I. ANALYSIS

Preliminarily, the Court notes that the appellant presents no argument challenging either the 70% disability rating for his PTSD, or the effective date of that rating. Therefore, the Court will not examine these issues. See Williams v. Principi, 15 Vet.App. 189, 199 (2001) ("[O]rdinarily this Court will not review issues that are not raised to it."); Ford v. Gober, 10 Vet.App. 531 (1997).

The Court reviews the Board's decision regarding the assignment of a disability rating and the assignment of an effective date under the "clearly erroneous" standard. See Link v. West, 12 Vet.App. 39, 46 (1998); Smallwood v. Brown, 10 Vet.App. 93, 97 (1997). A finding is clearly erroneous when, although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed. Gilbert v. Derwinski, 1 Vet.App. 49, 52 (1990).

In rendering a decision, the Board must include in its decision a written statement of the reasons or bases for its findings and conclusions on all material issues of fact and law presented on the record. That statement must be adequate to enable an appellant to understand the precise basis for the Board's decision and to facilitate informed review in this Court. See 38 U.S.C. § 7104(d)(1); Allday v. Brown, 7 Vet.App. 517, 527 (1995); Gilbert, 1 Vet.App. at 57. To comply with this requirement, the Board must analyze the credibility and probative value of the evidence, account for the evidence it finds persuasive or unpersuasive, and provide the reasons for its rejection of any material evidence favorable to the claimant. See Caluza v. Brown, 7 Vet.App. 498, 506 (1995).

A. Separate Rating for Alcohol and Substance Abuse

The appellant repeats an argument asserted below that he is entitled to a separate and distinct rating for his alcohol and substance abuse, even though these conditions are compensable only if they are secondary to another service-connected condition. See Allen v. Principi, 237 F.3d 1368, 1381 (Fed. Cir. 2001) ("Veterans can only recover if they can adequately establish that their alcohol or drug abuse disability is secondary to or is caused by their primary service-connected disorder."). Presumably, he believes that such a separate rating might have assisted him in meeting the percentage standards for a TDIU award at an earlier date. See Appellant's Brief at 16 (claim for earlier effective date for TDIU inextricably intertwined with separate disability rating for alcohol and substance abuse); 38 C.F.R. § 4.16(a) (2008).

The Secretary responds that the Board sufficiently explained why it rated the secondary alcohol and substance abuse disability as part of the rating process for the primary disability. VA made it clear that it was following that procedure:

[T]he RO granted entitlement to an initial evaluation of 50 percent for PTSD. At the Board hearing the veteran raised the intertwined issue of service connection for substance abuse on a secondary basis but the record shows that the rating increase of 50 percent was based in part on a VA medical opinion that linked the veteran's substance abuse to his PTSD. Thus, the veteran should understand that this matter has been resolved in his favor and this rating was explained to him in the September 2002 Statement of the Case.

R. at 360. Neither the Board nor the Secretary, however, cites any authority giving it the latitude to choose this approach to the rating process.

At first glance, the regulatory language and caselaw thereunder appear to be consistent with the Board's rating approach in this case. See Allen, 237 F.3d at 1381 ("[Regulation defines] secondary disabilit[y] as part of the original disability even when the secondary disability involves an alcohol or drug disability . . . ."); 38 C.F.R. § 3.310(a) (2008) ("When service connection is thus established for a secondary condition, the secondary condition shall be considered a part of the original condition.").

This Court, however, has analyzed § 3.310(a) and concluded that "we can find no evidence, persuasive or otherwise, that § 3.310(a) was intended to govern any element of disability benefit decisionmaking other than the determination of service connection." Roper v. Nicholson, 20 Vet.App. 173, 180 (2006). The Roper Court went on to reject a claimant's assertion that § 3.310(a) governed the rating process for secondarily service-connected conditions, id., and held that the language of this regulation did not preclude VA from separately rating a secondarily service-connected condition and then combining the ratings under 38 C.F.R. § 4.25, id. at 181. See also Ellington v. Peake, 541 F.3d 1364 (Fed. Cir. 2008) (endorsing reasoning of Roper).

There is also dicta in the Roper case that could be read to require a separate rating of a secondarily service-connected condition. See, e.g., id. at 178 ("[[T]he regulatory language] actually requires this result."); see also Ross v. Peake, 21 Vet.App. 528, 532 (2008) ("[T]he additional [secondary] disability is separately rated under the VA schedule for Rating Disabilities."); but see Suttman v. Brown, 5 Vet.App. 127, 134 (1993) (Board may determine whether to increase rating for beriberi due to heart disease, or alternatively, to assign a separate rating for the heart disability as a secondary condition).

While VA may definitely issue a separate rating for a secondarily service-connected condition, it is not clear from this Court's caselaw whether or under what conditions VA may elect to rate that condition by increasing the rating of the primary disability. It is also not clear that a claimant is entitled to his or her choice of rating approaches, as the appellant contends.

The Court is unaware of what regulations, or interpretations thereof, that the Board may have contemplated in choosing to rate the appellant's secondary alcohol and substance abuse in the manner it did. Accordingly, the best course is to remand for a statement of reasons and bases that analyzes the question, which was presented below but not discussed in the Board decision. In this manner, the Court will likely obtain a better record for appellate review and "' protect agency administrative authority' regarding the interpretation of its own issuances." McCormick v. Gober, 14 Vet.App. 39, 45 (2000) (quoting Maggitt v. West, 202 F.3d 1370, 1377 (Fed. Cir. 2000)).

B. Extraschedular Consideration

The appellant argues that remand is required because the Board provided inadequate reasons and bases for not referring the case for extraschedular consideration pursuant to 38 C.F.R. § 3.321(b)(i). The purpose of extraschedular evaluation is "[t]o accord justice . . . to the exceptional case where the schedular evaluations are found to be inadequate." 38 C.F.R. § 3.321(b)(i). "The governing norm in these exceptional cases is: A finding that the case presents such an exceptional or unusual disability picture with such related factors as marked interference with employment or frequent periods of hospitalization as to render impractical the application of the regular schedular standards." Id.

The issue with respect to "marked interference with employment" is whether the interference is both unusual and beyond that contemplated by the rating code(s) employed. In order to obtain extraschedular evaluation, a claimant must submit evidence that the disability affects employability in ways not contemplated by the rating schedule. See Thun v. Peake, 22 Vet.App. 111, 115 (2008) (citing with approval VA Gen.Coun. Prec. 06-96 (August 16, 1996)). If the rating criteria "reasonably describe the claimant's disability level and symptomatology, then the claimant's disability picture is contemplated by the rating schedule . . . and no referral [for extraschedular consideration] is required." Id.

In essence, the appellant argues that his pattern of alcohol and substance abuse constitutes marked interference with employment that merits extraschedular consideration. All of the compensable ratings for PTSD, however, consider varying degrees of occupational impairment. See 38 C.F.R. § 4.130, Diagnostic Code 9411 (2008). Moreover, regardless of how VA ultimately approaches the rating question for the alcohol and substance abuse, that rating will also reflect the occupational impairment from that condition.

The appellant advances no other evidence as constituting marked interference with employment, beyond that contemplated by the rating codes. Therefore, the Court affirms the Board's decision that referral for extraschedular consideration was not indicated. The Court also views the Board's statement that "[there is] nothing in the presentation on the examinations, and outpatient reports that would reasonably raise a question as to the applicability of the schedular standards in this case" (R. at 14), to be sufficient from the standpoint of reasons and bases.

The appellant alludes briefly to 38 C.F.R. § 4.16(b), which deals with another type of extraschedular consideration grounded on a finding that the claimant is "unable to secure and follow a substantially gainful occupation." There seems to be a widespread and unfortunate tendency to conflate this provision with 38 C.F.R. § 3.321, discussed above. This Court has recognized that "the effect of a service-connected disability appears to be measured differently for purposes of extraschedular consideration under 38 C.F.R. § 3.321(b)(1) . . . and for purposes of a TDIU claim under 38 C.F.R. § 4.16." Kellar v. Brown, 6 Vet.App. 157, 162 (1994). While the former requires marked interference with employment, the latter requires evidence of unemployability. Id. Insofar as there is evidence pertaining to the appellant's ability to "secure and follow a gainful occupation," it may be considered on remand of the TDIU claim discussed below. Such evidence, however, is irrelevant to extraschedular consideration under 38 C.F.R. § 3.321.

C. Effective Date for TDIU

The parties are in agreement that the Board's statement of reasons and bases is inadequate to sustain its determination of an effective date for TDIU. In particular, the Secretary accepts the appellant's argument that the effective date is inextricably intertwined with the remanded claims for depression and dysthymia. There are other defects in the Board's statement of reasons and bases, which reinforce the need for remand.

The Board did not analyze whether the appellant was unable to secure and follow a substantially gainful occupation. See 38 C.F.R. § 4.16(b). Although the RO noted, in multiple rating decisions, the appellant's reported income of $4,000 in a single year from odd jobs, it is not clear that this income amounted to a substantially gainful occupation. Although the term "substantially gainful occupation" is not defined in VA regulations, the Court has held that it requires that a person's earnings "exceed[] the poverty threshold for one person." Bowling v. Principi, 15 Vet.App. 1, 7 (2001) (citing Faust v. West, 13 Vet.App. 342, 355-56 (2000)). There is no evidence in the record whether this reported income exceeded the poverty threshold, and more generally whether and at what times the appellant may have been able to secure and follow a substantially gainful occupation. The Board has therefore not sufficientlyexplained its implied finding that referral for extraschedular evaluation under 38 C.F.R. § 4.16(b) was not warranted.

The Court also agrees with the appellant that the VA medical examination reports were inadequate to support the determination the Board made with respect to the effective date for the TDIU claim. The Board placed heavy reliance on the GAF scores reported by the VA examiners without examining the basis for the assignment of those ratings. The appellant correctly noted below that in the years leading up to his March 15, 2001, claim, various VA treatment records indicate GAF scores ranging from 30 to 45. When the VA examiners in August and December 2001 stated that the GAF score was 55, they included no explanation for the apparent improvement and pointed to no evidence supporting a significant change. See Stefl v. Nicholson, 21 Vet.App. 120, 124-25 (2007) (holding that "a mere conclusion by a medical doctor is insufficient to allow the Board to make an informed decision as to what weight to assign to the doctor's opinion"). Similarly, when another VA examiner indicated a GAF score of 45 in June 2005, there was no discussion of the evidence that called for such a significant downward change, the result of which is more consistent with the evaluations in the earlier treatment notes.

The Board stated that it gave "more weight to the functioning assessment on the formal examinations since they were derived from a review of the claims file and interview reports and took place over an extended period that showed a consistent level of functioning." R. at 12. The claims file review does little to enhance the reasoning of the examiner or the Board. A claims file review would have included all of the GAF scores in previous VA treatment records that were significantly different and therefore called for some explanation of the difference. Neither the Board nor the VA examiners point to any other evidence from the appellant's medical history that supports the GAF score of 55. Moreover, the passage of four months between the two reports that record a GAF of 55 is hardly "an extended period." It is also not clear that the December 2001 examiner made any independent assessment of the GAF score, merely noting that "[s]ince [the appellant's] GAF score is 55, [he] is not unemployable." R. at 276.

The Board further noted that the appellant was not participating in a formal treatment program or taking any medication at the time of the 2001 VA examinations and that he had briefly begun a treatment program at the time of the June 2005 examination that included a GAF score of

45. These factors are not meaningful as explanatory factors for at least two reasons. First, neither the symptoms listed in the VA rating code nor the factors for assessing GAF scores in the Diagnostic and Statistical Manual of Mental Disorders (4th ed. 1994) (DSM-IV) include treatment history. Second, it is not clear whether the appellant's decision to forgo formal treatment indicates an improvement in his condition or impaired judgment.

In sum, the reasons for the wide variance and gyrations in GAF scores over a four-to-five year period have not been adequately explained by medical reasoning. Therefore, the Board's statement of reasons and bases for its assignment of the effective date for TDIU is unsatisfactory in this respect as well.

For all of the foregoing reasons, the Court remands the rating for the secondarily service-connected alcohol and substance abuse and the effective date for the TDIU claim for readjudication. On remand, the appellant will be free to submit additional evidence and argument as to his request for a separate disability rating for his secondarily connected substance abuse claim and as to unemployability resulting from his service-connected disabilities at any time since he filed his 2001 TDIU claim. The Board is required to consider any such evidence and argument. See Kay v. Principi, 16 Vet.App. 529, 534 (2002). A final decision by the Board following the remand herein ordered will constitute a new decision that, if adverse, may be appealed to this Court upon the filing of a new Notice of Appeal with the Court not later than 120 days after the date on which notice of the Board's new final decision is mailed to the appellant. Marsh v. West, 11 Vet.App. 468, 472 (1998).