UNPUBLISHED

UNITED STATES COURT OF APPEALS
               FOR THE FOURTH CIRCUIT


TROY HARRIS,                            
                  Plaintiff-Appellee,
                v.
MICHAEL HOLLAND, Trustee; MARTY
HUDSON, Trustee; JOSEPH BRENNAN,                No. 02-2173
Trustee; B. V. HYLER, Trustee;
TRUSTEES OF THE UNITED MINE
WORKERS OF AMERICA 1974 PENSION
PLAN,
             Defendants-Appellants.
                                        
            Appeal from the United States District Court
      for the Western District of Virginia, at Big Stone Gap.
             Glen M. Williams, Senior District Judge.
                          (CA-01-148-2)

                     Argued: October 29, 2003

                     Decided: February 4, 2004

     Before WILLIAMS, MOTZ, and KING, Circuit Judges.



Affirmed by unpublished per curiam opinion. Judge Williams wrote
a dissenting opinion.


                            COUNSEL

ARGUED: Christopher Francis Clarke, Senior Assistant General
Counsel, Office of the General Counsel, UMWA HEALTH AND
2                        HARRIS v. HOLLAND
RETIREMENT FUNDS, Washington, D.C., for Appellants. Dawn
Allison Mullins, LEE & PHIPPS, P.C., Wise, Virginia, for Appellee.
ON BRIEF: Glenda S. Finch, Deputy General Counsel, Office of the
General Counsel, UMWA HEALTH AND RETIREMENT FUNDS,
Washington, D.C., for Appellants. Lewey K. Lee, Paul L. Phipps,
LEE & PHIPPS, P.C., Wise, Virginia, for Appellee.



Unpublished opinions are not binding precedent in this circuit. See
Local Rule 36(c).


                             OPINION

PER CURIAM:

   The Trustees of the United Mine Workers of America 1974 Pen-
sion Plan (the "Trustees") seek review of the district court’s decision
to award summary judgment in this benefits dispute to plaintiff Troy
F. Harris, a disabled coal miner. Harris v. Holland, No. 2:01CV00148
(W.D. Va. Sept. 13, 2002) (the "Opinion"). By its Opinion, the court
concluded that the Trustees had erred in denying Harris’s claim for
disability pension benefits. 

we agree with the
district court and affirm.

                                  I.

                                  A.

   On September 4, 1987, while working as a mine electrician for
Island Creek Coal Company in southwestern Virginia, Harris injured
his lower back while lifting a 250-pound bull gear (the "1987 Mine
Incident"). As a result of this injury, Harris immediately fell to the
ground and was later hospitalized. An orthopedic surgeon, Dr. Jean
M. Eelma, initially diagnosed Harris’s injury as a strain in the lumbar
region of his back. Dr. Eelma discharged Harris three days later, on
September 7, 1987, with instructions that he not return to work with-
out her permission.
                           HARRIS v. HOLLAND                            3
   Harris returned to the hospital soon thereafter, on September 10,
1987, complaining of continuous pain in his lower back and episodes
of numbness in both legs. Dr. Eelma observed that Harris suffered
from "diffuse lumbar tenderness, worse at the L-5/S-1 level."1 As a
result, on September 16, 1987, an MRI was performed on Harris’s
lumbar spine. Based on the MRI, a radiologist diagnosed Harris with
degenerative disc disease at the L-5/S-1 level, with disc bulging
resulting in stenosis of the L-5 neural exit foramina.2 The following
day, September 17, 1987, Dr. Eelma also reviewed the MRI results
and diagnosed Harris with a "fair sized disc herniation at the L-5/S-
1 level."3

   Harris returned to Dr. Eelma for followup appointments on October
8, 1987, and on October 29, 1987, during which he was treated for
disc disease at the L-5/S-1 level. Finally, on November 12, 1987, Dr.
Eelma gave Harris permission to return to his work in the coal mines.
Although Harris worked the next day and on three subsequent days,
he found his back pain to be intolerable. Two weeks later, Dr. Eelma
determined that Harris had suffered a "flare of his degenerative disc
disease," and she authorized him to leave work temporarily.

   Before the end of the year, Harris returned to the hospital for four
treatments of the injury he suffered in the 1987 Mine Incident — on
November 30, and December 7, 21, and 30, 1987. Following a bone
  1
     The symbol L-5 represents one of five lumbar vertebrae which are
located between the thorax and the pelvis. Dorland’s Illustrated Medical
Dictionary 961 (28th ed. 1994). The symbol S-1 represents one of five
sacral vertebrae which form the sacrum, a triangular bone located
between the lumbar vertebrae and the coccyx. 

The S-1
is located just below the L-5.
   2
     Stenosis identifies the "narrowing or stricture of a duct or canal."
Dorland’s Illustrated Medical Dictionary 1576 (28th ed. 1994). Foram-
ina is the plural of foramen, which means "a natural opening, or passage;
a general term for such a passage, especially one into or through a bone."

     The term herniation is used to describe "the abnormal protrusion of
an organ or other body structure through a defect or natural opening in
a covering, membrane, muscle, or bone." Dorland’s Illustrated Medical
Dictionary 759 (28th ed. 1994).
4                         HARRIS v. HOLLAND
scan on December 30, 1987, doctors concluded that Harris suffered
from degenerative osteoarthritis, "particularly evident at the level of
L-5/S-1, with likely bulging of disc on right, and arthrosis of facets."
On January 5, 1988, during yet another examination, Dr. Eelma
observed that Harris had persistent low lumbar tenderness. Although
she did not view Harris to be a surgical candidate, she opined that
Harris could not "continue in his present state." Dr. Eelma therefore
recommended that Harris see her for anti-inflammatory and pain
reliever injections at the L-5/S-1 level, and Harris did so on January
8, 1988. Finally, after nearly fifteen hospital visits and related doctor
examinations and treatments, Harris was authorized to again return to
work, which he did on February 22, 1988.

   In 1995 and 1996, Harris was injured in two other incidents while
working in the coal mines. First, on February 25, 1995, Harris injured
his neck when he hit his head on a piece of mining equipment (the
"1995 Mine Incident"). Harris nevertheless remained on the job and
received no first-aid treatment. Second, on November 8, 1996, while
attempting to lift a 150-pound piece of emonorail, Harris felt a sharp
pain in his lower back (the "1996 Mine Incident"). As a result, Harris
visited the hospital and was treated by Dr. Joshua P. Sutherland, Jr.,
a general practitioner. A back x-ray revealed degenerative changes in
Harris’s lower back, and he received injections for back pain and was
diagnosed with lumbar strain.

   After approximately twenty-four years in the coal mines, Harris
ceased working for Island Creek in February of 1997. On June 19,
1997, while pushing a lawnmower, Harris felt sharp pain in his lower
back, accompanied by pain in his right hip, radiation down his leg and
groin, and stomach pain (the "Lawnmower Incident"). Harris visited
the hospital the next day, where he advised the attending physicians
that he had been experiencing back problems since the 1987 Mine
Incident. A radiologist took x-rays of Harris’s spine and reported
degenerative changes in the lumbar region.

   On June 23, 1997, Harris returned to the hospital for a followup
examination in connection with the Lawnmower Incident. Dr. Suther-
land again examined Harris, and he reported that Harris suffered from
a severely decreased range of motion in his lumbar spine. Following
an MRI, a radiologist observed disc protrusion at the L-5/S-1 level.
                            HARRIS v. HOLLAND                               5
Harris continued to see Dr. Sutherland for monthly followup exami-
nations until May of 1998. As a result, Dr. Sutherland concluded that
Harris was unable to work and that he was totally disabled. On March
16, 2000, in writing to the Trustees, Dr. Sutherland opined that "[t]he
patient [Harris] is unable to do gainful employment as a direct result
of advanced degenerative lumbar disc dysfunction" (the "Dr. Suther-
land Letter").

                                     B.

   On August 8, 1997, Harris applied for Social Security Disability
Insurance ("SSDI") benefits. The appropriate state agency denied
Harris’s application for benefits on October 22, 1997, and that denial
was upheld on January 20, 1998. Harris then requested an administra-
tive hearing and, in September 1998, an Administrative Law Judge
("ALJ") changed the initial decision and approved Harris’s applica-
tion for SSDI benefits. In so doing, the ALJ concluded that Harris had
been permanently disabled since the Lawnmower Incident, and he
found that Harris’s "severe impairments" were "residuals of traumatic
low back injury with radiculopathy in the lower extremities; arthritis
in the neck, arms, and hands; and situational depression."4 In re Har-
ris, Decision of Social Security Administration (Sept. 8, 1998).

   On January 11, 1999, Harris applied to the Trustees for an award
of disability pension benefits pursuant to the 1974 Pension Plan.5 The
  4
     The term traumatic refers generally to "a wound or injury." Dorland’s
Illustrated Medical Dictionary 1735 (28th ed. 1994). Radiculopathy is
used to describe a "disease of the nerve roots." 

     The 1974 Pension Plan governs the 1974 Pension Trust, which is a
multi-employer employee pension plan covered by the Employee Retire-
ment Income Security Act ("ERISA"), 29 U.S.C. §§ 1001-1461. The
1974 Pension Plan provides benefits to plan participants who become
totally disabled as a result of mine accidents. It provides, in relevant part:
      A Participant who (a) has at least 10 years of signatory service
      prior to retirement, and (b) becomes totally disabled as a result
      of a mine accident . . . shall, upon retirement . . . be eligible for
      a pension while so disabled. A Participant shall be considered to
      be totally disabled only if by reason of such accident such Partic-
      ipant is subsequently determined to be eligible for Social Secur-
      ity Disability Insurance Benefits. . . .
1974 Pension Plan, Art. II pt. C.
6                              HARRIS v. HOLLAND
Trustees’ Disability Pension Analyst, a registered nurse (the "Nurse
Analyst"), denied Harris’s claim, relying on interpretive guidelines
promulgated by the Trustees.6 These guidelines are in question and
answer format ("Funds Q&As"), and they provide for the consistent
interpretation and application of the 1974 Pension Plan. The specific
Funds Q&A of concern here, Funds Q&A 252, provides, inter alia,
that a "disability must have been caused by the exertion or impact of
some external physical force or object against the body or by the exer-
tion or impact of the body against some external physical object . . . ."
Funds Q&A 252(3).7 The Nurse Analyst relied on this provision in
deciding that the 1987 Mine Incident was not a "mine accident" under
the 1974 Pension Plan. She based her decision on the premise that
Harris was not in the process of lifting the bull gear when his 1987
injury occurred.
    6
     The 1974 Pension Plan authorizes the Trustees to "promulgate rules
and regulations to implement this Plan . . . ." 1974 Pension Plan, Art.
VIII pt. B.1.
   7
     Pursuant to the guidelines promulgated by the Trustees, a miner is
totally disabled as a result of a mine accident if three conditions are satis-
fied. The miner’s disability must satisfy the criteria of (1) unexpected-
ness, (2) definiteness, and (3) force or impact. More specifically, Funds
Q&A 252 provides as follows:
        Q. The 1950 and 1974 Benefit and Pension Plans provide ben-
        efits to certain persons who are totally disabled as a result of a
        mine accident. For purposes of these provisions, what test must
        be met for a finding that a miner was "disabled as the result of
        a mine accident?"
        A. The following three characteristics must be present: (1)
        Unexpectedness: The disability must have been unlooked for and
        unforseen; (2) Definiteness: The disability must be traceable to
        a definite time, place, and occasion which occurred within the
        course of the mine worker’s employment. A progressive disease
        does not meet this test and therefore cannot be a disability that
        resulted from a mine accident; (3) Force or Impact: The disabil-
        ity must have been caused by the exertion or impact of some
        external physical force or object against the body or by the exer-
        tion or impact of the body against some external physical object;
        i.e., not simply as a result of the mine worker’s own physical
        condition.
Funds Q&A 252.
                          HARRIS v. HOLLAND                            7
   The Nurse Analyst further concluded that, even if the 1987 Mine
Incident constituted a mine accident under the 1974 Pension Plan, the
medical evidence failed to establish that the 1987 Mine Incident had
resulted in Harris’s disability. The Nurse Analyst also rejected Har-
ris’s claim that the 1996 Mine Incident had caused his disability, con-
cluding that the Lawnmower Incident was solely responsible for
Harris’s disability. On February 9, 2000, the Nurse Analyst, acting on
behalf of the Trustees, denied Harris’s claim.

   About a month later, on March 16, 2000, the Trustees received the
Dr. Sutherland Letter regarding Harris’s back problems. Dr. Suther-
land opined that "Mr. Harris’s current presenting complaints and
qualifications for disability is [sic] directly related to the injury of
1987 . . . . There is a direct cause and effect associated with the
patient’s present back pain and the injury of 1987." A week thereafter,
on March 24, 2000, a Program Specialist for the Trustees sent Harris
another denial letter, stating simply:

    We regret to inform you that the letter submitted by Dr.
    Sutherland dated March 16, 2000 does not add any new
    information to your disability case. The letter will be added
    to your file and your disability pension remains denied.

   On May 5, 2000, Harris requested that the Trustees conduct a hear-
ing on his claim for benefits. Accompanying this request, Harris sub-
mitted, for the first time, a copy of the accident report on the 1996
Mine Incident. The Nurse Analyst again recommended the denial of
benefits, asserting that the accident report failed to establish a causal
link between the 1996 Mine Incident and Harris’s disability, and
again asserting that the 1987 Mine Incident was not a mine accident.
Although the Nurse Analyst reviewed Harris’s entire file, she failed
to mention the Dr. Sutherland Letter, and she accordingly failed to
assess its impact on Harris’s claim for disability benefits. On July 27,
2000, Harris was again notified by the Trustees that his claim for ben-
efits was denied.

   In October of 2000, Harris requested that the Trustees reconsider
his claim for disability benefits and, in support thereof, submitted an
accident report on the 1995 Mine Incident. In response, the Nurse
Analyst conducted another review of Harris’s file and, on November
8                         HARRIS v. HOLLAND
30, 2000, she again recommended the denial of Harris’s claim. As
bases therefore, she asserted the following: (1) no treatment records
were on file for the 1995 Mine Incident, and it therefore could not
have caused Harris’s disability; (2) Harris continued to work after the
1996 Mine Incident, and it thus could not have caused his disability;
and (3) the 1987 Mine Incident did not constitute a mine accident
under the 1974 Pension Plan. Although the Nurse Analyst acknowl-
edged the existence of the Dr. Sutherland Letter, she did not make any
findings or draw any conclusions concerning its contents.

   On February 5, 2001, Harris was informed by the Trustees’ Hear-
ing Officer that he would receive a hearing concerning the Nurse
Analyst’s finding that the 1987 Mine Incident did not constitute a
mine accident under the 1974 Pension Plan. During a March 27, 2001,
telephonic hearing on this issue, Harris and Glenn Herbert, a
coworker and an eyewitness to the 1987 Mine Incident, each testified
as to the specific events of that day. Herbert, who was aiding Harris
in lifting the bull gear, testified that, contrary to the finding of the
Nurse Analyst, Harris was in the process of lifting the bull gear when
his back was injured.8 In light of this evidence, the Hearing Officer
overruled the Nurse Analyst and found that Harris was in the process
of lifting the bull gear when he injured his back. Based upon this find-
ing, the Hearing officer reversed the Trustees on the 1987 Mine Inci-
dent, concluding that it constituted a "mine accident" under Funds
Q&A 252.

   In light of the Hearing Officer’s decision, the Trustees reconsidered
Harris’s claim for disability pension benefits. Once again, however,
the Nurse Analyst recommended denying Harris’s claim. On this
occasion, she asserted that, although the 1987 Mine Incident consti-
tuted a mine accident under the 1974 Pension Plan, it did not result
in Harris’s disability. Her conclusion on this point was premised
solely on the fact that Harris was able to perform his regular work
duties, without further treatment on his injured back, from January of
1988 until February of 1998. The Nurse Analyst asserted that this fact
    8
   Harris also submitted affidavits to the Hearing Officer from three
other employees, the Shift Foreman, the Lampman and the Loadout
Operator. All three assisted in transporting Harris out of the mine after
he was injured; however, none witnessed the accident.
                           HARRIS v. HOLLAND                            9
mandated the conclusion that the 1987 mine accident did not result in
Harris’s disability. Again, however, she made no analysis with respect
to the Dr. Sutherland Letter, stating only that "[a]lthough Dr. Suther-
land drew a direct relationship between the 1987 injury and the dis-
ability of June 21, 1997, the records do not support this conclusion."
As a result, the Trustees, on June 12, 2001, again denied Harris’s
claim for disability benefits.

   On December 10, 2001, Harris appealed the Trustees’ decision to
the district court for the Western District of Virginia. On September
13, 2002, that court reversed the Trustees and awarded summary
judgment to Harris, thereby granting disability pension benefits to
him. Harris v. Holland, No. 2:01CV00148 (W.D. Va. Sept. 13, 2002).
The Trustees have filed a timely notice of appeal, and we possess
jurisdiction pursuant to 28 U.S.C. § 1291.

                                   II.

   We review de novo a district court’s award of summary judgment,
employing the same standards applied by the district court. Brogan v.
Holland, 

, 161 (4th Cir. 1997). When benefits plan trust-
ees are vested with discretionary authority concerning eligibility for
benefits, a reviewing court examines their decisions for abuse of dis-
cretion. Firestone Tire & Rubber Co. v. Bruch, 

, 111
(1989). Because the Trustees exercise such discretion, we review their
denial of disability benefits for abuse of discretion, and a decision of
the Trustees is not to be disturbed if it is supported by substantial evi-
dence. See Bernstein v. CapitalCare, Inc., 

, 787-88 (4th
Cir. 1995).

                                   III.

   Pursuant to the 1974 Pension Plan, an applicant for a disability
pension must, in order to secure benefits, satisfy three criteria: (1) he
must be totally disabled, as evidenced by an SSDI award; (2) he must
have been injured in a mine accident; and (3) he must demonstrate
that his disability is "a result of a mine accident." See 1974 Pension
Plan, Art. II pt. C. The parties agree that the first two elements of this
three-part test are satisfied, i.e., (1) Harris is disabled, as evidenced
by his SSDI award; and (2) the 1987 Mine Incident constitutes a mine
10                         HARRIS v. HOLLAND
accident. The sole issue in this appeal concerns the third criterion —
whether Harris’s disability was a result of the mine accident. As
explained below, we agree with the district court that the 1987 Mine
Incident resulted in Harris’s disability, and we also agree that the
Trustees’ decision is not supported by substantial evidence.9

                                    A.

   In their appeal, the Trustees maintain that the 1987 Mine Incident
did not result in Harris’s disability because he thereafter worked in
the mines for ten years without receiving further treatment for his
1987 back injury. In order to properly assess this contention, we must
analyze and apply our precedents in Boyd v. Trustees of the United
Mine Workers Health & Retirement Funds, 

(4th Cir.
1989), and Robertson v. Connors, 

(4th Cir. 1988),
which discuss the proper procedures for determining whether a mine
accident resulted in disability under the 1974 Pension Plan.

   In our Robertson decision, the claimant had strained his neck in
1980 when he was thrown from a coal shuttle car. 

. Robertson subsequently received an SSDI award dating
back to his mine accident, but the Trustees nevertheless denied his
claim for benefits under the 1974 Pension Plan. They reasoned that
Robertson had been hospitalized for anxiety and tension in 1972, for
a neck injury in 1977, and for chest pains in 1979, and that his physi-
cal and psychological ailments thus existed prior to the 1980 mine
accident. 

n.2. On appeal, however, we ruled that the
proper analysis in such a proceeding is not whether the mine accident
was the sole cause of disability, or even the most significant cause
thereof, but whether the mine accident was a proximate cause of dis-
ability. 

Judge Sprouse explained:

      The only reasonable interpretation of the requirement that
      total disability be ‘the result of a mine accident,’ therefore,
      is that it requires total disability to have been proximately
      caused by the mine accident. That is, if the plaintiff was
  9
   Because we agree with the district court that the 1987 Mine Incident
caused Harris’s disability, we need not assess whether the 1995 or 1996
Mine Incidents also resulted in his disability.
                          HARRIS v. HOLLAND                             11
    injured in a mine accident and that injury, whether in com-
    bination with a previous or subsequent condition, is sub-
    stantially responsible for plaintiff’s inability to perform his
    job . . . then his total disability results from a mine accident.

Mullins, 

, 1200 (W.D. Va.
1980), aff’d, 

(4th Cir. 1981)) (emphasis added). Although
Robertson’s mine accident caused a relatively minor neck injury, and
although his neck injury aggravated an earlier non-mine-related neck
injury, we found that his accident was nevertheless substantially
responsible for his disability. And, because a disability need not be
solely caused by a mine accident in order to be "the result of a mine
accident," Robertson was entitled to benefits under the 1974 Pension
Plan. Accordingly, we reversed the district court, which had upheld
the Trustees’ denial of disability benefits.

   We also addressed the causation analysis process in our 1989 Boyd
decision. In Boyd, the claimant had suffered a cervical spine sprain
when timber fell on her shoulder while she worked in the coal mines.

. Boyd was later awarded SSDI benefits based
on her mental and emotional impairments. 

benefits to Boyd under the 1974 Pension Plan, however, on the
ground that her mine accident did not result in disability because her
symptoms had appeared three years earlier when she was injured in
an automobile accident. 

again reversed the denial of
benefits, emphasizing the reasoning of Robertson. See 

we observed, "[w]hatever the contributions of her prior mental and
emotional condition and whatever the actual physical injury directly
suffered in the mine accident, Boyd’s mine accident was ‘substan-
tially responsible’ for her total disability as found for SSDI purposes."

Trustees’ denial of Harris’s claim for benefits contravenes the
mandate of Robertson and Boyd. First, the fact that Harris was able
to return to work following the 1987 mine accident does not necessar-
ily mean that the accident was not substantially responsible for his
disability. While the 1987 Mine Incident may not have been the sole
cause of Harris’s disability, the proper inquiry is whether the accident
was substantially responsible for his disability, not whether it was last
in chronological order. Boyd and Robertson support the proposition
12                        HARRIS v. HOLLAND
that a mine accident can result in the total disability of a miner, even
though it combined with a previous (as in Boyd and Robertson) or
subsequent injury (as in this case).10 

; 

. Although the 1987 Mine Incident may have com-
bined with the Lawnmower Incident and the other two incidents to
render Harris totally disabled, the 1987 accident is nonetheless sub-
stantially responsible for Harris’s disability. As the district court
explained, "clearly that first back injury in 1987 was a continuous
problem until a final exam by Doctors in 1998. The first Doctor who
saw Claimant found that he had, according to the MRI, a disc problem
at L-5/S-1 and the last Doctor who treated claimant found he had a
disc problem at L-5/S-1." Opinion at 15. Importantly, the ALJ con-
cluded that Harris was permanently disabled as a result of a "trau-
matic" (i.e., "a wound or injury") low back injury. It is thus illogical
to conclude that Harris’s 1987 traumatic low back injury, for which
he was treated approximately fifteen times, was not a proximate cause
of his permanent disability.

   It is not insignificant to our analysis that the Trustees disregarded
the opinion of Dr. Sutherland. In his letter, Dr. Sutherland opined that
"Mr. Harris’ current presenting complaints and qualifications for dis-
ability is [sic] directly related to the injury of 1987." Although we
cannot and do not mandate that the Trustees accord special weight to
the views of a claimant’s treating physician, the Trustees are not enti-
tled to "arbitrarily refuse to credit a claimant’s reliable evidence,
including the opinions of a treating physician." Black & Decker Dis-
ability Plan v. Nord, 

, 1972 (2003). And although we
cannot "impose on plan administrators a discrete burden of explana-
  10
    The circumstances underlying the Robertson and Boyd decisions dif-
fer slightly from the circumstances here. In those situations, the Social
Security Administration determined that the applicant’s disability com-
menced on the date of his mine accident. Here, the ALJ concluded that
Harris’s disability commenced on the date of the Lawnmower Incident.
Both Robertson and Boyd make clear, however, that an applicant’s dis-
ability onset date need not be the same as his mine accident. The proper
inquiry is whether the mine accident proximately caused the disability,
even if the accident combined with a subsequent non-mine-related acci-
dent to render the miner totally disabled. See 

; Rob-

.
                           HARRIS v. HOLLAND                            13
tion when they credit reliable evidence that conflicts with a treating
physician’s evaluation," 

to ensure that the Trustees
"‘provide adequate notice in writing to any participant or beneficiary
whose claim for benefits . . . has been denied, setting forth the spe-
cific reasons for such denial, written in a manner calculated to be
understood by the participant.’" 

29 U.S.C.
§ 1133(1)). In this setting, the Trustees simply decided to ignore the
essentials of the Dr. Sutherland Letter, and they failed to provide any
basis for doing so.11 As the district court aptly explained, "Dr. Suther-
land has treated the plaintiff for some years during the time he has
received his social security, and his opinion . . . would be entitled to
great weight in this case. The trustees apparently have given no con-
sideration whatsoever to Dr. Sutherlands [sic] opinion." Opinion at 3.
Because the Trustees ignored Dr. Sutherland’s opinion, they failed to
properly analyze the proximate cause issue.

                                    B.

   Finally, the Trustees maintain that a progressive disease, such as
degenerative disc disease, is specifically excluded from the 1974 Pen-
sion Plan. Pursuant to Funds Q&A 252(2), a miner is totally disabled
as a result of a mine accident only if his disability is definite, meaning
that it is "traceable to a definite time, place, and occasion which
occurred within the course of the mine worker’s employment." Funds
Q&A 252(2) goes on to exclude progressive diseases from this defini-
tion, stating that "[a] progressive disease does not meet this test and
therefore cannot be a disability that resulted from a mine accident."
In light of the progressive disease proviso in Funds Q&A 252(2), the
Trustees contend that Harris’s disability resulted from a degenerative
process not covered by the 1974 Pension Plan. To the contrary, how-
ever, the Trustees failed to properly apply the guidelines it adopted
in Funds Q&A 252(2). As explained below, its progressive disease
proviso does not apply to the circumstances underlying Harris’s
claim.
  11
    We note with interest the paradox that, although the Nurse Analyst
ignored Dr. Sutherland’s conclusion that the 1987 Mine Incident caused
Harris’s disability, she relied heavily on Dr. Sutherland’s views in reach-
ing her conclusions as to Harris’s medical condition.
14                         HARRIS v. HOLLAND
   As we have previously explained, "the ‘progressive diseases’ pro-
viso should be read not as an exception to the [enumerated] . . . exam-
ple[s] in Q&A 252[ ], but rather as an instruction on how to handle
situations that do not fall into any of the enumerated examples of
‘mine accidents’ in Q & A 252." Richards v. United Mine Workers
of Am. Health & Ret. Fund, 

, 137 (4th Cir. 1990). In
Richards, the claimant suffered chest pains while lifting a hundred
pounds of timber at a mine site. He was later hospitalized, and the
doctors advised him that he had experienced a heart attack and suf-
fered from progressive heart disease. 

on the pro-
gressive disease proviso, the Trustees contended that heart disease,
not mine work, caused Richards to suffer his heart attack, and they
denied his claim for benefits. 

district court reversed,
and we agreed with the district court. We explained that, under the
example found in Funds Q&A 252(k), a mine accident occurs when
"‘a miner suffers a heart attack while pushing a heavy object in the
normal course of his job.’" 

Funds Q&A 252(k)).
Because Richards was lifting a heavy object at the mine site when his
heart attack occurred, he fell within the example enumerated in Funds
Q&A 252(k), and the progressive disease proviso did not apply. 

explained, "[i]n essence, Q&A 252(k) establishes that
heart attacks that occur during heavy lifting in the mines are caused
in substantial part by the miner’s work. If the conditions of Q & A
252(k) are met, the Trustees are foreclosed from further inquiry into
the actual cause of the heart attack." 

Richards applies with equal force here. Under the
example set forth by the Trustees in Funds Q&A 252(j), a coal miner
is disabled as a result of a mine accident if he "injures his back lifting
a heavy object in the normal course of his job." This is precisely what
happened to Harris in the 1987 Mine Incident. The Hearing Officer,
in her decision, specifically found that Harris was lifting the 250-
pound bull gear when he injured his back in 1987. Harris’s condition
therefore falls under the example enumerated in Funds Q&A 252(j),
because he was injured while "lifting a heavy object in the normal
course of his job" in the coal mines. Accordingly, the progressive dis-
ease proviso does not apply to Harris’s claim, and his disability satis-
fies the requirement of definiteness in Funds Q&A 252(2). Therefore,
the Trustees’ contention on this point must also be rejected.
                          HARRIS v. HOLLAND                           15
                                  IV.

  Pursuant to the foregoing, we affirm the Opinion of the

                                                            AFFIRMED

WILLIAMS, Circuit Judge, dissenting:

   In this case, our sole duty is to determine whether the Trustees’
decision to deny Harris benefits was an abuse of discretion. Whether
the Trustees abused their discretion turns, in this case, on whether
their determination that Harris’s disability was not caused by a mine
injury was supported by substantial evidence. Because I conclude that
the Trustees’ decision was supported by substantial evidence, and
because I believe that my colleagues in the majority apply the now-
discredited treating physician rule, I respectfully dissent.

                                   I.

   I essentially agree with the majority’s recitation of the facts, so I
will recount only those facts that I believe that my colleagues omit or
de-emphasize. Harris unquestionably injured his back in a mine acci-
dent in September of 1987. As the majority notes, he received various
treatments for the back injury during the four months following the
accident, and in the course of those treatments, he was diagnosed with
degenerative disc disease, a form of osteoarthritis.1 By November,
Harris was "quite a bit better" and "anxious to return to work." (J.A.
at 220.) He suffered a relapse, however, which his physician attri-
buted to "a flare of his degenerative disc disease." (J.A. at 221.) He
then underwent a bone scan, because his physician was concerned
about the possibility of "arthritic change." (J.A. at 223.) The bone
scan confirmed the earlier diagnosis of osteoarthritis. After January of
1988, Harris received no additional medical treatment for his back for
nearly nine years. During this time, he continued to be actively
employed as a miner.
  1
    Osteoarthritis is "a noninflammatory degenerative joint disease . . .
characterized by degeneration of the articular cartilage, hypertrophy of
bone at the margins, and changes in the synovial membrane." Dorland’s
Illustrated Medical Dictionary 1286 (29th ed. 2000).
16                         HARRIS v. HOLLAND
   During the 1990’s, Harris was involved in two other minor mine
accidents. In February of 1995, Harris ran into a brow and "stove [his]
neck." (J.A. at 43.) He required no medical treatment after this acci-
dent and remained on the job until the end of his shift. In November
of 1996, Harris injured his back while attempting to lift the end of a
monorail. The injury occurred at 1:50 A.M., and Harris continued
working until the end of his shift. He did not seek medical attention
until 11:30 A.M., over nine hours after the accident, at which time he
was diagnosed with a lumbar strain.

   Despite these accidents, Harris continued to work as a miner until
he was laid off in February of 1997. He sought no medical treatment
for his back between the time he was laid off and June 16, 1997, when
he severely injured his back on June 16, 1997, while mowing his
lawn. After the lawn-mowing accident, Harris’s range of motion in
his legs and back was greatly decreased, and he was in substantial
pain. He was subsequently treated for muscle spasms and acute sciatica.2
He was once again diagnosed with degenerative osteoarthritis in
August 1997. His treating physician, Dr. Sutherland, opined that Har-
ris was "totally disable[d] . . . due to lumbar disc syndrome from 6-
19-97." (J.A. at 261.) In September, Sutherland diagnosed Harris with
"advanced degenerative joint disease," "lumbar disc syndrome," "bi-
lateral sciatica," and "chronic pain disorder." (J.A. at 262.)

   Harris applied for an award of Social Security Disability Insurance
Benefits (SSDI) in August 1997. An administrative law judge (ALJ)
determined that Harris was totally disabled and thus qualified for
SSDI. The ALJ attributed Harris’s disability to "residuals of traumatic
low back injury with radiculopathy3 in the lower extremities; arthritis
in the neck, arms, and hands; and situational depression." (J.A. at
147.) Notably, the ALJ determined that Harris had been disabled
since June 21, 1997, a few days after Harris’s lawn-mowing accident.
  2
     Sciatica is "a syndrome characterized by pain radiating from the back
into the buttock and into the lower extremity." Dorland’s Illustrated
Medical Dictionary 1609 (29th ed. 2000). It is "most commonly caused
by protrusion of a low lumbar intervertebral disk." 

"a disease of the nerve roots." Dorland’s Illustrated
Medical Dictionary 1511 (29th ed. 2000).
                           HARRIS v. HOLLAND                            17
   Harris then applied to the Trustees for a disability pension. The
Trustees denied Harris’s application. They concluded that to the
extent Harris’s disability was attributable to "traumatic low back
injury," "the disability . . . [wa]s related to the non-work-related [i.e.,
lawn-mowing] injury of June 1997." (J.A. at 32.) They concluded that
"the records d[id] not support [Dr. Sutherland’s] conclusion" that
there was a "direct relationship between the 1987 injury and the dis-
ability of June 21, 1997." (J.A. at 32.) Specifically, the Trustees deter-
mined that, given Harris’s "work history and lack of evidence of
[ongoing] back problem[s]," it "would not be reasonable to infer" that
Harris’s disability was related to the "back injuries of 1987 and 1996"
as opposed to the back injury of 1997. (J.A. at 32.)

                                    II.

                                    A.

   When an ERISA disability pension plan commits eligibility deter-
minations to the discretion of the plan administrator or fiduciary, we
review those determinations for abuse of discretion. Firestone Tire &
Rubber Co. v. Bruch, 

, 115 (1989). "It is well-established
that the abuse of discretion standard under [Firestone] is applicable
to our review of the Trustees’ decisions under the UMWA pension
plans." Hale v. Tr. of United Mine Workers of Am. Health & Retire-
ment Funds, 

, 901(4th Cir. 1994). Under this standard,
"we will not disturb such a decision if it is reasonable." Booth v. Wal-
Mart Stores, Inc. Associates Health and Welfare Plan, 

,
342 (4th Cir. 2000).

   Although we may consider many factors in determining the reason-
ableness of a fiduciary’s discretionary decision, see 

; Brogan v. Holland, 

, 161 (4th Cir. 1997);
Lockhart v. United Mine Workers of Am. 1974 Pension Trust, 

, 77 (4th Cir. 1993), the only factor at issue here is the degree to
which the considered materials support the Trustees’ decision. See

. In other words, given the facts of this case,
the Trustees’ decision is reasonable if it is supported by substantial
evidence. See Bernstein v. CapitalCare, Inc., 

, 788 (4th
Cir. 1995).
18                         HARRIS v. HOLLAND
                                    B.

   In order to be eligible for a disability pension under the 1974
UMWA pension plan, a claimant must establish three things. See
Boyd v. Tr. of United Mine Workers Health & Retirement Funds, 

, 59 (4th Cir. 1989). First, the claimant must establish that he
was involved in a mine accident. 

show that he
has been awarded SSDI disability benefits, thus conclusively estab-
lishing that he has a disability ("Qualifying Disability"). 

he must prove that the mine accident proximately caused, or
was substantially responsible for, the Qualifying Disability. 

v. Connors, 

, 475 (4th Cir. 1988).

   As my colleagues note, the Trustees use a set of questions and
answers, called "Q & As," to help them interpret the terms of the 1974
pension plan. See 

. We "afford the Trustees’
interpretation of these rules the same deference that we give the
Trustees’ interpretation of the language of the Plan itself." 

Q & A 252, "miners who become disabled by progressive dis-
eases or conditions such as black lung, silicosis, tuberculosis, arthritis,
rheumatism, etc., cannot be considered ‘disabled as the result of a
mine accident.’" Q & A 252; (Supp. J.A. at 2.). When a progressive
disease combines with a mine accident to proximately cause a claim-
ant’s Qualifying Disability, however, the claimant is eligible for a dis-
ability pension. See 

(explaining that Q
& A 252(k) requires benefits when a mine worker "suffers a heart
attack while pushing a heavy object in the normal course of his job");
Chicarelli v. UMWA Health & Ret. Funds, 

, 462 (4th
Cir. 1991) (same).

  In determining whether the Trustees’ decision is reasonable, we
accord great weight to the disability date specified in an SSDI award
when ascertaining the date and cause of a Qualifying Disability. Rich-

; Horn v. Mullins, 

, 37 (4th Cir.
1981). In fact, we have described the "date of disability determined
by the ALJ" as the "critical factor" to be considered when deciding
whether substantial evidence supports the Trustees’ decision. 

. This "critical factor" clearly supports the Trustees’
decision in this case. The ALJ found the date of disability to be June
21, 1997, within four days of the lawn-mowing accident, but almost
                         HARRIS v. HOLLAND                          19
ten years after the mine accident. Moreover, in addition to bring con-
sistent with the ALJ’s determination, the Trustees’ conclusion is sup-
ported by the following undisputed facts:

    • Harris continued to work following each of his three
      mine accidents until he was laid off in 1997;

    • Harris had been diagnosed with a progressive disease,
      osteoarthritis of the spine, several times since 1987;

    • Harris continued working in the mines for nearly ten
      years after his first, and most severe, mine accident;

    • After treatment for the 1987 mine accident ended in Jan-
      uary 1988, and prior to his lawn-mowing accident, Harris
      sought medical treatment for back pain on only one day,
      the day of the 1996 mine accident;

    • Dr. Sutherland, Harris’s treating physician, opined that
      Harris was "totally disable[d] . . . from 6-19-97," three
      days after the lawn-mowing accident (J.A. at 260-61);

    • The ALJ attributed Harris’s disability to a combination
      of arthritis, lower back trauma, and depression;

    • Dr. Sutherland, the only doctor to opine that the 1987
      mine accident caused Harris’s disability, did not first
      evaluate Harris until nearly ten years after the accident;

    • Harris has required extensive, continuing medical treat-
      ment since the lawn-mowing accident, but he did not
      require such treatment before the accident; and

    • Harris has been unable to work since the lawn-mowing
      accident, but he was able to work before the accident.

In the face of this evidence, substantial evidence supported the Trust-
ees’ decision, and it was reasonable for the Trustees to conclude that
Harris’s disability was caused by a combination of his osteoarthritis
20                         HARRIS v. HOLLAND
and the lawn-mowing accident, not a mine accident.4 Instead of not-
ing the abundance of evidence supporting the Trustees’ determina-
tion, the majority engages in a de novo fact-finding expedition and
determines that the 1987 mine accident substantially caused Harris’s
disability.5 Even if we were reviewing the Trustees’ decision de novo,
  4
     Because my colleagues focus exclusively on the 1987 mine accident,
I focus on that accident as well. I wish to note however, that there is
absolutely no evidence that the later two mine accidents, neither of which
were severe enough to require Harris to leave work, contributed in any
way to his disability. Accordingly, substantial evidence supports the
Trustees’ decision that these minor accidents did not substantially cause
Harris’s disability.
   5
     The majority relies heavily on our opinions in Robertson and Boyd,
holding that the Trustees’ decision contravenes the "mandate" of those
cases. See Robertson v. Connors, 

(4th Cir. 1988); Boyd v.
Trs. of United Mine Workers Health & Ret. Funds, 

(4th Cir.
1989). Although the majority accurately states the facts and holdings of
those cases, they fail to explain how those cases are relevant here. The
parties did not dispute, nor do I, that a mine accident can substantially
cause a Qualifying Disability even if it is not the sole cause of that dis-
ability. This rather unremarkable proposition of law, however, does not
help us determine whether the Trustees’ decision in this case was sup-
ported by substantial evidence. Moreover, if my colleagues mean to
imply that Boyd and Robertson are factually similar to this case, they
simply are mistaken. The majority relegates to a footnote what we have
called the "critical factor" in lower back injury cases — in both of those
cases, the ALJ determined that the miners had become disabled on or
near the date of their mine accidents. See 

n.2
("Our decision is bolstered by the ALJ’s determination that Social Secur-
ity disability for Robertson’s depressive syndrome commenced on the
date of the mine accident."); 

(noting that the ALJ
determined that Boyd was disabled since 1983 following her 1982 mine
accident). In this case, that same "critical factor" supports the Trustees’
decision — the ALJ determined that Harris became disabled within days
of his lawn-mowing accident. Moreover, unlike Harris, both Robertson
and Boyd persistently complained of debilitating pain from the date of
their mine injuries, and neither was able to return to work for any sub-
stantial period of time after his or her mine accident. See 

("Importantly, however, prior to the mine accident, [the mine
worker’s] manifestations of pain did not prevent his return to work.");

("[F]rom the date of her mine injury in July 1982,
Boyd continuously complained of debilitating pain, repeatedly sought
medical assistance for this pain, and finally was unable to continue
work.").
                           HARRIS v. HOLLAND                             21
however, the evidence relied upon by the majority would provide slim
support for its conclusion.

   First, the majority quotes with approval the district court’s finding
that "clearly the first back injury in 1987 was a continuous problem
until a final exam by Doctors in 1998." Ante at 12. To the contrary,
I believe that it is anything but "clear" that Harris had a continuous
problem with his back following the 1987 injury. Harris sought abso-
lutely no medical treatment for his back during the nine years after
the accident, and he continued his employment in the mines during
the same period. In my mind, this hardly evidences a "continuous
problem."

   Next, the majority again quotes with approval the district court’s
finding that "the first Doctor who saw [Harris] found that he had . . .
a disc problem at L-5/S-1 and the last Doctor who treated [Harris]
found he had a disc problem at L-5/S-1." Ante at 12. While this state-
ment is true, the majority fails to note that the "disc problem" found
by the doctors was degenerative osteoarthritis, a progressive disease
that can never be caused by a mine accident. I fail to understand how
the fact that Harris had osteoarthritis at L-5/S-1 in 1987 and continued
to have it in 1998 is at all relevant. Because osteoarthritis is a progres-
sive disease, one would expect Harris’s osteoarthritis to have wors-
ened during the time period between 1987 and 1998, which it in fact
did. A doctor’s finding that Harris had osteoarthritis at L-5/S-1 in
1998 is thus quite unextraordinary and does not increase the likeli-
hood that Harris’s disability was caused by the 1987 mine accident.
Furthermore, because degenerative osteoarthritis is a progressive dis-
ease that cannot be caused by a mine accident, it is unclear how these
findings support the majority’s analysis.

   My colleagues also believe that "[i]t is . . . illogical to conclude that
Harris’s 1987 traumatic low back injury . . . was not a proximate
cause of his permanent disability," because "the ALJ concluded that
Harris was permanently disabled as a result of a ‘traumatic’ . . . low
back injury." Ante at 12. The majority fails to explain why it is logical
to assume that the trauma to which the ALJ referred was a ten-year-
old mine accident, for which Harris had not sought medical treatment
in almost a decade, rather than the lawn-mowing injury that occurred
just days before Harris became totally disabled. Indeed, I believe that
22                           HARRIS v. HOLLAND
the only reasonable conclusion to draw from the date of disability
specified in the SSDI award, is that the "trauma" to which it refers is
the lawn-mowing accident.

   The final arrow in my colleagues’ quiver is Dr. Sutherland’s letter,
in which he opined that "Mr. Harris’s . . . disability is directly related
to the injury of 1987." (J.A. at 287.) The majority avers that the Trust-
ees "ignore[d] the essentials of . . . Dr. Sutherland[’s] Letter." Ante
at 12. This statement is belied by the record. The analyst specifically
considered and rejected Dr. Sutherland’s opinion.6 Moreover, not-
withstanding the fact that Dr. Sutherland’s letter supports the majori-
ty’s factual conclusion, it is not entitled to nearly the evidentiary
weight that my colleagues give it. Black & Decker Disability Plan v.
Nord, 

, 1972 (2003). ("[C]ourts have no warrant to
require administrators automatically to accord special weight to the
opinions of a claimant’s physician. . . .")

     The Trustees obviously, and correctly, questioned how Dr. Suther-
  6
   The paragraph in which the Trustees rejected Dr. Sutherland’s opin-
ion reads as follows:
         The records show Mr. Harris sustained a work-related back
      injury in September 1987. He had treatment over a four-month
      period with no evidence of an ongoing problem. There are no
      treatment records after January 8, 1988. . . . Mr. Harris worked
      until he was laid off on February 13, 1997 with no evidence of
      an ongoing back problem. The work history and lack of evidence
      of a back problem would indicate that he was not disabled.
      Therefore, it would not be reasonable to infer that the back inju-
      ries of 1987 and 1996 are related to the disability basis of residu-
      als of traumatic low back injury with radiculopathy in the lower
      extremities. Although Dr. Sutherland drew a direct relationship
      between the 1987 injury and the disability of June 21, 1997, the
      records do not support this conclusion. . . . [Harris] sustained a
      back injury in June 1997 when he was mowing the lawn. The
      records show he had treatment thereafter. The records support a
      conclusion that the disability basis of residuals of traumatic low
      back injury with radiculopathy in the lower extremities is related
      to the non-work-related injury of June 1997.
(J.A. at 32.)
                          HARRIS v. HOLLAND                           23
land was able, in 2000, to conclude that the 1987 mine accident
directly caused Harris’s disability in light of the intervening degenera-
tive changes caused by Harris’s osteoarthritis and in light of Harris’s
much more recent back injury. The Trustees’ apparent skepticism was
undoubtedly reinforced by the fact that Dr. Sutherland never exam-
ined Harris in connection with the 1987 mine accident or, indeed, at
any time during the nine years after the accident. Moreover, Dr. Suth-
erland’s letter gave no explanation or reasoning as to how he was able
to reach his unlikely conclusion. Nevertheless, resolving this factual
dispute is not our task. Given the evidence in their possession, the
Trustees were clearly entitled to disagree with Dr. Sutherland’s opin-
ion, 

while the analyst’s written findings might not
be a model of explanatory writing, we may not "impose on plan
administrators a discrete burden of explanation when they credit reli-
able evidence that conflicts with a treating physician’s evaluation." 

both the district court and my colleagues chose to
apply the "treating physician rule," which the Supreme Court explic-
itly rejected just last term. 

("ERISA does not support
judicial imposition of a treating physician rule, whether labeled ‘pro-
cedural’ or ‘substantive.’") The treating physician rule, as it had been
developed, embodied both "a ‘procedural’ rule, which require[d] a
hearing officer to explain why she rejected the opinions of a treating
physician, and a ‘substantive’ rule, which require[d] that ‘more
weight’ be given to the medical opinions of a treating physician." 

opinion explicitly disclaims reliance on the
treating physician rule, its holding speaks louder than its disclaimer.
See ante at 13 ("Because the Trustees ignored Dr. Sutherland’s opin-
ion, they failed to properly analyze the proximate cause issue.").

                                  III.

   Because the Trustees’ decision that Harris’s disability was not
caused by a mine accident was supported by substantial evidence, I
would reverse the holding of the district court. In my view, my col-
leagues rely on the now-discredited treating physician rule and ignore
the abundant evidence supporting the Trustees’ decision. Accord-
ingly, I respectfully dissent.