MELLOY, Circuit Judge.

Kimberly Anne Nowling applied for disability benefits under Title II of the Social Security Act (the "Act"), 42 U.S.C. § 401, et seq., and supplemental security income benefits under Title XVI of the Act, 42 U.S.C. § 1381, et seq. An ALJ denied her application, and the Commissioner declined further review. Nowling unsuccessfully challenged her denial of benefits in the district court. We remand to the agency for further proceedings.

I.

Nowling suffers from conversion disorder manifesting itself as somatoform, non-epileptic "pseudo-seizures." In addition, she is obese and suffers from migraine headaches, mood disorder, anxiety disorder, and personality disorder.

Before turning to the details of Nowling's medical and administrative history, we describe briefly the phenomenon of conversion disorder and the somatoform, non-epileptic "pseudo-seizures" that form the core of Nowling's medical issues and alleged disability. Conversion disorder is a phenomenon in which a person actually and subjectively experiences symptoms without a known underlying medical cause. Easter v. Bowen, 867 F.2d 1128, 1129 (8th Cir.1989). It is believed the symptoms, such as non-epileptic seizures, result from an unconscious, involuntary conversion of mental stress into a physiological symptom. Id.; see also Miller v. Colvin, No. 12-2293, 2014 WL 641714 at *3 (W.D.Ark. Feb. 19, 2014) ("Pseudoseizures, or psychogenic non-epileptic seizures . . . are believed to occur as a part of a conversion disorder in which the patient unconsciously converts emotional dysfunction into physical symptoms." (citing Ronald P. Lesser and S. Marc Testa, Symptoms That Mimic Epilepsy Linked to Stress, Poor Coping Skills, http://www.hopkinsmedicine.org/news/media/releases/symptoms_that_mimic_epilepsy_linked_to_stress_poor_coping_skills (last accessed January 26, 2016))). In prior opinions, we reviewed these or similar disorders and noted the difficulty of assessing how such disorders limit a person's activities. Metz v. Shalala, 49 F.3d 374, 377 (8th Cir.1995).1 In particular, we noted that a prime feature of conversion disorder may be a disconnect between the actual severity of symptoms demonstrated by clinical evidence and the way the applicant subjectively perceives the symptoms. See, e.g., Easter, 867 F.2d at 1130 ("[H]er primary disorder, as clinically diagnosed, causes her to exaggerate her physical problems in her mind beyond what the medical data indicate."). That is not to say this exaggerated experience of symptoms amounts to malingering. Id. Rather, the applicant actually believes herself to be experiencing symptoms at a greater level of severity than clinical evidence can support.

Given this disconnect, an obvious difficulty arises when it becomes necessary to make credibility assessments in cases involving somatoform phenomena. Subjective perceptions of somatoform effects may, in fact, be debilitating even when clinical or diagnostic medical evidence does not fully support the claimed symptoms. It nevertheless remains necessary to make credibility assessments in these settings, and "[i]n cases involving somatoform disorder. . . an ALJ is not free to reject subjective experiences without an express finding that the claimant's testimony is not credible." Metz, 49 F.3d at 377. Where such a finding has been made, "[w]e will not disturb the decision of an ALJ who considers, but for good cause expressly discredits, a claimant's complaints . . . even in cases involving somatoform disorder." Gowell v. Apfel, 242 F.3d 793, 796 (8th Cir.2001).

Given the difficulty in this area, if an ALJ expressly accepts that a claimant suffers from a somatoform disorder, but also finds the claimant at least partially non-credible, the ALJ ideally should set forth the credibility determination with sufficient detail to expressly inform the reviewing court as to the factual details of the petitioner's limitations as accepted or believed by the ALJ. And in this type of case, even more so than in other cases, corroborating testimony from actual witnesses such as family members or former employers regarding the nature of the symptoms may hold particular value for a fact finder.

With this background in mind, we turn to the record. Nowling began experiencing seizure-like episodes prior to 2007. She worked as a nurse's aide and bathing assistant. Over time, her employer cut her hours and duties because it was not safe for her to be alone with patients or lift or bathe patients given her potential for seizure-like episodes. In December 2007, following an episode of seizure-like symptoms, her employer asked her to see a neurologist and instructed her not to return to work until she did. Nowling visited neurologists, who identified no neurological cause for her symptoms and identified no signs of brain activity suggestive of actual epilepsy. She returned to work on a reduced basis. Eventually, in August 2008, she quit her job due to an inability to work alone with patients.

On March 6, 2009, she applied for Title II and Title XVI benefits. She originally alleged an onset date of March 6, 2009. Her applications were denied, and at a hearing before an ALJ on March 16, 2011, she amended her alleged onset date to February 28, 2008. She did not allege her condition materially worsened in late February 2008. Rather, February 28, 2008, is a date when she asked for and received a doctor's note recommending that she restrict her working hours to 7 ½ hours per day, three days per week.2

Medical evidence from treating physicians was largely consistent between 2007 and 2011. Nowling saw neurologist Mignon Makos in 2007. Dr. Makos found no evidence of epileptic seizures. Nowling experienced a seizure-like episode during her exam, and Dr. Makos indicated that Nowling was able to sit but a nurse needed to hold her head and neck. Dr. Makos also indicated that Nowling's eyes appeared to react voluntarily to stimulus during the pseudo-seizure and that she appeared to recoil voluntarily in response to discomfort. She recovered quickly after her seizure and experienced little physical change other than temporarily slightly elevated blood pressure. Dr. Makos diagnosed her with conversion disorder/somatoform pseudo-seizures and took her off seizure medication. Nowling reported to Dr. Makos that she frequently suffered migraine headaches and that her headaches often preceded her more severe seizures. Dr. Makos placed her on medication to treat her headaches in an effort to control the pseudo-seizures.

Nowling received treatment from general practitioner Eduardo Fernandez, M.D., from October 2007 through July 2010; psychiatrist Jose Menendez, M.D., from August 2006 through January 2011 (at least 29 office visits); and licensed social worker Nancy Miller from April 2009 through January 2011 (at least 38 visits or therapy sessions). These care providers consistently diagnosed and treated Nowling with conversion disorder manifesting as non-epileptic/pseudo-seizures. During treatment, Miller consistently described Nowling with a GAF score of 45 or 50; and Dr. Menendez described a GAF score of 50.3 No care provider diagnosed her with epileptic seizures nor identified a physically identifiable cause for her seizures. No care provider assigned a GAF score greater than 50 other than Dr. Menendez who, in a "medical source statement form" completed on November 30, 2010, listed a GAF score of 56 as both her current level of functioning and as the highest level of functioning in the preceding year. In the same form, Dr. Menendez described Nowling's impairments as severe, noted that she was not a malingerer, and stated that her impairments would cause her to be absent from work "more than four days per month." Even though he described her impairments in this fashion, the GAF score of 56 generally reflects only a moderate level of impairments.

Treatment notes from all caregivers reflect that Nowling described the frequency of her seizures as waxing and waning over the years. She described them as occurring as often as two or three per week and as infrequently as two to three per month. Nowling reported that headaches of varying intensity, including migraine headaches, often preceded her seizures. She received medicine for anxiety and personality disorder as well as medicine to control headaches. Over the course of her treatment her physicians frequently changed her drugs and dosages in an attempt to better control her headaches and seizures. Although many treatment notes indicate mild improvement, read as a whole they show her symptoms waxed and waned with some short term improvement but without substantial long-term worsening or improvement during the 2008 to 2011 time frame.

In treatment notes, Dr. Fernandez, Dr. Menendez, and Ms. Miller often recited Nowling's reports of her symptoms as well as her reports of how she had been spending her time. These reports were largely consistent with a later description of daily activities Nowling provided during her ALJ hearing. She did not drive, other than a one-mile route to her brother's home, for fear of having an episode and causing an accident. She did not shop or generally go out in public alone. Her father or her sister-in-law generally drove her when needed. She could do yard work "at her own pace," did her own laundry, and cooked for herself and her father twice per week. One day per week, her father would drive her to her sister's home in the morning where she would help her sister's children prepare for school and get on a bus. She typically would watch television for one to one and a half hours per day. She liked to read and work on the computer, but she could only work on the computer for about fifteen to twenty minutes or read for about half an hour before headaches made her stop.

In addition, Nowling described several vacations and family events she attended: at least one instance of camping; an event where she helped to host a large group of family; monthly trips to see her brother in Nebraska; attending and helping with children's sporting events; and vacationing once in western Nebraska and once in North Carolina. Following the trip to western Nebraska, Nowling reported to Dr. Menendez that she felt relaxed and had not suffered a seizure-like episode at all during her trip. Dr. Menendez's notes from that office visit indicate that Nowling "did not appear to have any symptoms in the absence of stress."

To her doctors, to Ms. Miller, and to the ALJ, she described her seizure-like episodes as varying in intensity and effect. Some episodes involved complete loss of consciousness, some involved tremors and shaking, and still others involved an inability to form words and articulate thoughts. The durations varied from five minutes or less to more than twenty minutes, and recovery times following episodes varied from minutes to a whole day. In her hearing testimony, Nowling described her seizures as occurring "in spurts" where she might have two or three in the same day but then go a month without having one. No treating physician suggested any belief or suspicion that Nowling was malingering or exaggerating her symptoms.

In the medical source statement form prepared on November 30, 2010, Dr. Menendez described Nowling as seriously limited or unable to meet competitive standards in several respects, including an inability to "[m]aintain regular attendance and be punctual within customary, usually strict tolerances," "[c]omplete a normal workday and workweek without interruptions from psychologically based symptoms," "[r]espond appropriately to changes in a routine work setting," or "[d]eal with stress of semiskilled and skilled work."

Consulting psychologist John Keough, MA, met with Nowling and conducted a psychological exam in May 2009. He concluded Nowling's general life activities and presentation during her examination did not demonstrate severe limitations. In particular, Mr. Keough recited several of Nowling's self-described activities, discussed her mannerisms and ability to respond to his questioning, and concluded she could function in a low stress/low demand workplace.

In addition, consulting psychologist Margaret Sullivan, PhD, reviewed Nowling's records and found only mild to moderate limitations, reaching conclusions consistent with those of Mr. Keough.

Nowling's sister-in-law, Dawn Nowling, testified extensively about her interactions with Nowling and her observations of Nowling's seizures. Dawn Nowling spent at least some time with Nowling on a daily basis, some days spending most of the day with her. Dawn Nowling reported witnessing approximately forty seizure-like events since 2010.4 She described most events as lasting five to ten minutes with relatively short recovery times, but approximately twenty-five percent of the events lasted for fifteen to twenty minutes or longer and required greater recovery times. Dawn Nowling described Nowling's activities in a manner consistent with Nowling's description: Nowling refused to drive herself or appear in public alone but managed self-care, housekeeping, and yard care on her own pace and participated in family events and other events when accompanied. Dawn Nowling described herself or other family members moving Nowling to a chair or laying her down when seizure-like symptoms would arise. She also described the more mild episodes as akin to someone who has had a stroke and cannot formulate words or focus. Finally, she described the seizure-like episodes as having become a fairly routine occurrence that family members were accustomed to dealing with because they had been happening for a long time and physicians could not determine physical cause or offer effective treatment.

In a written opinion issued after the hearing, the ALJ applied the five-step sequential analysis of 20 C.F.R. § 404.1520 and § 416.920. The ALJ found Nowling non-credible but did not wholly reject her claimed limitations or the existence of the pseudo-seizures.5 Rather, the ALJ found that some level of limitations and medically supported symptoms existed but at a lesser severity than that claimed by Nowling. The ALJ concluded the medical evidence could be expected to cause Nowling's general symptoms. The ALJ, however, did not address the primary feature of conversion disorder and somatoform symptoms, namely, the extent to which Nowling actually perceived symptoms and the extent to which conversion disorder rather than a lack of credibility might explain an absence of objective medical support for her symptoms.

In reaching his conclusions, the ALJ determined Dr. Menendez's opinion (the medical source statement) was entitled to "little weight" because the opinion was internally inconsistent in that it listed a GAF score of 56 but described Nowling's impairments as severe. The ALJ noted that Dr. Menendez's opinion was inconsistent with the opinions of the consulting psychologists and "not well supported." The ALJ also stated Dr. Menendez's opinion was inconsistent with the balance of Dr. Menendez's own treatment notes. The ALJ, however, relied heavily upon: (1) the comment from Dr. Menendez that Nowling's symptoms did not appear to arise in the absence of stress, and (2) comments in Dr. Menendez's treatment notes indicating Nowling showed "improvement." The ALJ gave "greatest weight" to the opinion of Mr. Keough, the psychologist who examined Nowling once. The ALJ also gave "great weight" to the opinion of Margaret Sullivan, PhD, the non-examining psychologist who reviewed Nowling's records. Finally, the ALJ commented upon Nancy Miller only to indicate that Nancy Miller repeated the GAF score of 50, and the ALJ did not mention the testimony from Dawn Nowling.6

The ALJ determined, pursuant to 20 C.F.R. §§ 404.1520(a)(4) and 416.920(a)(4) that: (i) Nowling had not engaged in substantial gainful activity since the amended onset date; (ii) she suffered from the following impairments that, at least in combination, were severe: conversion disorder with pseudo-seizures, headaches, mild obesity, mood disorder, anxiety disorder (not otherwise specified), and personality disorder (not otherwise specified), but medical evidence did not show them to be so severe as to preclude work for 12 months; (iii) she did not have a listed impairment or a combination of impairments that met or medically equaled a listed impairment; (iv) she could not perform her past relevant work; but (v) she retained the residual functional capacity to:

At the hearing, the ALJ presented this residual functional capacity to a qualified vocational expert in three separate hypothetical scenarios. In a first scenario, the ALJ described Nowling's restrictions exactly as quoted. With these restrictions, the vocational expert opined that Nowling could work as a price marker, a router, or a mail folding machine operator—DOT listed positions deemed consistent with these limitations and available in substantial numbers in the state and national economies. 20 C.F.R. §§ 404.1520(e), (g); id. §§ 416.920(e), (g).

In a second scenario, the ALJ again asked the vocational expert to consider the limits described in the residual functional capacity assessment but added seizures with subsequent recovery periods that would make Nowling unable to work for half-day periods on an unpredictable basis, with such periods arising once per week. And in a third scenario, the ALJ described a similar, periodic inability to perform any work but with the inability to work caused by headaches rather than pseudo-seizures. As to both of the latter scenarios (her baseline restrictions as quoted above coupled with a periodic and unpredictable wholesale inability to work due to either seizures or headaches), the vocational expert opined that the resulting absenteeism would be unacceptable to any employer such that there were no jobs Nowling could perform if she had those limitations. The ALJ concluded Nowling was not disabled, accepting the vocational expert's response to the first hypothetical situation and finding Nowling could perform the cited jobs. The Commissioner denied further review, and the ALJ's opinion is the final administrative order. The district court affirmed.

II.

We review de novo the district court's judgment affirming the ALJ's denial of benefits. Grable v. Colvin, 770 F.3d 1196, 1201 (8th Cir.2014). We owe substantial deference to the agency, however, and will "affirm if the ALJ made no legal error and the ALJ's decision is supported by substantial evidence in the record as a whole." Id. "Substantial evidence is less than a preponderance of the evidence, but is `such relevant evidence as a reasonable mind would find adequate to support the commissioner's conclusion.'" Id. (quoting Davis v. Apfel, 239 F.3d 962, 966 (8th Cir.2001)). "In determining whether existing evidence is substantial, this court looks at both evidence that supports and evidence that detracts from the Commissioner's decision." Bergmann v. Apfel, 207 F.3d 1065, 1068 (8th Cir.2000). "We may not reverse simply because we would have decided differently or because substantial evidence supports a contrary outcome." Grable, 770 F.3d at 1201.

Several matters, as determined by the ALJ, are undisputed at this time: Nowling suffers from conversion disorder manifesting itself as somatoform, non-epileptic "pseudo-seizures." In addition, she is obese and suffers from migraine headaches, mood disorder, anxiety disorder, and personality disorder. She cannot continue her past work, and she is subject to physical and work-environment limitations due to her conditions.

Notwithstanding these determinations, several matters remain unclear. The ALJ stated Nowling was non-credible in part, finding she exaggerated her symptoms. But it is unclear whether this determination was a finding of malingering or an exaggeration related to the nature of conversion disorder. Further, the ALJ neither explained how extensive he determined Nowling's symptoms to be nor addressed the effect of her conversion disorder upon her perception of her own symptoms. It therefore remains unclear what symptoms the ALJ actually believed Nowling experiences and what, if any, effect her conversion disorder has upon her ability to work.

In the context of this uncertainty, Nowling argues the ALJ committed legal error in three respects resulting in a conclusion not supported by substantial evidence on the record as a whole. First, she argues the ALJ failed to consider testimony from her sister-in-law, Dawn Nowling, who witnessed Nowling's non-epileptic seizures. Second, she argues the ALJ improperly discounted Dr. Menendez's opinion. Third, she argues the ALJ improperly discounted Nancy Miller's opinion. We address these arguments below and conclude remand is required.

In particular, we conclude the errors identified by Nowling show the ALJ failed to consider Mental Impairment Listing 12.07 Somatoform Disorders (including Conversion Disorder) when assessing Nowling's residual functional capacity. The ALJ's opinion asserts that he "considered whether the claimant's pathology and attending symptoms met or equaled" Listing 12.07. The ALJ, however, did not consider the listing in relation to Nowling's residual functional capacity in the manner required by the regulations governing Mental Disorders, Subpt. P, App. 1 § 12.00.

Section 12.00.D.1. addresses sources of evidence in considering mental impairments. The ALJ's conclusory finding that Nowling was not credible did not take § 12.00.D.1.b. into account. The failure to consider Dawn Nowling's testimony and limited consideration of Nancy Miller's treatment records was contrary to § 1200.D.1.c. The failure to consider the medical evidence "longitudinally" was contrary to § 12.00.D.2., leading to unwarranted emphasis on the one-time evaluation of Keogh and an evaluation of Dr. Menendez based on the November 2010 medical source statement rather than Dr. Menendez's entire course of treatment. Further, as noted above, supra at n. 3, the excessive (almost exclusive) reliance on the GAF scores was an improper basis upon which to discount the substantial longitudinal treatment records of Dr. Menendez. Finally, the ALJ failed to consider whether Nowling's chronic impairment resulted in a structured life that masked the extent to which she is impaired from working, see §§ 1200.E. and F.

We also conclude the ALJ failed to evaluate fully the vocational expert's testimony. In posing the second and third hypotheticals regarding seizures and headaches and recovery times, the ALJ described lesser impairments than Nowling claimed. The vocational expert nonetheless said both conditions were disabling because they would cause unscheduled, excessive absenteeism—the exact reason Dr. Menendez's medical source statement concluded Nowling was disabled. Yet the ALJ made no finding that Nowling's conversion disorder was of less severity than what he described in his hypotheticals.

A. Dawn Nowling

Dawn Nowling is not a medical source, but her testimony reflects on the severity of Nowling's impairments and is evidence the ALJ is to consider in applying the five-step sequential analysis. See 20 C.F.R. § 404.1513(d). It is undisputed the ALJ's opinion failed to address Dawn Nowling's testimony expressly or describe what weight, if any, the ALJ accorded her testimony. In general such an omission need not lead our court to reverse an ALJ's otherwise-supported decision. See, e.g., Buckner v. Astrue, 646 F.3d 549, 560 (8th Cir.2011) (rejecting an argument that the failure to address a lay-witness's description of symptoms required remand, stating, "the ALJ's `arguable deficiency in opinion-writing technique,' had no bearing on the outcome of [the] case and does not require remand" (quoting Robinson v. Sullivan, 956 F.2d 836, 841 (8th Cir. 1992))); Wildman v. Astrue 596 F.3d 959 (8th Cir.2010) ("Although required to develop the record fully and fairly, an ALJ is not required to discuss every piece of evidence submitted." (quoting Black, 143 F.3d at 386)). When coupled with other "errors and uncertainties in the opinion," Willcockson v. Astrue, 540 F.3d 878, 880 (8th Cir.2008), however, we have remanded for reconsideration and clarification by the agency, id. at 881. And, here the failure to consider Dawn Nowling's testimony was contrary to the governing regulations. See 20 C.F.R. Pt. 404, Subpt. P, App. 1, § 1200.D.1.c ("If necessary, information should also be obtained from non-medical sources, such as family members and others who know you, to supplement the record of your functioning in order to establish the consistency of the medical evidence and longitudinality of impairment severity. . . .").

Dawn Nowling testified regarding Nowling's daily activities in a manner consistent with Nowling's testimony and statements to her physicians. In particular, Dawn Nowling stated that Nowling did not shop or otherwise go out publicly by herself for fear of having a seizure-like episode when unaccompanied. The ALJ, however, recited his understanding of Nowling's activities as including shopping by herself.7 The failure to consider Dawn Nowling's testimony and the misstatement of the record in this regard demonstrates a failure to properly analyze the effects of a structured setting as required by the regulations. Id. § 1200.F (describing the potentially mitigating effect of life in structured setting, whether an institution or a home, upon symptoms).8 Simply put, the nature of the medical condition and the nature of the life activities, including such considerations as independence, should be considered against the backdrop of whether such activities actually speak to claimant's ability to hold a job. Participation in activities with family or activities at home and at "your own pace" may not reflect an ability to perform at work. And, "a claimant need not be bedridden to qualify for disability benefits." Hutsell v. Massanari, 259 F.3d 707, 713 (8th Cir.2001).

Citing Buckner, the agency argues any error in failing to discuss Dawn Nowling's testimony does not require remand because the same evidence that discredits Nowling's testimony also discredits Dawn Nowling's testimony. Buckner, 646 F.3d at 559-60 (denying remand where it was not possible to determine whether the ALJ had considered lay testimony but where it was clear the same evidence that discredited the claimant's testimony would also discredit the lay witness's testimony). The agency's argument in this regard misses the mark. First, it is unclear what exactly the ALJ found not credible about Nowling, other than the conclusion that her symptoms were some unarticulated degree of severity less than she described them to be. In this setting, even without taking into account the peculiarities of a somatoform disorder, Dawn Nowling's testimony is neither redundant with Nowling's testimony, nor is it discredited by the same evidence that purportedly discredits Nowling's testimony. Rather, Dawn Nowling's testimony serves as a third-party's observation of the symptoms the ALJ appears to have rejected as non-credible subjective complaints. See Willcockson, 540 F.3d at 881 ("[W]e question whether witness statements corroborating a claimant's subjective complaints can logically be treated as cumulative by assuming that they would have been rejected for the same reasons that the claimant statements were rejected, where the agency itself says that because subjective complaints are hard to document, it will `carefully consider' evidence from other persons addressing the extent of the claimant's pain and how it affects his or her ability to function." (quoting 20 C.F.R. § 404.1529(c)(3))).

Further, the ALJ's opinion and the agency's argument on appeal ignore entirely the general nature of somatoform disorders. As in Easter, "the ALJ's opinion. . . cites . . . somatoform disorder [but] does not adequately consider the effects of that mental condition." 867 F.2d at 1130. The ALJ found Nowling suffered from conversion disorder but at no point accorded any consequence to that fact or to the nuance our court recognized in Easter that somatoform disorders may be disabling and may result in "a distorted perception of . . . physical ailments." Id. at 1131. In this situation, and in light of the additional concerns detailed below, we cannot find the failure to address Dawn Nowling's testimony harmless nor characterize it merely as an "arguable deficiency in opinion-writing technique." Robinson, 956 F.2d at 841

B. Dr. Menendez & Nancy Miller

"A treating physician's opinion `should be granted controlling weight if it is well-supported by medically acceptable clinical and laboratory diagnostic techniques and is not inconsistent with the other substantial evidence in the record.'" Miller v. Colvin, 784 F.3d 472, 477 (8th Cir.2015) (quoting Cunningham v. Apfel, 222 F.3d 496, 502 (8th Cir.2000)). A treating physician's opinion, however, "does not automatically control or obviate the need to evaluate the record as a whole." Id. (quoting Hogan v. Apfel, 239 F.3d 958, 961 (8th Cir.2001)). Rather, "an ALJ may discount or even disregard the opinion of a treating physician where other medical assessments are supported by better or more thorough medical evidence or where a treating physician renders inconsistent opinions that undermine the credibility of such opinions." Id. (quoting Wildman, 596 F.3d at 964 (further citation omitted)).

The ALJ discounted Dr. Menendez's opinion because the GAF score of 56 in the medical source statement was inconsistent with the other experts' opinions and with Dr. Menendez's other opinions and statements. As stated above, supra at n. 3, however, GAF scores are of little value. As such, it is error to disregard such substantial longitudinal treatment based merely upon one of several reported GAF scores.

Further, we note the ALJ improperly accorded great weight to statements in Dr. Menendez's treatment notes indicating that Nowling demonstrated "improvement" without acknowledging that Nowling's symptoms waxed and waned throughout the substantial period of treatment, without acknowledging the unpredictable and sporadic nature of Nowling's symptoms, and without assessing the effect of her structured living environment. See Hutsell, 259 F.3d at 712 (stating that an ALJ may have "relied too heavily on indications in the medical record that [the claimant] was `doing well,' because doing well for the purposes of a treatment program has no necessary relation to a claimant's ability to work or to her work-related functional capacity"); see also 20 C.F.R. Pt. 404, Subpt. P, App. 1, § 1200.F ("If your symptomatology is controlled or attenuated by psychosocial factors, we must consider your ability to function outside your highly structured settings." (emphasis added)). The ALJ's superficial references to the GAF score and Nowling's "improvement" fail to satisfy the regulations' "require[ment] that the ALJ `always give good reasons' for the weight afforded to a treating physician's evaluation." Reed v. Barnhart, 399 F.3d 917, 921 (8th Cir. 2005) (quoting 20 C.F.R. § 404.1527(d)(2)). Here those reasons must be articulated with acknowledgment of the nature of the disorder at issue and with consideration given to Dawn Nowling's corroborating testimony.

Finally, Nowling argues the ALJ failed to consider the extensive treatment records from Nancy Miller. Nancy Miller, the licensed clinical social worker/therapist who treated Nowling over the course of 38 visits and who consistently rated Nowling's impairments as severe, is not an "acceptable medical source," 20 C.F.R. § 404.1513(a), but is an "other medical source," id. § 404.1513(d), whose opinion the ALJ is to consider when assessing the severity of an impairment and how it affects the ability to work. See id. § 416.913. Generally, "[i]n determining what weight to give `other medical evidence,' the ALJ has more discretion and is permitted to consider any inconsistencies found within the record." Raney, 396 F.3d at 1010 (quoting 20 C.F.R. § 416.927(d)(4)). The agency itself, however, has instructed that the opinion of an "other medical source" generally should be considered pursuant to several pertinent factors, including the length of treatment. See Social Security Ruling (SSR) 06-3p, 71 Fed.Reg. 45,593, 2006 WL 2329939, at *6 (Aug. 9, 2006);9 20 C.F.R. Pt. 404, Subpt. P, App. 1 § 1200.C. ("Other professional health care providers . . . can normally provide valuable functional information, which should be obtained when available and needed."). In light of the extensive treatment history between Nancy Miller and Nowling, and the consistency between Nancy Miller and Dr. Menendez's treatment records, it was error to disregard Nancy Miller's testimony based solely on the reference to GAF scores. See supra n. 3 & 6.

III. Conclusion

For the reasons stated, we remand with instructions to the district court to remand this case to the agency for further consideration.