psychiatric standpoint; the omitted reference to episodes of syncope states that they support her inability to be able to work but does not suggest that the episodes in themselves make Merklin unable to work. Dr. Forbes’s report, which Liberty furnished to Merklin when she requested it, reflects Dr. Ken-nett’s view of Merklin’s ability to do light work limited to four hours per day; Dr. Kennett’s view was already known to Merklin because Dr. Kennett was her treating physician.

II

Liberty had discretion to find that Merklin did not meet the policy’s definition of “disabled” given the reports of Drs. Webber, Isenberg, Gaziano, Nettrour, and Massie. None found a basis for restricting Merklin from sedentary work based on physical medical conditions. There were conflicting indications, for example, Dr. Kennett believed that Merklin was unable to work and Dr. Monka believed that she shouldn’t work. However, Merklin worked through her physical ailments until she became disabled on account of psychological issues. Drs. Isenberg, Gaziano, and Webber indicated that she was capable of performing sedentary work. As there was substantial evidence supporting Liberty’s determination, we cannot say that its determination was based on clearly erroneous findings of fact. See Atwood, 45 F.3d at 1323-24.

III

Whether or not the district court relied too heavily on Dr. Forbes’s report, Liberty’s determination that Merklin could perform sedentary work full time is supported by the reports of Drs. Webber, Sanders, and Nettrour, as well as the absence of any physical restriction identified by others.

*32rv

Liberty sought documentation from Merklin’s treating physicians, considered what they furnished, and evaluated whether their opinions were supported by objective medical data. It was not obliged to accord special weight to their opinions and was entitled to credit reliable evidence adduced by other physicians. Black & Decker Disability Plan v. Nord, 538 U.S. 822, 834, 123 S.Ct. 1965, 155 L.Ed.2d 1034 (2003).

V

To the extent that Merklin argues that Liberty based its decision on the unsupported theory that her physical problems had not prevented her from working until the onset of depression, we disagree that support is lacking. Dr. Monka’s original correspondence so indicated. Nor do we agree that Liberty failed to consider all of Merklin’s ailments together. It reviewed her claims on at least five occasions, and explained why, based on medical findings, Merklin was unable to work from a psychiatric standpoint but that her physical limitations were not severe enough to be disabling.1

AFFIRMED.