MEMORANDUM OPINION AND ORDER

HENRY KENNEDY JR., District Judge

Jane G. Fitts, ("Fitts"), was formerly employed as an attorney for the Federal National Mortgage Association ("Fannie Mae"), whose employee disability insurance provider is defendant Unum Life Insurance Company of America ("Unum"). Fitts brings this action claiming that Unum violated the Employee Retirement Income Security Act of 1974 ("ERISA"), 29 U.S.C. § 1001 et seq., because it determined that she was eligible to receive disability payments for only twenty-four months under a policy that limits disability payments for any Fannie Mae employee who develops a disability that is due to a "mental illness." Before the court are the parties' cross motions for summary judgment. Fitts, who has been diagnosed with bipolar disorder, claims that Unum improperly classified her disability as a mental illness and thus improperly subjected her to the disability policy's twenty-four month limitation of benefits provision. Unum, also seeking summary judgment, argues the opposite. Upon consideration of the parties' motions, the oppositions thereto, and the record in this case, the court concludes that Fitts's motion must be granted, and Unum's motion denied.

I. BACKGROUND

Fitts began to work as an attorney for Fannie Mae in 1982. In 1988, Fitts started seeing her current psychiatrist and was ultimately diagnosed with bipolar disorder. Pl's. Mot. for Summ. J., Ex. 8 at 2. In 1995, Fitts was unable to continue working due to her illness. Compl. ¶ 17. Part of the employee welfare benefit plan that Fannie Mae offers to its workers includes a long-term disability insurance policy that is issued and administered by Unum. Under the policy, any employee who develops a disability is eligible for a certain package of benefits until age sixty-five. The policy contains an exception, however. If the employee's disability is "due to a mental illness," the employee's benefits are discontinued after twenty-four months. The policy defines mental illness as a "mental, nervous or emotional disease or disorder of any type." Def's. Opp'n to Pl's. First Mot. for Summ. J., Ex. 1.

Fitts applied for and began receiving disability benefits in 1995, but she was informed that her benefits would cease in 1997 because her bipolar disorder was subject to the twenty-four month limitation for mental illnesses. Compl. ¶ 20. After Unum rejected Fitts's challenge to its determination that bipolar disorder was a mental illness for purposes of receiving disability benefits, Fitts filed this lawsuit.

II. PROCEDURAL HISTORY

This is the third time Fitts's claim has been considered by this court. When this suit was first filed, Fitts claimed that Unum and Fannie Mae had breached certain common law duties, as well as violated the Americans with Disabilities Act ("ADA"), ERISA, and the District of Columbia Human Rights Act ("DCHRA"). With the exception of Fitts's ERISA claim, this court (Urbina, J.) dismissed all of Fitts's claims in an opinion dated March 29, 1999. See Fitts v. Federal Nat'l Mortgage Ass'n, 44 F. Supp. 2d 317, 331 (D.D.C. 1999) (Urbina, J.).

In a subsequent decision, Judge Urbina granted defendants' motion for summary judgment on Fitts's ERISA claim (Count 3). See Fitts v. Federal Nat'l Mortgage Ass'n, 77 F. Supp. 2d 9, 24 (D.D.C. 1999) (Urbina, J.). Judge Urbina, employing an abuse of discretion standard of review, found that defendants' classification of Fitts's infirmity as a mental illness was reasonable, and therefore did not constitute an abuse of discretion. See id. at 25. On appeal, the D.C. Circuit affirmed the district court's dismissal of Fitts's non-ERISA claims but reversed the grant of summary judgment on her ERISA claim, holding that the district court should have reviewed the classification determination de novo. See Fitts v. Federal Nat'l Mortgage Ass'n, 236 F.3d 1, 6 (D.C. Cir. 2001). Consequently, this case was remanded for a de novo determination of whether defendants properly classified Fitts's bipolar disorder as a mental illness. See id.

Pursuant to the mandate and based on the summary-judgment record created before the circuit court's decision, this court granted summary judgment in Fitts's favor. See Fitts v. Federal Nat'l Mortgage Ass'n, No. 98-00617, slip. op. (Feb. 26, 2002). Thereafter, this court granted Unum's motion for reconsideration in order to provide the parties an opportunity to supplement the record concerning whether bipolar disorder fits within the "mental illness" exclusion in Unum's long term disability agreement. See Fitts, 236 F.3d at 6 ("In light of the change in the standard of review, the parties will be free to supplement the existing record by, inter alia, submitting current medical evidence regarding bipolar disorder."). In addition, to facilitate the efficient resolution of this issue, the court bifurcated discovery and created two phases to the case: Phase I, addressing the applicability of the "mental illness" limitation, and Phase 2 (if necessary), addressing whether or not Fitts remains disabled within the terms of Unum's disability plan.

Following the close of Phase I discovery, the parties have once again moved for summary judgment. Now, with the benefit of a supplemented record, the court must re-examine whether bipolar disorder is included in Unum's long term disability limitation for "mental illness."

III. ANALYSIS

Fitts maintains that the policy's definition of mental illness is ambiguous and that the doctrine of contra proferentem requires ambiguous contract terms to be interpreted against the drafter, which in this case is Unum. She asserts that as long as her construction of the policy's provision regarding mental illness is reasonable, she should prevail. She goes on to argue that it is reasonable to read the policy's mental illness exception to exclude bipolar disorder because of the physical characteristics of the illness. Specifically, she contends that she herself has experienced some of the physical conditions said to correlate with bipolar disorder. In response, Unum asserts that bipolar disorder is quite plainly and unambiguously a mental illness because it is defined in the medical community and by ordinary laypeople as such, and because it has no physical causes. Furthermore, Unum contends that Fitts does not actually have bipolar disorder, nor does she exhibit any of the physical conditions that some bipolar patients experience.

The court agrees with Fitts that the term "mental illness" is ambiguous and therefore must rule in Fitts's favor as to the issue of whether bipolar disorder is encompassed within the "mental illness" limitation. However, to the extent that the parties ask the court to rule on Phase 2 issues concerning Fitts's continuing disability, or what affliction Fitts actually suffers from, the court must deny summary judgment given the existence of material issues of fact.

A. Principles of Construction in ERISA Benefit Plans

Because Fitts's claim involves an employee benefit plan, it is governed by ERISA rather than state contract law. ERISA requires that terms in benefit plans "be written in a manner calculated to be understood by the average plan participant." 29 U.S.C. § 1022(a). Consequently, those terms should be given the meaning normally attributed to them by a person of average intelligence and experience. See Phillips v. Lincoln Nat'l Life Ins. Co., 978 F.2d 302, 308 (7th Cir. 1992). While an expert's definition of a contract term is not controlling, the court can rely on expert opinion as a way of determining a term's ordinary meaning. See, e.g., Kunin v. Benefit Trust Life Ins. Co., 910 F.2d 534, 536 (9th Cir. 1990).

The reach of ERISA is broad and sweeping, as the Act was designed to preempt state laws regarding employee benefit plans. See 29 U.S.C. § 1144(a); Brewer v. Lincoln Nat'l Life Ins. Co., 921 F.2d 150, 153 (8th Cir. 1990). Congress, in passing ERISA, anticipated that "a federal common law of rights and obligations under ERISA-regulated plans would develop." Pilot Life Ins. Co. v. Dedeaux, 481 U.S. 41, 56 (1987). Thus, federal law controls the disposition of this suit.

The doctrine of contra proferentem requires courts to construe ambiguous terms in insurance contracts strictly in favor of the insured and against the insurer. See Germany v. Operating Eng'rs Trust Fund, 789 F. Supp. 1165, 1169-70 (D.D.C. 1992). While contra proferentem is a state law doctrine of contract interpretation, it has been applied in the ERISA context as a matter of federal common law. See id. at 1170. The doctrine is applied to insurance contracts because insurance contracts are typically drafted by the insurance company, because insurance companies tend to be repeat players with greater expertise and experience in insurance matters than plan beneficiaries, and because beneficiaries have no opportunity for arms-length negotiation over the terms of the plan. See, e.g., Phillips, 978 F.2d at 307; Germany, 789 F. Supp. at 1170. Because of the disparity in bargaining power and experience between the parties, insurers must clearly delineate any limitations on coverage in a manner easily understood by a layperson, and they cannot take advantage of ambiguities of their own creation. See Kunin, 910 F.2d at 540; Germany, 789 F. Supp. at 1170.

The doctrine is applied only in situations where the contract language is ambiguous, however, and the court will not create an ambiguity where none exists. See Phillips, 978 F.2d at 308. A contract provision is ambiguous if it is susceptible to two or more reasonable interpretations. See Pipage Tribal Utility Auth. v. FEC., 723 F.2d 950, 955 (D.C. Cir. 1983); see also Carey Canada, Inc. v. Columbia Cas. Co., 940 F.2d 1548, 1556 (D.C. Cir. 1991) (defining ambiguity with regard to terms in insurance contracts). As a threshold matter then, the court must make a determination as to whether the "mental illness" exclusion is ambiguous. Accordingly, the court examines various definitions of "mental illness."

B. Defining the term "mental illness"

1. Fitts's Definition

Fitts alleges that the term "mental illness" should be defined to exclude any ailment that has a physical or biological basis. Pursuant to that definition, she maintains that her sickness, bipolar disorder, is not a mental illness because it has physical, biological and genetic components. In support of this claim, Dr. Frederick T. Goodwin testified that bipolar disorder is a neurobiological disorder that affects the physical and chemical structure of the brain. He explains that it may be characterized by certain physical occurrences, including degenerative changes observed in the brain, Goodwin Dep. 59:8-9, and a progressive loss of hippocampal cells in the brain. Id. at 61:4-5. In addition, he stated that depressive episodes associated with bipolar disorder are generally accompanied by large outpourings of corticosteroids (stress hormones) from the adrenal gland, which are damaging to a number of areas of the brain. Id. at 61:10-16. Dr. Goodwin also maintains that bipolar disorder runs in families and that certain people have a genetic predisposition for developing the disease. Id. at 130:11-13, 131:20-21. Furthermore, he points out that bipolar disorder is not successfully treated with even the most intense psychotherapy. Id. at 115:15-16. Dr. Goodwin ultimately concludes that bipolar disorder is a physical illness because it is a disease afflicting a physical organ of the body, just like diseases affecting the heart, the kidneys, or the liver.

Dr. Suzanne J. Griffin, a practitioner in the fields of psychology and pharmacology who has treated Fitts since 1996, also testified that bipolar disease is a physical illness. She characterizes bipolar disorder as "a biological disorder of brain function with a genetic inheritance pattern." Griffin Dep. 92:19-93:9. Dr. Griffin maintains that there is evidence that people with bipolar have certain genetic dysfunctions and "individuals may inherit an increased vulnerability for developing bipolar disorder." Id. at 93:17-19, 95:5. While Dr. Griffin admits that "the manifestations symptomatic of bipolar disorder are more obviously behavioral and emotional," she maintains that the physical changes in the brain that result in those manifestations make bipolar disorder a physical illness. Griffin Decl. ¶ 3. She also notes that bipolar disorder is linked to changes in blood flow to the brain similar to the changes exhibited in Alzheimer's disease or heart disease. Thus, Dr. Griffin concludes that "[d]epression is as physical as heart disease, the difference being that because of the function of the affected organ — i.e. the brain, depression gives rise to behavioral symptoms, more pronounced that [sic] those that also attend heart disease." Id. at ¶ 6.

2. Unum's Definition

Unum contends that any condition listed as a mental illness in the most recent edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV), which includes bipolar disorder, should be deemed a mental illness for purposes of interpreting disability policies. Unum argues that the categorizations in the DSM-IV "supply the common terminology used in discussing mental health issues" and "informs the definitions of mental illness in legislation and administrative regulations." Def's. Mot. for Summ. J. at 22-23.

Alternatively, Unum asserts that bipolar disorder is a mental illness because there are no physical conditions that must be present for a person to be diagnosed with the disorder. Id. at 25. Unum points out that witnesses on both sides agree that there is no physical test which can be used to diagnose bipolar disorder. Griffin Dep. 96:13-16; Goodwin Dep. 117:21-118:4; Goldberg Dep. 21:6-12; Hegarty Decl. ¶ 15; Mirkin Decl. ¶¶ 12-14; Haines Decl. ¶ 11. It also notes that the witnesses agree that while there are certain physical correlates that tend to accompany bipolar disorder, none are present in all patients with the illness. Goodwin Dep. 165:3-167:11; Griffin Dep. 97:2-12; Hegarty Decl. ¶ 14. Finally, Unum maintains that while genes may increase vulnerability to developing bipolar disorder, nobody has identified a particular gene that either causes or strongly correlates with bipolar disorder. Goodwin Dep. 132:19; Griffin Dep. 93:10-95:5; Goldberg Dep. 22:2-23:3.

In support of its contention that bipolar disorder is a mental illness, Unum also presents the declaration of Dr. Peter T. Mirkin. First, Dr. Mirkin points out that bipolar disorder is recognized in the DSM-IV, which is accepted as "the official nomenclature of mental disorders" by many medical professionals. Mirkin Decl. ¶¶ 4-5. Second, Dr. Mirkin observes that bipolar disorder is typically treated by psychiatrists utilizing psychotherapy and psychotropic drugs. Id. at ¶ 4. Dr. Mirkin agrees with Dr. Griffin and Dr. Goodwin that "biological and neurochemical changes can accompany bipolar disorder," he admits that there is evidence that individuals may inherit a genetic predisposition for the disease, and he concludes that while the cause of bipolar disorder remains unknown, the disorder probably results from a combination of physiological, psychological, and social factors. Id. at ¶¶ 6, 8. He cautions, however, that the mere existence of biological markers does not make bipolar disorder a physical illness because all illnesses recognized in DSM-IV may correlate with biological changes. Id. at ¶ 6. Dr. Mirkin also states that the relationship between blood flow changes and bipolar disorder is unclear and that blood flow changes are not always present in individuals suffering from bipolar disorder. Id. at ¶ 12.

Unum also presents the declaration of Dr. Robert A. Haines, a board certified psychologist employed by Unum, in support of its position. Dr. Haines concurs with Dr. Mirkin that bipolar disorder is a mental illness because it is defined as such in the DSM-IV and because it is treated with psychotropic drugs or psychotherapy. Haines Decl. ¶ 6. Dr. Haines admits that psychotropic drugs alter the chemical balance in the brain, though he insists that there is no consensus that the chemical imbalance in the brain is the cause of bipolar disorder. Id. at ¶ 10. He further states that it is a mental illness because it is "characterized predominantly by a cognitive, emotional or behavioral abnormality," regardless of any physiological connections to the disease. Id. at ¶ 12. Dr. Haines acknowledges that although the psychiatric community has not yet identified the causes of bipolar disorder, there is some evidence favoring a genetic predisposition. Id. at ¶ 9. 3. Definitions used by courts

Courts are split over how to define the term mental illness and whether ailments like bipolar disorder fall within that definition when it is contained in employee benefit plans. Courts that have addressed the issue have taken three different approaches. The first approach is a symptom-based analysis. Under this model, mental illness insurance limitations "apply to any abnormal condition that manifests itself in symptoms that an untutored layperson without benefit of medical advice or diagnosis would call mental disorder or illness." Lynd v. Reliance Standard Life Insurance Co., 94 F.3d 979, 987 (5th Cir. 1996) (Dennis, J. dissenting). The rationale is that laypeople are more likely to recognize the symptoms of an illness than to understand its causes. See, e.g., Lynd, 94 F.3d at 983-84; Brewer, 921 F.2d at 154 ("[L]aypersons are inclined to focus on the symptoms of an illness; illnesses whose primary symptoms are depression, mood swings and unusual behavior are commonly characterized as mental illnesses regardless of their cause."); Equitable Life Assurance Soc'y v. Berry, 212 Cal. App. 3d 832, 840-41 (1989) (holding that "[m]anifestation, not cause, is the yardstick" for defining mental illness). One pillar of this position is the view that a cause-based definition of mental illness would effectively read out a disability plan's mental illness exception because some physical cause can be identified for any illness. See, e.g., Lynd, 94 F.3d at 983 n. 5 ("In identifying the `causes' and `symptoms' of illnesses, it seems that an argument could always be fashioned that the illness itself should be viewed as a `symptom' of some underlying `physical' cause; this is particularly true if one is willing to trace the origins of the illness ad infinitum."). Because bipolar disorder, depression and other similar illnesses manifest themselves primarily through behavioral and emotional changes, courts using a symptom-based definition of mental illness have routinely found such disorders to be mental illnesses. See, e.g., Tolson v. Avondale Indus., Inc., 141 F.3d 604, 610 (5th Cir. 1998) (defining depression as a mental illness); Lynd, 94 F.3d at 983-84 (classifying "major depressive disorder" as a mental illness); Brewer, 921 F.2d at 154 (considering mood disorders to be mental illnesses); Pelletier v. Fleet Fin. Group, 2000 WL 1513711 (D.N.H. Sept. 19, 2000) (classifying major depressive disorder as a mental illness); Attar v. Unum Life Ins. Co., 1997 WL 446439 (N.D. Tex. July 19, 1997) (defining bipolar disorder as a mental illness); Berry, 212 Cal. App. 3d at 840-41 (defining manic depression as a mental illness).

The second method courts have used to interpret the term "mental illness" is the cause-based approach. Courts using a cause-based classification have held that defining mental illness as an illness having no organic cause whatsoever is reasonable. See, e.g., Phillips, 978 F.2d at 310-11; Kunin, 941 F.2d at 536. These courts find a symptom-based definition for mental illness to be unsatisfying principally because such a definition logically would exclude many ailments that lay people would commonly consider to be physical illnesses, such as abnormal behavior caused by a head injury or brain trauma, brain cancer, Alzheimer's disease, or delirium resulting from a fever or staph infection. See, e.g., Phillips, 978 F.2d at 306 n. 2. Courts taking a cause-based approach have typically decided in favor of insureds, either because they found the definition of mental illness in a disability policy to be ambiguous, see, e.g., Phillips, 978 F.2d at 308, or because they found that diseases such as bipolar disorder were physical illnesses. See, e.g, Arkansas Blue Cross Blue Shield, Inc. v. Doe, 733 S.W.2d 429, 432 (Ark.App. 1987) (affirming the district court's determination that bipolar disorder is a physical illness).

A third recognized interpretation of "mental illness" is the treatment-based approach, where the method of treatment is considered in determining whether an illness is mental or physical. Thus, illnesses treated by psychiatrists employing psychotherapy and psychotropic medication have been considered to be mental illnesses. See, e.g., Blake v. Unionmutual Stock Life Ins. Co., 906 F.2d 1525, 1530 (11th Cir. 1990) (noting that the plaintiff's postpartum depression was properly considered a mental illness because "she was treated primarily by psychiatrists receiving well recognized psychiatric treatment, including individual psychotherapy, psychoactive drug therapy, electroconvulsive therapy and participation in group sessions."); see also Simons v. Blue Cross Blue Shield of Greater New York, 536 N.Y.S.2d 431, 434 (N.Y.Sup.Ct. 1989) (finding that coverage for the plaintiff's hospitalization for anorexia nervosa was not excluded by the insurance contract's mental illness limitation because the patient received treatment for the physical problems from anorexia — malnutrition and hypotension — even though anorexia could be considered a psychiatric condition).

4. Contra proferentum

The lack of consensus on the meaning of mental illness and the prevalence of different definitions for the term indicate that it is ambiguous. The court has outlined five different approaches for defining the term mental illness. All of these approaches have a reasonable basis, yet given the same set of facts, they could all reach a different conclusion. There is no dispute over the possible causes and manifestations of bipolar disorder. Both parties' doctors acknowledge that bipolar disorder is characterized by a combination of physical, psychological, and social factors, and they generally agree as to what those factors are. The primary area of contention is whether those factors make bipolar disorder a mental illness or a physical illness. Thus, the dispute is not over the factors affecting the cause, manifestation, or treatment of the disorder, but over how those factors should be interpreted. That the dispute is over definitions, rather than facts, demonstrates that the term mental illness is ambiguous.

Applying the rule of contra proferentum, mental illness must therefore be construed strictly against Unum and in a manner that is reasonable and most favorable to Fitts. Germany, 789 F. Supp. at 1169; Kunin, 910 F.2d at 539. In this case, the definition of mental illness that is most favorable to Fitts is one sanctioned by both the Ninth and Seventh Circuits, which have held that "mental illness refers to a behavioral disturbance with no demonstrable organic or physical basis. . . . It stems from reaction to environmental conditions as distinguished from organic causes." Phillips 978 F.2d at 310 (quoting Kunin, 910 F.2d at 538). This definition is appropriate because it was reasonable for Fitts, who believes her condition has physically manifested itself, to conclude that her illness was not mental and therefore not included in the mental illness provision of Unum's disability policy. Accordingly, Fitts's motion for summary judgment must be granted with respect to the issue of the inapplicability of the "mental illness" exception to bipolar disorder.

C. Characterizing Fitts's Illness

The question that remains then, is whether Fitts has been properly diagnosed as someone suffering from bipolar disorder, and as such, an individual who should be appropriately excluded from the policy's mental illness exception. As the record reflects, the parties vehemently disagree on the answer to this question.

Fitts has presented testimony from six doctors who agree that she has bipolar disorder. Griffin Dep. 87:10; Goodwin Dep. 50:20; Ketter Letter, Pl's. Mot. for Summ. J., Ex. 7; Polk Letter, Pl's. Mot. for Summ. J., Ex. 8; Johns Hopkins Hospital Discharge Diagnosis, Pl's. Mot. for Summ. J., Ex. 9; Georgetown University Hospital Discharge Diagnosis, Pl's. Mot. for Summ. J., Ex. 10. Fitts asserts that her condition is undeniably bipolar disorder, a diagnosis that she argues is supported by physical manifestations of the disease including short term memory loss, Hyde Dep. 41:4-5, and headaches, Id. 105:5-6; two physical conditions said to correlate with bipolar disorder. One of Fitts's physicians stated that she suffers from a "disorder of brain metabolism" that has a "physiological basis." Ketter Letter, Pl's. Mot. for Summ. J., Ex. 7. In addition, Fitts had an MRI on February 2, 1997 that indicated atrophy in her brain. Def's. Mot. for Summ. J., Ex. 10. This is consistent with the results of some studies that found that bipolar disorder is associated with anatomical brain changes. Griffin Dep. 95:10-12.

Unum presents conflicting evidence that Fitts does not have bipolar, but rather she has a much less severe form of personality disorder that is devoid of any physical or organic bases. Diagnosis of Dr. Frank Moscarillo, Def's. Mot. for Summ. J., Ex. 13 (ruling out bipolar disorder); Letter from Dr. Polk, Def's. Mot. for Summ. J., Ex. 15 ("There may well be an Axis II Borderline Personality Disorder"); Def's. Mot. for Summ. J., Ex. 16-18. In support of its argument, Unum asserts that Fitts herself exhibits no physical manifestations whatsoever. Dr. Robert Madsen, who Unum hired to review Fitts's medical record, believes that Fitts is feigning the memory loss she claims to experience. Def's. Mot. for Summ. J., Ex. 3. Dr. Angela Hegarty, a forensic psychologist and neurologist, also concludes that there is evidence that Fitts has not really experienced any memory loss. Hegarty Decl. ¶ 5. In addition, Unum submits that Fitts does not have any brain atrophy because an MRI of her brain on March 17, 1998, showed that the lesions from the previous MRI were no longer seen. Def's. Mot. for Summ. J., Ex. 11. Unum ultimately concludes that Fitts has exhibited none of the biological conditions said to correlate with bipolar disorder.

Because there is a genuine issue of material fact with respect to the precise nature of Fitts's condition, summary judgment is inappropriate. The parties dispute whether Fitts has experienced any of the physical manifestations said to correlate with bipolar disorder, and each has presented evidence that would provide a basis for a reasonable trier of fact to believe its position. Accordingly, the court finds that a evidentiary hearing is required in order to resolve this factual issue.

IV. CONCLUSION

For the aforementioned reasons, it is this 23rd day of February, 2006, hereby

ORDERED that Fitts's "Motion for Summary Judgment on the `Phase I' Issue of the Inapplicability of the `Mental Illness' Exclusion" (Dkt. #127) is GRANTED, and it is further

ORDERED that Unum's "Motion for Summary Judgement on Count Three of Plaintiff's Complaint (ERISA Claim)" (Dkt. #128) is DENIED.