Richard Wurzinger, age fifty-eight, was sentenced to 262 months in prison, the longest sentence within the range indicated by the Sentencing Guidelines for his role at the center of a conspiracy to manufacture methamphetamine. Given his age and his failing health, he will likely die in prison. He appeals his sentence, arguing that the district court did not follow proper sentencing procedure and that his sentence is unreasonably long. We affirm.
Wurzinger first used methamphetamine in the summer of 2000, when he was fifty-two years old. Two years later he arranged to have Justin Rzentkowski teach him and his daughter-in-law, Colleen Wur-zinger, how to move on to manufacture the drug themselves. Their first attempt burned down Colleen’s barn, but Wurzinger and Colleen were soon “cooking” methamphetamine regularly. Over the next two years Wurzinger was at the center of a loose methamphetamine manufacturing conspiracy, whose members were primarily his family — his daughter-in-law Colleen, his wife Penny Wurzinger, his daughter Kole Lipski and her husband Jeff Lipski. The co-conspirators occupied various roles. Jeff Lipski distributed Wurzinger’s methamphetamine, both the Lipskis and Colleen helped Wurzinger cook the stuff at various times and almost everyone collected Sudafed and other over-the-counter drugs containing pseudoephedrine, a chemical used in the production of methamphetamine.
Police arrested Wurzinger on December 8, 2004. On July 13, 2005, he pleaded guilty to one count of conspiring to manufacture a controlled substance prohibited by 21 U.S.C. § 841(a)(1). He agreed with a modified presentence report that assigned him an adjusted offense level of thirty-five and a criminal history category of III, resulting in a Sentencing Guidelines-recommended range of 210 to 262 months in prison. Nonetheless, he argued that he was entitled to a below-guidelines sentence under 18 U.S.C. § 3553(a) for a number of reasons, among them his age, his diabetes and the difference between his sentence and those of his co-conspirators, some of whom had not been prosecuted and many of whom had received shorter sentences in state court. To put it mildly, the district court did not agree. It imposed a sentence of 262 months, the top of the guidelines range.
Wurzinger now appeals, arguing both that the district court improperly presumed that an appropriate sentence for him falls within the guidelines range and that his sentence is unreasonably long. The first claim stumbles coming out of the gate. Just prior to oral argument, this court decided that a presumption in favor of a guidelines sentence is appropriate and that a court need only consider a non-guidelines sentence when a defendant provides “cogent reasons” for one. United *651States v. Hankton, 463 F.3d 626, 629 (7th Cir.2006).
This leaves Wurzinger’s claim that his sentence is unreasonably long in light of the sentencing goals and factors enumerated in 18 U.S.C. § 3553(a). United States v. Booker, 543 U.S. 220, 261, 125 S.Ct. 738, 160 L.Ed.2d 621 (2005). He requires particularly compelling arguments to succeed, since a sentence within the range recommended by the Sentencing Guidelines carries a rebuttable presumption of reasonableness. United States v. Gonzalez, 462 F.3d 754, 756 (7th Cir.2006); United States v. Mykytiuk, 415 F.3d 606, 608 (7th Cir.2005). Wurzinger spends a good portion of his brief arguing that this established feature of circuit law improperly makes the guidelines mandatory, but he does not offer the “compelling reasons” necessary for us to revisit our own precedent. Goshtasby v. Bd. of Trs. of the Univ. of Ill., 141 F.3d 761, 766 (7th Cir.1998), overruled on other grounds by Kimel v. Fla. Bd. of Regents, 528 U.S. 62, 92, 120 S.Ct. 631, 145 L.Ed.2d 522 (2000). Other courts have disagreed with our view, believing that courts may come to treat the rebuttable presumption as effectively irre-buttable. See, e.g., United States v. Jimenez-Beltre, 440 F.3d 514, 518 (1st Cir.), petition for cert. filed, — U.S.L.W. — (U.S. Aug. 4, 2006) (No. 06-5727). This may be a valid concern but we believe that courts are familiar with the concept of rebuttable presumption and capable of applying it. Hankton, 463 F.3d at 630 n. 5; see also United States v. Zavala, 443 F.3d 1165, 1169 (9th Cir.) (distinguishing between “bursting bubble,” mandatory conclusive and mandatory rebuttable presumptions), reh’g en banc granted sub nom. United States v. Carty, 462 F.3d 1066 (9th Cir.2006). Wurzinger’s arguments must be strong enough to rebut the presumption that his sentence, at the top of the recommended guidelines range but still within it, is reasonable.
His strongest argument that his sentence is unreasonable is that his diabetes will kill him before he is free. Wurzinger was diagnosed with diabetes forty years ago; it is severe enough to require medication. He does not claim that the federal prison system will be unable to treat his condition,1 but he does argue that it leaves him with a life expectancy of “116.4 to 176.4 months,” well short of his sentence’s duration. His life expectancy statistics seem to be flawed,2 but the government *652does not dispute them, and he is already experiencing disturbing complications of his disease. Wurzinger has lost sensation in his hands and feet, an effect of diabetes in older people that like the similar loss of sensation in leprosy can lead to untreated infections and even amputations. Nat’l Inst, of Diabetes & Digestive & Kidney Diseases, Diabetic Neuropathies: The Nerve Damage of Diabetes 3 (2002), available at http:/diabetes.niddk.nih.gov/dm/ pubs/neuropathies/neuropathies.pdf. At sentencing, Wurzinger had an infected sore on his foot which he said was related to his diabetes. If science is the standard, it is unlikely that Wurzinger will live out his sentence.
There is a worthy tradition that death in prison is not to be ordered lightly, and the probability that a convict will not live out his sentence should certainly give pause to a sentencing court. United States v. Crickon, 240 F.3d 652, 656 (7th Cir.2001); United States v. Jackson, 835 F.2d 1195, 1200 (7th Cir.1988) (Posner, J., concurring); United States v. Watson, 385 F.Supp.2d 534, 538 (E.D.Pa.2005); United States v. Gigante, 989 F.Supp. 436, 441-43 (E.D.N.Y.1998); United States v. Baron, 914 F.Supp. 660, 662 (D.Mass.1995); United States v. Maltese, No. 90 CR 87-19, 1993 WL 222350, at *10 (N.D.Ill. June 22, 1993); United States v. Garrett, 712 F.Supp. 1327, 1335 (N.D.Ill.1989); United States v. Gomez, 742 F.Supp. 407, 411-12 (E.D.Mich.1989); see also Kathleen Dean Moore, Pardons: Justice, Mercy, and the Public Interest 166-67, 173-74 (1989) (describing the tradition of “deathbed pardons”). Wurzinger’s key argument is not the nonstarter that “age per se is a mitigating factor” (though he does attempt that argument as well), United States v. Bullion, 466 F.3d 574, 576 (7th Cir.2006), but that a sentence of death in prison is notably harsher than a sentence that stops even a short period before. Death is by universal consensus a uniquely traumatic experience, and prison often deprives defendants of the ability to be with their families or to otherwise control the circumstances of death. John A. Beck, Compassionate Release from New York State Prisons: Why Are So New Getting Out? 27 J.L. Med. & Ethics 216, 223-24 (1999); Jason S. Ornduff, Releasing the Elderly Inmate: A Solution to Prison Overcrowding, 4 Elder L.J. 173, 192 (1996); see also Sasha Abramsky, Prisoner’s Dilemma: Lifers, Legal Aff., Apr. 2004, at 40, 41, 43 (describing death at Angola Prison in Louisiana). A sentence that forces this experience on a prisoner is quantitatively more severe than a sentence that does not consume the entirety of a defendant’s life, inflicting greater punishment and creating a stronger deterrent effect. See, e.g., United States v. Patriarca, 948 F.2d 789, 793 (1st Cir.1991) (holding that an increase in penalty that ensured death in prison would likely deter bail jumping); Elizabeth Rapaport, Retribution and Redemption in the Operation of Executive Clemency, 74 Chi.-Kent L.Rev. 1501, 1521 (2000). Additionally, of course, the physical constraints of a dying illness will incapacitate some defendants as effectively as imprisonment, making such a long sentence unnecessary. Beck, supra, at 224.
But to the extent that Wurzinger seeks a sentence that will expire before he is likely to, he argues for a sentence below *653the recommended guidelines range, and on what was under the mandatory guidelines regime a discouraged ground. U.S.S.G. § 5H1.1; see also id. § 5H1.4; Moore, supra, at 174. A court may now impose such a sentence, but only with a “very good explanation” rooted in the circumstances of the case. United States v. Wallace, 458 F.3d 606, 608 (7th Cir.2006). Wurzinger must argue, in the face of the presumption that his sentence is reasonable, that it was unreasonable for the district court not to impose such an exceptional sentence, or at least a shorter sentence within the guidelines range that he would be more likely to survive.
As this court recently observed in affirming the afroue-guidelines, likely de fac-to life sentence of another fifty-eight year old insulin-dependent diabetic, this type of argument rarely succeeds. Bullion, 466 F.3d at 576-77. While the present case is a less compelling candidate for a severe sentence — though he shared Wurzinger’s age and disease, Bullion was additionally a serial armed robber, serial kidnapper and serial rapist sentenced for the attempted shotgun-murder of the person he somehow persuaded to many him, id. at 575-76— Wurzinger’s argument is another failure. The district court reasonably concluded that Wurzinger’s sentence was necessary to punish his crime of not merely manufacturing methamphetamine, but corrupting his extended family to manufacture most of the methamphetamine available in his region of Wisconsin. (See, e.g., Sentencing Tr. at 16 (“[I]f indeed he is in ill health, you can think of the thousands that he may have affected in his major drug dealing who are perhaps in that same position were it not for him being on the drug scene.”).) The court also reasonably thought a long sentence was necessary to keep Wurzinger from cooking more methamphetamine. Wurzinger argues that older offenders are generally less likely to commit crime, but as we recently observed what matters is whether the court reasonably concluded that Wurzinger in particular is a risk for further crimes. Bullion, 466 F.3d at 577. Here it was not unreasonable to think Wurzinger had an above-average tendency to crime in his old age; he was fifty-two when he embarked upon what is to all appearances his most serious offense.
While we say nothing about whether a lower sentence would have been equally reasonable, age and illness do not, in the face of the circumstances presented here, make Wurzinger’s sentence unreasonable. While some of the district court’s comments were a mite strange — most notably the claim that Wurzinger, at fifty-eight, was “a pretty young guy” (Sentencing Tr. at 15) — on the whole, it offered a reasonable explanation of why Wurzinger’s conduct justified a sentence at the top of the recommended guidelines range.
Wurzinger urges three additional grounds for a shorter sentence, none of which come as near as his age and illness to outweighing the need to punish and incapacitate him. The first is that it was unreasonable to sentence him to 262 months when some of his co-conspirators prosecuted in Wisconsin state court received sentences far shorter than what they would probably have received in federal court. He focuses on his daughter-in-law Colleen, sentenced to thirty-six months in state prison for conduct he argues was not significantly different from his own under the Sentencing Guidelines.
Courts should reduce “unwarranted sentence disparities among defendants with similar records who have been found guilty of similar conduct,” 18 U.S.C. § 3553(a)(6), but in most cases “disparities are at their *654ebb when the Guidelines are followed,” United States v. Boscarino, 437 F.3d 634, 638 (7th Cir.2006), petition for cert. filed, 74 U.S.L.W. 3629 (U.S. Apr. 27, 2006) (No. 05-1379). Reducing a federal prisoner’s sentence to accord with that of a similarly situated state convict may decrease one sentencing disparity but simultaneously enlarges another: that between the federal convict and all similarly situated federal convicts. Id. Because penalties vary from state to state, sentence reductions to approach state penalties similarly vary with the state in which the federal sentencing court sits, unjustifiably creating disparities among federal convicts. See United States v. Branson, 463 F.3d 1110, 1112 (10th Cir. 2006); United States v. Williams, 282 F.3d 679, 681-82 (9th Cir.2002). We cannot say that the court’s failure to narrow the gap between Wurzinger and his co-conspirators was unreasonable.
Next, Wurzinger argues that a shorter sentence was required because his criminal history has been largely caused by alcoholism and methamphetamine addiction. Substance dependency was a discouraged ground for departure, U.S.S.G. § 5H1.4, and mere addiction, without more, does not compel a below-guidelines sentence, United States v. Hankton, 463 F.3d 626, 630 (7th Cir.2006). Nor is a sentence at the top of the guidelines range unreasonable. The most persuasive case Wurzinger cites permits shorter sentences when a defendant makes an extraordinary effort to overcome drug addiction that might be sabotaged by a long prison stay. United States v. Maier, 975 F.2d 944, 946-48 (2d Cir.1992). But as far as the record reveals, Wur-zinger has made no efforts to defeat an addiction. At the time of sentencing he did not think he needed treatment for substance abuse and did not believe he was physically addicted to methamphetamine.
Finally, Wurzinger argues that his sentence was unreasonable because of the “gross disparity” between it and any of the prior sentences he received. The disparity is reasonable: Wurzinger has committed a crime more serious than any of those he has previously committed, such as burglarizing a grocery store for a few cases of beer or drunkenly firing a gun. In fact, though Wurzinger’s argument is made in terms of length of sentence rather than seriousness of prior crimes, it is arguably a disguised disagreement with the Sentencing Commission’s determination regarding the seriousness of his criminal history and the implications of that history for an appropriate sentence. United States v. Gonzalez, 462 F.3d 754, 755 (7th Cir.2006).
Wurzinger could still argue that a sentence shorter than the guidelines would recommend is required in his case, and he cites cases dealing with the guidelines policy statement authorizing downward departures when a criminal history calculation overstates the seriousness of a defendant’s past crimes. See U.S.S.G. § 4A1.3(b)(l). Wurzinger’s calculation does not overstate the seriousness of his history; his many crimes committed before the year 2003 earned him no criminal history points. His score was based entirely on a recent conviction for carrying a switchblade knife and his commission of the present crime while on probation for that offense. He presents no reason to believe that this truncated history overstates his tendency to serious criminality. See United States v. Carrasco, 313 F.3d 750, 757-58 (2d Cir.2002) (“[I]t would be unusual for a [criminal history calculation] based on one prior offense to overstate the seriousness of a criminal record.”).
The penalty imposed upon Wurzinger was harsh but his crime was very serious. We do not encourage de facto life sen*655tences but here the decision can be regarded as reasonable and we affirm it.