RECOMMENDED FOR FULL-TEXT PUBLICATION
                                 Pursuant to Sixth Circuit Rule 206
                                        File Name: 08a0353p.06

                     UNITED STATES COURT OF APPEALS
                                   FOR THE SIXTH CIRCUIT
                                     _________________


                                                     X
                               Plaintiff-Appellant, -
 CANDICE DAVIS,
                                                      -
                                                      -
                                                      -
                                                          No. 07-1512
          v.
                                                      ,
                                                       >
 MICHIGAN BELL TELEPHONE COMPANY,                     -
                              Defendant-Appellee. -
                                                     N
                      Appeal from the United States District Court
                     for the Eastern District of Michigan at Detroit.
                      No. 06-10513—Sean F. Cox, District Judge.
                                       Argued: June 3, 2008
                             Decided and Filed: September 29, 2008
                  Before: ROGERS, COOK, and McKEAGUE, Circuit Judges.
                                       _________________
                                            COUNSEL
ARGUED: Charles Gottlieb, GOTTLIEB & GOREN, Bingham Farms, Michigan, for Appellant.
David R. Deromedi, DICKINSON WRIGHT, Detroit, Michigan, for Appellee. ON BRIEF:
Charles Gottlieb, GOTTLIEB & GOREN, Bingham Farms, Michigan, for Appellant. David R.
Deromedi, Ryan K. Mulally, DICKINSON WRIGHT, Detroit, Michigan, for Appellee.
                                       _________________
                                           OPINION
                                       _________________
        ROGERS, Circuit Judge. Plaintiff Candice Davis appeals the district court’s grant of
summary judgment in favor of the defendant, Michigan Bell Telephone Company, in this FMLA
action. The defendant terminated Davis from her job as a customer service representative in
February of 2005 because of excessive unexcused absences. Davis had sought FMLA leave with
respect to those absences, but her request was denied on the ground that she had not worked enough
hours in the preceding twelve months to be eligible for FMLA benefits. After being fired, Davis
brought this action against the defendant, alleging that the defendant had interfered with her rights
under the FMLA, retaliated against her for exercising those rights, and failed to give proper notice
of ineligibility as required by the FMLA. Finding that Davis was ineligible for FMLA benefits in
2005, and that Davis could not maintain a cause of action based on improper notice of ineligibility,
the district court granted summary judgment in favor of the defendant. On appeal, Davis argues that
the district court erred in finding her ineligible for FMLA benefits because her eligibility from 2004
should have carried over into 2005. In the alternative, she argues that the defendant should be


                                                  1
No. 07-1512                 Davis v. Mich. Bell Tel. Co.                                                       Page 2


estopped from challenging her FMLA eligibility, and that she is entitled to recover because of the
defendant’s failure to provide proper notice of ineligibility. Because these arguments are unavailing,
the district court judgment is affirmed.
        Davis was hired by the defendant as a customer service representative in November of 1997.
She began suffering from depression in 1999. Prior to September of 2004, Davis requested FMLA
benefits on several occasions, but was denied on the ground that she was ineligible for FMLA
benefits. An eligible employee is one who has been employed for at least twelve months by the
employer with respect to whom leave is requested, and who has been employed by that employer
for at least 1,250 hours of service during the twelve-month period immediately preceding the
commencement of the leave. 29 U.S.C. § 2611(2)(A); 29 C.F.R. § 825.110(a)(2). Davis failed to
qualify for FMLA eligibility prior to September of 2004 because she did not meet the 1,250-hour
requirement. On September 24, 2004, however, she reached that mark and therefore became eligible
for FMLA leave. As soon as she became eligible, her healthcare provider certified that Davis
suffered from depression, and that the depression was a chronic condition that would require
intermittent FMLA leave. Based on this certification, the defendant approved intermittent FMLA
leave for Davis’s depression.
        Between September 24 and December 13, Davis’s depression caused her to take several
discrete absences from work. Each time she was absent, she sought FMLA leave by submitting the
medical certification forms required to obtain FMLA leave. On each of these forms, Davis’s
healthcare provider indicated that Davis’s absences were caused by her depression. The defendant
accepted this documentation and considered these absences to be a part of the intermittent FMLA
leave that had begun on September 24. These absences were approved as FMLA leave without the
defendant’s having reevaluated Davis’s eligibility for FMLA leave on the date of each absence
because the defendant concluded that Davis’s eligibility should only be evaluated at the
commencement of the intermittent leave, which occurred on September 24.
        On December 13, 2004, Davis began an absence that continued into 2005.1 She never
submitted the medical certification form required for FMLA leave, although she notified the
defendant of her intent to seek FMLA leave during the first week of this absence. After the seventh
day of her absence — including weekends — the defendant filed a claim for short-term disability
benefits on Davis’s behalf. At that time, the defendant also notified Davis that she had the right to
apply for FMLA benefits in the event that she was denied short-term disability leave. Short-term
disability leave is an employment benefit that is provided by the defendant, but not required by law.
As is allowed by the FMLA, time spent on short-term disability leave is counted against an
employee’s annual twelve-week entitlement to FMLA leave.



         1
           Davis argues that the continuous absence lasting into the new year actually began on December 7 rather than
December 13. The district court, however, found that Davis’s absence began on December 13. See Davis v. Mich. Bell
Tel. Co., No. 06-10513, 

, at *1 (E.D. Mich. March 19, 2007). This finding cannot be reversed because
Davis has not satisfied her burden of demonstrating it to be clearly erroneous. See Bailey v. USF Holland Inc., 

, 885 (6th Cir. 2008) (citing Isabel v. City of Memphis, 

, 411 (6th Cir. 2005)). In fact, the evidence in
the record convincingly shows that Davis returned to work after December 9 and then began another period of absence
on December 13. For example, the disability approval notice generated by the defendant identifies December 13 as the
first day of Davis’s absence. See J.A. at 25. The defendant’s attendance and FMLA records also show that Davis was
absent from December 7 until December 9 and then was absent again from December 13 through the end of the year.
The records do not, however, indicate that Davis was absent on December 10, 11, and 12. See 

297, 379-80.
The most convincing evidence, though, is found in Davis’s own words. In a letter to her mental healthcare provider,
Davis said, “I have been off since Mon. 12/13,” 

and in a letter to the defendant’s FMLA Processing Unit,
Davis said, “Also, the fmla eligibility form did not include the first date of absence which commenced on 12/13/04 and
continued through 1/14/05.” 

On this record, the district court’s conclusion that Davis’s absence began on
December 13 is beyond question.
No. 07-1512                  Davis v. Mich. Bell Tel. Co.                                                       Page 3


       During her short-term disability leave, Davis informed her therapist that she wanted to return
to work on December 27, 2004. Nevertheless, Davis remained absent into the new year. On
January 7, 2005, Davis’s therapist — according to the defendant — informed the defendant that
Davis was no longer disabled and could have returned to work on January 3. As a result, the
defendant notified Davis on January 12 that only those absences from December 20, 2004, until
January 2, 2005, would be approved as paid short-term disability leave; every absence after January
2 was to be considered an unexcused absence unless Davis sought and received FMLA leave for
those absences. Davis was further notified that she would be considered to have resigned if she did
not report to work on January 14. See J.A. at 99.
       January 14 came and went without Davis’s reporting to work. She did, however, show up
for work on January 15, but by that time, she had missed her opportunity to save her job. As she had
already been disciplined on three separate occasions for poor attendance, the defendant’s attendance
policy mandated that Davis be dismissed. Accordingly, she was suspended pending her dismissal.
        Davis requested a review board hearing regarding her pending dismissal. On February 8,
2005, she also requested that her unexcused absences be approved as FMLA leave. Before Davis
submitted any medical documentation to support her request, the defendant determined that Davis
was not eligible to receive FMLA benefits in 2005. This determination was based on the
defendant’s assumption that FMLA eligibility could not carry over from one twelve-month FMLA
period to the next. Because the defendant measured its twelve-month FMLA period according to
the calendar year, it concluded that Davis’s eligibility should be reevaluated as of the first unexcused
absence in 2005, which occurred on January 3.2 This reevaluation revealed that Davis had not
worked 1,250 hours in the twelve months preceding January 3. The defendant therefore refused to
convert Davis’s unexcused absences from January 2005 into FMLA leave. Davis was also denied
relief by the review board and a grievance proceeding, and she was therefore terminated on February
14, 2005.
       After her termination, Davis continued to seek FMLA leave. On March 16, 2005, her
healthcare provider submitted the requisite medical certification form. It indicated that her absences
from December 13, 2004 to January 14, 2005 had been caused by her depression. In response, the
defendant sent Davis a letter on March 18, 2005 that again informed her that she was not eligible
for FMLA leave with respect to her absences from January 3 to January 14.
        Davis subsequently filed suit against the defendant on the grounds that it had retaliated
against her for asserting her rights under the FMLA, interfered with her FMLA rights by denying
FMLA benefits for her absences occurring after January 2, 2005, and failed to provide proper notice

         2
           The defendant selected this date for reevaluation of Davis’s eligibility because Davis’s absences on January
1 and 2 were covered by short-term disability leave. According to the defendant, short-term disability leave is separate
and distinct from FMLA leave, and therefore, FMLA leave would not have begun until after the short-term disability
leave ended. Thus, according to the defendant, January 3 is the appropriate date for determining Davis’s eligibility
because that is the date on which her FMLA leave would have commenced in 2005. See 29 C.F.R. § 825.110(d) (stating
that eligibility determination is made as of commencement of leave). One could argue, however, that Davis’s FMLA
leave in 2005 commenced on January 1 instead of January 3 because January 1 was the first date in the new year on
which she was absent due to her FMLA-qualifying serious health condition of depression. Cf. Bloom v. Metro Heart
Group of St. Louis, Inc., 

, 1031 (8th Cir. 2006) (holding that employee’s FMLA leave began on the date
she qualified for a serious health condition, and date on which she received paid leave is irrelevant to determining when
FMLA leave began). The problem with the defendant’s position is that it could allow employers to interfere with FMLA
rights by placing employees with FMLA-qualifying illnesses on leave that is nominally not FMLA leave, and then
cancelling that leave and terminating the employees once they become ineligible for FMLA benefits. The question need
not be resolved here, however, because it is evident that Davis would have been found ineligible on either January 1 or
January 3. Therefore, because it makes no difference, and because the parties have framed the alternative dates for
evaluating eligibility as either September 24, 2004, or January 3, 2005, we assume — without deciding — that January
3 would be the appropriate date for reevaluating Davis’s FMLA eligibility in the new year.
No. 07-1512             Davis v. Mich. Bell Tel. Co.                                           Page 4


of her ineligibility. Both sides submitted motions for summary judgment. Davis argued that
because her absences in January of 2005 were part of a continuous absence that began in 2004, the
absences in January of 2005 were part of the intermittent FMLA leave that had been approved in
September of 2004. She claimed that she was therefore entitled to judgment as a matter of law
because her eligibility should have been evaluated as of the beginning of that leave in September
of 2004. The defendant, however, argued that FMLA eligibility for intermittent leave cannot be
carried over from one twelve-month FMLA period to the next. According to the defendant, Davis’s
eligibility therefore had to be reevaluated as of her first request for FMLA leave in 2005. The
district court agreed with the defendant, holding that “[a]n employer may recalculate an employee’s
FMLA eligibility at the commencement of the first absence in a new 12-month period as designated
by the employer.” Davis, 

, at *4. Based on this reasoning, the district court
concluded that the defendant had not interfered with Davis’s FMLA rights because Davis had not
worked enough hours in the preceding twelve months to be eligible for FMLA benefits with respect
to her absences occurring after January 2, 2005. The district court also found that her ineligibility
precluded her retaliation claim. Finally, the district court held that a cause of action cannot be
maintained on the basis of a failure to provide proper notice of ineligibility. Based on these rulings,
the district court granted summary judgment in favor of the defendant.
        The only question with respect to Davis’s FMLA interference claim is whether Davis was
eligible for FMLA leave with respect to her unexcused absences in January of 2005. If she were
eligible, then she would be entitled to proceed with her lawsuit because she would be able to satisfy
the other elements an FMLA interference claim. See Wysong v. Dow Chem. Co., 

, 447
(6th Cir. 2007) (reciting elements of FMLA interference claim (quoting Cavin v. Honda of Am. Mfg.,
Inc., 

, 719 (6th Cir. 2003))). However, because the district court correctly determined
that Davis was not eligible, the defendant is entitled to judgment as a matter of law since an
employee’s ineligibility precludes the employee from pursuing an FMLA interference claim. See
Killian v. Yorozu Auto. Tenn., 

, 556 (6th Cir. 2006) (citing Walton v. Ford Motor Co.,

, 485 (6th Cir. 2005)).
       The determination of FMLA eligibility “must be made as of the date leave commences.”
29 C.F.R. § 825.110(d). If the date is measured as of September 24, 2004, as Davis suggests, then
she was eligible for FMLA leave. If the date is measured in January of 2005, as the defendant
suggests, then Davis was not eligible. Therefore, the question boils down to when Davis’s leave
commenced.
         When an employee has a chronic health condition for which intermittent FMLA leave has
been approved, the leave commences upon the occurrence of the first absence caused by that
condition, and it extends to cover every other absence caused by that condition during the same
twelve-month FMLA period. See Barron v. Runyon, 

, 681 (E.D. Va. 1998). In
other words, each absence subsequent to the first absence is not treated as a separate period of
FMLA leave with its own commencement date. To hold otherwise would render the term
“intermittent leave” meaningless and would effectively read it out of the FMLA since a period of
intermittent leave “must, by definition, comprise periods . . . in which the employee is present at
work.” 

series of absences, separated by days during which the employee is at work,
but all of which are taken for the same medical reason, subject to the same notice, and taken during
the same twelve-month period, comprises one period of intermittent leave.” 

however, can only extend to the end of the twelve-month FMLA period in which it began. See 

Once a new twelve-month FMLA period begins, any additional absences caused by that
same chronic condition would constitute a new period of intermittent FMLA leave. See 

Otherwise, there would be no point at which the initial period of intermittent FMLA leave ended and
a new period commenced. Under that scenario, employees would never have to reestablish their
eligibility for FMLA leave and would therefore be perpetually entitled to twelve weeks of FMLA
leave per year based on a single eligibility determination. See 

Because Congress
No. 07-1512              Davis v. Mich. Bell Tel. Co.                                              Page 5


could not have intended this absurd result, it must be true that a period of intermittent leave cannot
last beyond the specific twelve-month FMLA period in which it begins. Therefore, absences caused
by the same chronic condition, but occurring in different twelve-month FMLA periods, must
constitute different periods of FMLA leave. And as different periods of leave, they must have
different times of commencement. The clear consequence of this is that Davis’s unexcused absences
in January of 2005, if approved as FMLA leave, would have constituted a new period of FMLA
leave that commenced in January of 2005. Therefore, Davis’s FMLA eligibility was appropriately
reevaluated in January of 2005, and the defendant was correct in determining that Davis was not
eligible for FMLA leave with respect to her unexcused absences.
         Davis suggests that it is not proper to reevaluate her eligibility in the middle of her leave, and
that therefore her eligibility should be allowed to carry over until the end of her leave — i.e.,
January 15, the date on which she returned to work. This argument is based on a misunderstanding
of the concept of intermittent leave because it assumes that Davis’s absence from December 13 until
January 15 should be considered a single period of leave simply because it is a continuous period
of absence. A period of intermittent leave, however, is not made up of a single continuous absence.
As explained above, an employee does not begin a new period of leave with each new absence. An
obvious corollary to this rule is that the simple act of returning from an absence does not itself
terminate a period of intermittent leave. Since a period of intermittent leave is not terminated solely
by the act of returning to work, there is no basis for saying that Davis’s intermittent leave terminated
when she returned to work on January 15. But it is obvious that the period of intermittent leave that
began in September of 2004 must end at some point. If the intermittent leave that began in
September of 2004 instead ended upon the beginning of a new twelve-month FMLA period, then
Davis’s request for FMLA leave in 2005, if approved, would have constituted a new period of
FMLA leave commencing in January of 2005. Thus, the ultimate question presented by Davis’s
argument is whether her intermittent leave in 2004 ended upon the occurrence of a new twelve-
month FMLA period, or whether it ended at some arbitrary point, such as her return to work on
January 15. Since the act of returning to work itself does not terminate a period of intermittent
leave, there is no principled reason to conclude that Davis’s intermittent leave should cover absences
up to January 15, but not those occurring thereafter. There is, however, a logical basis for
concluding that Davis’s intermittent leave terminated upon the beginning of a new twelve-month
FMLA period. Because the FMLA speaks in terms of twelve-month periods, see 29 U.S.C. §
2612(a), the most reasonable conclusion is that a period of intermittent leave terminates when a new
twelve-month FMLA period begins. See Sills v. Bendix Commercial Vehicle Sys. LLC, No. Civ.
1:04-CV-149, 

, at *8 (N.D. Ind. Oct. 20, 2005). This means that Davis’s
intermittent leave that began on September 24, 2004 could not have carried over into 2005. As a
result, any intermittent leave sought in 2005 would have been a new period of leave commencing
in January of 2005, therefore requiring a reevaluation of Davis’s eligibility at that time.
      This conclusion is supported by a Department of Labor opinion that provides the following
example:
        Assume an employee is diagnosed with an FMLA-qualifying chronic condition, such
        as MS [multiple sclerosis] as in your example, that results in an employee needing
        intermittent leave due to the episodic nature of the condition. For example, if an
        employee with MS who was eligible to take intermittent FMLA leave in April and
        May needed leave again when the episodes of incapacity recurred in July and again
        in October, the employee would be entitled to FMLA leave without having to re-
        qualify under the 1,250-hour eligibility test so long as the absences occurred within
        the same 12-month period and the employee had not exhausted the 12-week leave
        entitlement for this or any other FMLA-qualifying reason. If the employee needed
        leave for MS again in a new 12-month period, the employee would have to re-qualify
No. 07-1512             Davis v. Mich. Bell Tel. Co.                                           Page 6


       under the 1,250-hour eligibility test to be entitled to take FMLA leave for the same
       chronic condition in the new 12-month period.
Dept. of Labor Op. Letter, FMLA-112, 

(Sept. 11, 2000). This further
demonstrates that intermittent leave cannot carry over into the next twelve-month FMLA period.
Because the example instructs that eligibility must be reevaluated in the new twelve-month FMLA
period, it must be the case that a new period of intermittent FMLA leave commences when a new
twelve-month FMLA period begins. Otherwise, there would be no justification for reevaluating
eligibility.
        The conclusion that Davis’s FMLA eligibility was properly reevaluated in January of 2005
is also supported by the general purposes of the FMLA. The FMLA is intended to balance the
demands of the workplace with needs of families, and entitle employees to reasonable leave for
medical-related reasons while doing so “in a manner that accommodates the legitimate interests of
employers.” 29 U.S.C. § 2601(b)(3). Additionally, the stated purposes of the FMLA demonstrate
“that Congress wanted to ensure that employees’ entitlement to leave and reinstatement did not
unduly infringe on employers’ needs to operate their businesses efficiently and profitably.”
Bachelder v. Am. W. Airlines, Inc., 

, 1120 (9th Cir. 2001). These considerations
suggest that the FMLA and its accompanying regulations should be interpreted to require that an
employee reestablish FMLA eligibility before taking leave in a new twelve-month FMLA period.
It would be unduly burdensome on a business’s need to operate efficiently and profitably if the
business were required to provide an employee with twelve weeks of intermittent leave per year
perpetually based on the fact that the employee was eligible for FMLA benefits on a single day. In
order to accommodate the reasonable interests of businesses, it must be possible to reevaluate
employees’ eligibility at some point, and the only logical method of finding that point is to conclude
that a new period of intermittent leave commences when a new twelve-month period begins. In this
case, Davis’s eligibility was properly reevaluated at such a point, and she was found to be ineligible.
Therefore, since she was not eligible for FMLA leave, she cannot claim that the defendant interfered
with her rights by refusing to approve her request for FMLA leave.
         Davis also argues in the alternative that the defendant should be equitably estopped from
denying her eligibility, and that she is entitled to recovery because the defendant failed to provide
proper notice of her ineligibility. Equitable estoppel does not apply to this case because the
defendant made no material misrepresentation of eligibility, much less a misrepresentation upon
which Davis reasonably relied. See United States v. Guy, 

, 937 (6th Cir. 1992) (citing
Heckler v. Cmty. Health Servs., 

, 59 (1984); United States v. Varani, 

,
1304 (6th Cir. 1986)). According to Davis, the defendant should be estopped from denying her
eligibility because her continuous absence lasting into the new year actually began on December 7
rather than December 13, and the defendant informed her that her absence beginning on December
7 would be covered by FMLA leave. There are two problems with this line of analysis. First, it is
true that Davis was approved for FMLA leave for December 7, 8, and 9, but as previously discussed,
the evidence indicates that she returned to work before beginning on December 13 the absence that
lasted into the new year. See supra note 1. Thus, it makes no difference that the defendant may
have informed Davis that her absence beginning on December 7 would be covered by FMLA leave.
Second, the defendant never made any forward-looking promises with respect to Davis’s FMLA
eligibility. On December 6, the defendant made a representation of eligibility, see J.A. at 294, but
this representation concerned Davis’s eligibility for FMLA leave with regard to her absence from
November 27 until December 4. Likewise, on December 11, the defendant made a representation
of eligibility with respect to Davis’s absences on December 7, 8, and 9. See 

Neither
of these representations addressed whether Davis would be eligible for FMLA leave in the future.
In short, the representations were nothing more than backward-looking statements that made no
promises concerning whether Davis would be able to satisfy the 1,250-hour requirement for
absences occurring after the dates on which the representations were made. Since no affirmations
No. 07-1512             Davis v. Mich. Bell Tel. Co.                                         Page 7


of future FMLA eligibility were made to Davis, the defendant cannot be equitably estopped from
denying Davis’s eligibility.
         Finally, Davis gains no traction from her argument concerning the defendant’s failure to
provide proper notice of ineligibility. Although not well developed in her briefs, Davis’s argument
can only be interpreted as a claim that defendant failed to provide notice of ineligibility in
compliance with 29 C.F.R. § 825.110(d). Two provisions of § 825.110(d) are potentially at play
here. One prohibits an employer from subsequently challenging an employee’s FMLA eligibility
if the employer confirmed the employee’s eligibility at the time the notice for leave was received.
The other says that an employee will be deemed eligible if the employee is not informed of her
ineligibility within two business days of the date on which eligibility is determined, or in some
cases, within two business days of the employer’s receipt of the employee’s notification of the need
for FMLA leave. Neither provision, however, is of aid to Davis. The first provision does not apply
here because the defendant did not confirm Davis’s eligibility for FMLA benefits at the time it
received her notice for leave. Moreover, the notification requirements in the latter provision are of
no aid to Davis because they are invalid due to the fact that they impermissibly expand FMLA
eligibility beyond the parameters established in 29 U.S.C. § 2611(2). See, e.g., Knapp v. Am. W.
Airlines, 207 F. App’x 896, 900 (10th Cir. 2006); Woodford v. Cmty. Action of Greene County, Inc.,

, 55-56 (2d Cir. 2001); Evanoff v. Minneapolis Pub. Sch., Special Sch. Dist. No. 1, 11
F. App’x 670, 671 (8th Cir. 2001); Brungart v. BellSouth Telecomms., Inc., 

, 796-97
(11th Cir. 2000); Dormeyer v. Comerica Bank-Illinois, 

, 582 (7th Cir. 2000);
Massengill v. Anderson County Bd. of Educ., 

, 1007 (E.D. Tenn. 2007).
Therefore, neither of Davis’s alternative arguments is sufficient to rescue her FMLA interference
claim.
        Davis’s retaliation claim likewise fails. The basis for this claim is rather unclear, and
Davis’s briefs shed little light on the issue. To the extent that she is claiming that she was
terminated because of her attempt to obtain FMLA leave in January of 2005, her claim must fail as
a matter of law because she was not eligible for FMLA benefits with respect to her January 2005
absences. See Humenny v. Genex Corp., 

, 905-06 (6th Cir. 2004) (holding that an
FMLA retaliation claim cannot be maintained by one who was not an “eligible employee”). This
appears to be the correct interpretation of Davis’s argument since she concedes that she cannot
prevail without proving that she was eligible for FMLA benefits. However, in the event that Davis
is claiming that she was retaliated against for taking FMLA leave in 2004 — when she was eligible
— then her claim would still fail because she has presented no evidence of a causal connection
between her FMLA leave and her termination. See 

(citing Arban v. West
Publ’g Corp., 

, 404 (6th Cir. 2003)). Thus, no matter how this claim is interpreted, it
must fail as a matter of law.
       For the foregoing reasons, the district court’s grant of summary judgment to the defendant
is AFFIRMED.