[PUBLISHED]
HANSEN, Circuit Judge.Robert Ellis appeals from the district court’s1 judgment affirming the Commissioner’s denial of Ellis’s claim for disability benefits. We affirm.
I.
Mr. Ellis filed an application for Supplemental Security Income (SSI) on October 10, 2000,2 and the ALJ held a hearing on Ellis’s application on November 14, 2001. The ALJ denied the application for benefits on January 25, 2002, and the Appeals Council denied Ellis’s request for review on April 11, 2002, making the ALJ’s decision the final agency decision.
Ellis alleged in his October 14, 2000, Disability Report that he has been disabled since 1993 due to back and leg problems, medications, chronic pain, hepatitis C, and limited physical activity. (Admin. R. at 102.) Ellis has been on strong narcotic pain medications for over ten years, and noted that “medicine is the only thing that relives [sic][his] pain.” (Id. at 123.) During the November 14, 2001, hearing, Ellis testified that in addition to the impairments he had during his previous disability hearing in 1998, he had been diagnosed with hepatitis C and non-insulin-dependent diabetes within the last three years. (Id. at 34-36.) Ellis also testified that he had been admitted to the hospital with a blood clot in his left leg, which was diagnosed as thrombosis, one month prior to the hearing. (Id. at 36.) Ellis testified that a recent CT scan of his shoulder and neck revealed either a pinched nerve or torn rotator cuff in his right shoulder, for which he planned to have an operation. (Id. at 37-38.) Ellis stated that problems with his back, particularly degenerative disc disease, and his inability to sit or stand in one position for any period of time would prevent him from doing the types of jobs suggested during his prior hearing, such as a small parts bench assembler or returned-goods sorter. (Id. at 39-40.)
Ellis’s treating physician, Dr. Patrick Johnson, provided an opinion dated November 21, 2001, in which he stated that Ellis had multiple chronic medical problems that rendered him incapable of performing any sustained gainful employment. (Id. at 279.) The opinion listed Ellis’s medical problems as severe chronic pain in his back and hip as a result of multiple vehicle accidents, diabetes, hepatitis C, recurrent episodes of deep vein thromboem-bolism (a blood clot in the thigh or leg) (DVT), severe dental caries, and hypertension. (Id.) Dr. Johnson opined that Ellis was “incapable of even the most sedentary work because he cannot sit or stand probably for more than one hour at a time due to his chronic pain and his tendency toward developing blood clots in his legs.” (Id.) Dr. Johnson also submitted a “Medical Source Statement of Ability to do Work-Related Activities (Physical),” dated October 1, 2001, in which he opined that Ellis could walk a total of two hours in an eight-hour period, but only a half hour *992without interruption, and that he could sit for a total of four hours in an eight-hour period, but only one hour without interruption. (Id. at 268.)
Ellis has seen Dr. Johnson on a regular basis since 1993, primarily for chronic back and hip pain. The vast majority of Ellis’s contacts with Dr. Johnson during the time period since Ellis’s previously denied application for disability benefits in 1999 have been routine examinations for his chronic back pain and refilling pain prescriptions. (Id. at 273-76.) During each visit, Dr. Johnson continued Ellis’s regular course of medication and filled prescriptions by phone on a monthly basis between visits. (Id.) Dr. Johnson noted that Ellis’s hepatitis C was quiescent during a September 5, 2000, visit. (Id. at 276.) He also noted at an April 9, 2001, visit that Ellis’s chronic back and hip pain were stable on his current medications. (Id. at 274.) Ellis first complained of shoulder pain during his August 9, 2001, checkup. (Id. at 272.) Dr. Johnson increased Ellis’s OxylR prescription and gave Ellis a DepoMedrol injection for the shoulder pain. Dr. Johnson noted probable rotator cuff tendinitis, as opposed to a rotator cuff tear, and scheduled an MRI. (Id.) The tendinitis and pinched nerve were confirmed by an MRI of the right shoulder on October 5, 2001, which revealed “[bjypertrophic AC joint with inferior impingement against the musculo-tendinous junction of the supraspinatus” and “[bjicipital tendinitis.” (Id. at 256.)
Ellis has had two episodes of DVT, most recently being hospitalized on September 9, 2001. (Id. at 260-61.) Within 24 hours the leg swelling and pain reduced dramatically, and he was walking without pain or difficulty by September 12. Dr. Johnson continued Ellis on Coumadin to control the swelling and released Ellis on September 12. The previous episode occurred in August 1993 when Ellis was immobile following a pelvic fracture from an automobile accident. (Id. at 137-38.) A follow-up examination on August 24, 1993, revealed that there was no warmth or swelling in the lower left leg and that Ellis was doing well following a week in the hospital on Heprin therapy. (Id. at 139.) Examinations subsequent to the 1993 episode revealed no evidence of recurrent DVT. (Id. at 143,144.)
As to Ellis’s back pain, the record includes Ellis’s regular complaints of back pain and medication refills, and the results of an MRI of the lumbar spine performed on July 1, 2000. The MRI revealed “[m]ild to moderate degenerative facet disease throughout the lumbar spine,” and “[degenerative disc disease L2-3 but without disc herniations, spinal canal or gross neuroforaminal stenosis.” (Id. at 254.)
Dr. Richard Secor performed a state agency consultative examination of Ellis on January 18, 2001, including various range of motion tests. Dr. Secor noted that Ellis had full range of motion in both shoulders, in his right elbow, both wrists, both knees, and both hips. (Id. at 220-21.) He also had full range of motion upon lateral and back flexion of his cervical spine and lumbar spine. (Id.) The examination revealed only slight range of motion limitations in Ellis’s left wrist, extension of his cervical spine, rotation of his neck, and extension of his lumbar spine. (Id.) Dr. Secor recorded grip strength in both hands and both upper extremities at 5 out of 5. (Id. at 220.) Dr. Secor noted that Ellis’s recent MRI showed no acute herniation, and that Ellis reported chronic pain, which was treated with narcotics. (Id. at 215.) Dr. Secor also noted that Ellis walked with part-time assistance of a cane, but opined that the cane was not medically necessary. He noted that Ellis could sit, stand, and lie down without assistance. (Id. at 216.)
The ALJ determined that Ellis had medically determined impairments of *993DVT, non-insulin-dependant diabetes mel-litus, and fibromyalgia, which amounted to severe impairments. (Add. at 39.) The ALJ determined that Ellis’s hepatitis C was non-severe, based on the medical records indicating that Ellis’s liver enzymes were normal in April 2000 and that the hepatitis C was quiescent in September 2000. The ALJ further found that Ellis did not have a severe impairment in his back and legs, other than as it related to the fibromyalgia. He also found that Ellis’s shoulder pain did not present a severe impairment unrelated to the fibromyalgia that was expected to last at least twelve months.
The ALJ then determined that Ellis’s impairments did not meet a listed impairment and proceeded to determine Ellis’s residual functional capacity (RFC). The ALJ found Ellis’s subjective complaints of the severity of his symptoms and limitations to be not fully credible, based on a lack of objective medical evidence or a treatment history that would support his allegations. Giving Ellis some benefit of the doubt about his limitations, the ALJ determined that Ellis could lift and carry no more than ten pounds, could sit for up to six hours in an eight-hour workday, and could stand for up to two hours in an eight-hour workday, resulting in a finding that Ellis could perform the full range of sedentary work. Based on Ellis’s age, education, and ability to perform the full range of sedentary work, the ALJ found that Ellis was not disabled. See 20 C.F.R. pt. 404, subpt. P, app. 2, table 1, rule 201.27.
The Appeals Council denied Ellis’s request for review, and Ellis appealed the denial of benefits to the district court. The experienced magistrate judge performed a thorough review of the medical evidence and determined that the ALJ did not err in denying Ellis’s application for benefits. Ellis appeals.
II.
We review de novo the district court’s decision upholding the Commissioner’s denial of disability benefits. We will affirm if substantial evidence on the record as a whole-that is, evidence which a reasonable mind would find adequate to support the Commissioner’s conclusion-supports the Commissioner’s findings. Stormo v. Barnhart, 377 F.3d 801, 805 (8th Cir.2004). We consider the whole record, including evidence that detracts from as well as evidence that supports the Commissioner’s decision, and we will not reverse as long as substantial evidence supports the outcome. Id.
The ALJ undertook the familiar five-part analysis in determining whether Ellis was disabled,3 finding at the fifth step that he was not. The Commissioner bears the burden at the fifth step of establishing that, given the applicant’s residual functional capacity, age, education, and work experience, there are a significant number of jobs available in the national economy which the applicant can perform. 20 C.F.R. § 404.1560(c). Ellis raises three issues on appeal. He claims that the ALJ erred by: failing to fully develop the medical record before refusing to give controlling weight to his treating physician; failing to properly apply the Polaski4 factors to his subjective complaints of pain; and *994failing to seek vocational expert testimony. We address each issue in turn.
A. ALJ’s Obligation to Develop the Record and Defer to Ellis’s Treating Physician
The ALJ declined to give controlling weight to Dr. Johnson’s opinion because the opinion was not well supported by medically acceptable clinical and laboratory diagnostic techniques and no examinations revealed signs indicative of Dr. Johnson’s opinion such as muscle atrophy. (Add. at 41.) Ellis argues that rather than rejecting Dr. Johnson’s opinion as conclu-sory, the ALJ had a duty to contact Dr. Johnson for clarification of his opinion before discrediting it.
A social security hearing is a non-adversarial proceeding, and the ALJ has a duty to fully develop the record. See Stor-mo, 377 F.3d at 806. Although that duty may include re-contacting a treating physician for clarification of an opinion, that duty arises only if a crucial issue is undeveloped. Id. Ellis does not allege that the record is missing any relevant medical records. In fact, the ALJ held the record open for 30 days following the hearing to allow Ellis to supplement it with a more recent opinion from Dr. Johnson, which gave the ALJ the benefit of Dr. Johnson’s November 21, 2001, letter. At oral argument before this court, Ellis’s attorney noted that he supplemented the record and asserted that the record was sufficiently developed to support Dr. Johnson’s opinion that Ellis was disabled. Without informing the' court what additional medical evidence should be obtained from Dr. Johnson, Ellis has failed to establish that the ALJ’s alleged failure to fully develop the record resulted in prejudice, and has therefore provided no basis for remanding for additional evidence. See Shannon v. Chater, 54 F.3d 484, 488 (8th Cir.1995) (“[Rjeversal due to failure to develop the record is only warranted where such failure is unfair or prejudicial.”). The ALJ did not err in failing to re-contact Dr. Johnson.
The ALJ refused to give controlling weight to Dr. Johnson’s opinion that Ellis was disabled and limited in his ability to stand for no more than two hours a day or to sit for no more than four hours per day. In assessing Ellis’s RFC, the ALJ determined that Ellis could sit for a total of six hours and stand for a total of two hours, but was limited to sedentary work. This in itself is a significant limitation, which reveals that the ALJ did give some credit to Dr. Johnson’s medical opinions. It is only Dr. Johnson’s opinion about how long Ellis could sit or stand and his opinion that Ellis was incapable of working with which the ALJ disagreed.
Generally, an ALJ is obliged to give controlling weight to a treating physician’s medical opinions that are supported by the record. See Randolph v. Barnhart, 386 F.3d 835, 839 (8th Cir.2004); 20 C.F.R. § 404.1527(d)(2). A medical source opinion that an applicant is “disabled” or “unable to work,” however, involves an issue reserved for the Commissioner and therefore is not the type of “medical opinion” to which the Commissioner gives controlling weight. See Stormo, 377 F.3d at 806 (“[Tjreating physicians’ opinions are not medical opinions that should be credited when they simply state that a claimant can not be gainfully employed, because they are merely opinions on the application of the statute, a task assigned solely to the discretion of the Commissioner.” (internal marks omitted)); 20 C.F.R. § 404.1527(e)(1). Further, although medical source opinions are considered in assessing RFC, the final determination of RFC is left to the Commissioner. See 20 C.F.R. § 404.1527(e)(2). Thus, to the extent that the ALJ discredited Dr. John*995son’s conclusion that Ellis could not work, he rightly did so.
The Commissioner defers to a treating physician’s medical opinions about the nature and severity of an applicant’s impairments, including symptoms, diagnosis and prognosis, what an applicant is capable of doing despite the impairment, and the resulting restrictions. 20 C.F.R. 404.1527(a)(2). “A treating physician’s opinion is due ‘controlling weight’ if that opinion is ‘well-supported by medically acceptable clinical and laboratory diagnostic techniques and is not inconsistent with the other substantial evidence in the record.’ ” Hogan v. Apfel, 239 F.3d 958, 961 (8th Cir.2001) (quoting Prosch v. Apfel, 201 F.3d 1010, 1012-13 (2000)).
This arguably includes Dr. Johnson’s opinion that Ellis could only stand for a total of two hours and sit for a total of four hours in a work day based on Ellis’s “severe chronic back and hip pain.” (Admin. R. at 268.) Dr. Johnson also opined that Ellis could neither sit nor stand for more than an hour at a time due to his chronic pain and tendency toward blood clots in his legs from the deep vein thromboembolism. (Id. at 279.) The ALJ applied the same two-hour limit to Ellis’s ability to stand, but found that he could sit for up to six hours (rather than four) in an eight-hour period. As noted by the ALJ, there is no medical evidence suggesting the limitations contained in Dr. Johnson’s letter. Ellis admittedly has chronic back and hip pain, but the record reveals that his medications alleviate that pain. Ellis testified that he spent his time reading and watching television, which is contrary to his assertion that he could sit only for a limited time. Dr. Johnson never ordered or even suggested to Ellis that he limit the time that he stood or sat, nor did Ellis ever suggest to Dr. Johnson that he was unable to stand or sit for any length of time. See Hogan, 239 F.3d at 961 (discounting medical source statement opining that applicant could not sit, stand, or walk for.more than 20 minutes at a time or one hour total per day where no similar restrictions were included in her treatment records, and the consulting doctor concluded that the applicant could sit, stand, or walk up to six hours per day). Although Dr. Johnson listed Ellis’s propensity for blood clots in his legs as medical support for his finding that Ellis could not sit or stand for more than an hour at a time, Dr. Johnson never cautioned Ellis or limited Ellis’s prolonged sitting or standing to avoid the possibility of a blood clot. See Hensley v. Barnhart, 352 F.3d 353, 356 (8th Cir.2003) (affirming ALJ’s decision discrediting treating physician’s opinion that applicant had significantly limited mobility where “few if any functional limitations” were placed on the applicant by his other physicians); Hogan, 239 F.3d at 961 (discrediting treating physician’s opinion of limitations'where “[n]one of these restrictions appear elsewhere in [the treating physician’s] treatment records.”). Dr. Johnson noted that anti-coagulants kept Ellis’s thromboembolism in control. Further, Dr. Secor noted that Ellis was able to “walk briskly” without use of the cane that he carried, opining that the cane was medically unnecessary, and Ellis had no problems getting around the examining room. Because Dr. Johnson’s opinion that Ellis could only stand for two hours and sit for four hours, and do neither for more than one hour at a time, is not supported by any medical evidence in the record and is contrary to other evidence in the record, the ALJ properly discredited the opinion.
B. Application of Polaski to Ellis’s Subjective Claims of Pain
It. is the ALJ’s duty to determine an applicant’s RFC. Before doing so, the ALJ must determine the applicant’s credibility, as his subjective complaints *996play a role in assessing- his RFC. Pearsall v. Massanari, 274 F.3d 1211, 1217-18 (8th Cir.2001). Applying the factors discussed in Polaski, 739 F.2d at 1322, the ALJ found that Ellis was not fully credible about the severity of his symptoms and limitations based on a lack of objective medical evidence in the record, as well as Ellis’s limited treatment record. (Add. at 40.) The ALJ noted that Ellis’s medications alleviated his pain, there was no record of adverse side effects from the medication, and no doctor observed signs consistent with the limited lifestyle claimed by Ellis. (Id. at 40-41.) While the ALJ may not discount Ellis’s complaints solely because they are not fully supported by objective medical evidence, Ellis’s complaints may be discounted based on inconsistencies in the record as a whole. Lowe v. Apfel, 226 F.3d 969, 972 (8th Cir.2000).
“Where adequately explained and supported, credibility findings are for the ALJ to make.” Id. The ALJ adequately explained the inconsistencies upon which he relied to discount Ellis’s subjective complaints, and we therefore uphold that finding. The record reveals that Ellis has not worked since 1993- (Admin. R. at 102.) See Woolf v. Shalala, 3 F.3d 1210, 1214 (8th Cir.1993) (noting that a sporadic work history is relevant to the ALJ’s credibility analysis); Polaski, 739 F.2d at 1322 (noting work history as one factor to consider in credibility determination). Ellis claimed that back and hip pain were severe enough to prevent him from sitting or standing for any length of time, yet Dr. Secor’s examination in January 2001 revealed that Ellis had normal, or near normal, range of motion in his shoulders, hips, cervical spine, and lumbar spine. (Admin. R. at 220-21.) The record also revealed that medication alleviated his pain. (Id. at 123, 274). Finally, Dr. Johnson noted his concern that Ellis was becoming addicted to - the narcotic medications as early as 1995 (id. at 140, 145), providing a further basis for disbelieving the severity of Ellis’s complaints, see Anderson v. Barnhart, 344 F.3d 809, 815 (8th Cir.2003) (noting that record supported ALJ’s finding concerning applicant’s possible overuse of narcotic pain medication in discrediting applicant).
C. Requirement for Vocational Expert Testimony
Ellis argues that the ALJ erred in relying on the Medical-Vocational Guidelines (grids) to determine whether he was disabled because of his non-exertional impairments. The ALJ may not rely on the grids if Ellis suffers from non-exertional impairments, but instead must obtain the opinion of a vocational expert. See Shannon, 54 F.3d at 488. Non-exertional impairments that “do[] not diminish or significantly limit the claimant’s residual functional capacity to perform the full range of Guideline-listed activities” do not prevent ■ use of the grids, however. Id. Because the ALJ was within his discretion, based on the record, to discredit Ellis’s subjective complaints of pain and find that Ellis’s pain did not diminish his ability to perform the full range of sedentary work, the ALJ properly relied on the grids without calling for vocational expert testimony. Id.
Ellis also argues that his inability to sit or stand for more than an hour at a time significantly limits the number of sedentary jobs available to him, and as such, the ALJ should have called a vocational expert to assess this non-exertional limitation. Sedentary jobs primarily involve sitting, although the category also includes jobs with occasional standing or walking. 20 C.F.R. § 404.1567(a). A Social Security Ruling explains that the sitting requirement allows for normal breaks, including lunch, at two hour intervals. See SSR 96-Op, 1996 WL 374185, at *6 (Soc.Sec.Admin. July 2, 1996). The Ruling likewise notes *997that the full range of sedentary jobs requires an applicant to be able to walk or stand for approximately two hours out of an eight-hour day. The need to alternate between sitting and standing more frequently than every two hours could significantly erode the occupational base for a full range of unskilled sedentary work. Id. at *7. The Ruling notes that the RFC assessment should include the frequency with which an applicant needs to alternate between sitting and standing, and if the need exists, that vocational expert testimony may be more appropriate than the grids. Id.
If the ALJ had credited Dr. Johnson’s opinion that Ellis could sit for no more than one hour at a time, then we would agree with Ellis that the ALJ should have sought the opinion of a vocational expert. We have already determined, however, that the ALJ properly discredited Dr. Johnson’s opinion, and the ALJ’s RFC assessment properly excluded that limitation. Having found, as supported by the record, that Ellis could sit for up to six hours during an eight-hour period, with no apparent need to alternate that position more frequently than every two hours, the ALJ appropriately relied on the grids. See Patrick v. Barnhart, 323 F.3d 592, 596 (8th Cir.2003) (affirming ALJ’s use of grids where ALJ properly discredited claimant’s non-exertional complaints of fatigue).
III.
We appreciate the fact that Ellis has had an unfortunate and difficult history involving numerous automobile accidents that resulted in significant injuries. Given the record, the ALJ’s explanation of his actions, and the district court’s thorough review, we are bound to uphold the Commissioner’s decision to deny benefits. The judgment of the district court is affirmed.
. The Honorable Terry I. Adelman, United States Magistrate Judge for the Eastern District of Missouri, to whom the case was referred for decision by consent of the parties pursuant to 28 U.S.C. § 636(c) (2000).
. Ellis’s application alleged a disability onset date of 1978. The ALJ construed the allegation of a 1978 onset date as a request to reopen prior applications which had been denied. The district court noted that the ALJ properly applied res judicata and limited the current disability determination to the period following the most recently denied application, or the period since May 28, 1999. On appeal, Ellis does not allege error in so limiting the period covered by his current application.
. "The five part test is as follows: 1) whether the claimant is currently employed; 2) whether the claimant is severely impaired; 3) whether the impairment is, or is comparable to, a listed impairment; 4) whether the claimant can perform past relevant work; and if not, 5) whether the claimant can perform any other kind of work.” Cox v. Barnhart, 345 F.3d 606, 608 n. 1 (8th Cir.2003).
. Polaski v. Heckler, 739 F.2d 1320, 1322 (8th Cir.1984).