Hirezz.com

P a t h o p h y s i o l o g y / C o m p l i c a t i o n s
O R I G I N A L
Benfotiamine Prevents Macro- and
Microvascular Endothelial Dysfunction and
Oxidative Stress Following a Meal Rich in
Advanced Glycation End Products in
Individuals With Type 2 Diabetes
LIN STIRBAN, MD
KNUT KLEESIEK, MD
ONICA NEGREAN, MD
MICHAELA MUELLER-ROESEL, MD
Endothelial dysfunction is an early
ERND STRATMANN, PHD
THEODOR KOSCHINSKY, MD
HOMAS GAWLOWSKI, MS
JAIME URIBARRI, MD
INA HORSTMANN
HELEN VLASSARA, MD
HRISTIAN G ¨
OTTING, PHD
DIETHELM TSCHOEPE, MD
(3), obesity (4), coronary artery disease(5), congestive heart failure (6), and type1 (7) and type 2 (8) diabetes. Postprandial OBJECTIVE — Diabetes is characterized by marked postprandial endothelial dysfunction
induced by hyperglycemia, hypertriglyceridemia, advanced glycation end products (AGEs), and dicarbonyls (e.g., methylglyoxal [MG]). In vitro hyperglycemia-induced MG formation and endothelial dysfunction could be blocked by benfotiamine, but in vivo effects of benfotiamine on and occurs not only in patients with car- postprandial endothelial dysfunction and MG synthesis have not been investigated in humansuntil now.
diovascular disease (10) or diabetes (11)but even in healthy subjects (12). Distinc- RESEARCH DESIGN AND METHODS — Thirteen people with type 2 diabetes were
tive and cumulative (11) effects of hyper- given a heat-processed test meal with a high AGE content (HAGE; 15.100 AGE kU, 580 kcal, 54 g protein, 17 g lipids, and 48 g carbohydrates) before and after a 3-day therapy with benfotiamine (14) on postprandial endothelial dysfunc- (1,050 mg/day). Macrovascular flow-mediated dilatation (FMD) and microvascular reactive hyperemia, along with serum markers of endothelial disfunction (E-selectin, vascular cell adhe- prandial state covers most of our daytime, sion molecule-1, and intracellular adhesion molecule-1), oxidative stress, AGE, and MG were measured during both test meal days after an overnight fast and then at 2, 4, and 6 h postpran- play a decisive role in prevention of ath- RESULTS — The HAGE induced a maximum reactive hyperemia decrease of Ϫ60.0% after
2 h and a maximum FMD impairment of Ϫ35.1% after 4 h, without affecting endothelium- independent vasodilatation. The effects of HAGE on both FMD and reactive hyperemia were completely prevented by benfotiamine. Serum markers of endothelial dysfunction and oxidative dysfunction, including insulin, folic acid, stress, as well as AGE, increased after HAGE. These effects were significantly reduced by ben- and statins (11). These approaches aim atreducing postprandial oxidative stress CONCLUSIONS — Our study confirms micro- and macrovascular endothelial dysfunction
(vitamins C and E, statins, and partly folic accompanied by increased oxidative stress following a real-life, heat-processed, AGE-rich meal in individuals with type 2 diabetes and suggests benfotiamine as a potential treatment.
Diabetes Care 29:2064 –2071, 2006
(statins), or have a direct effect on endo-thelial nitric oxide (NO) production (folicacid, insulin, and tetrahydrobiopterin) ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● From the 1Heart and Diabetes Center NRW, Bad Oeynhausen, Ruhr-University, Bochum, Germany; the 2German Diabetes Center, Heinrich-Heine University, Duesseldorf, Germany; and the 3Division of Diabetes and Aging, Mount Sinai School of Medicine, New York, New York.
Address correspondence and reprint requests to Prof. Dr. Diethelm Tschoepe, Heart and Diabetes Center NRW, Georgstrasse 11, 32545 Bad Oeynhausen, Germany. E-mail: [email protected].
Received for publication 9 March 2006 and accepted in revised form 8 June 2006.
A.S. and M.N. contributed equally to this work.
Abbreviations: AGE, advanced glycation end product; CML, carboxymethyllysine; CRP, C-reactive pro-
group of moieties, one of the most repre- tein; FMD, flow-mediated dilatation; HAGE, high AGE content; HAGEϩBT, HAGE plus benfotiamine; ICAM, intracellular adhesion molecule; IL, interleukin; MG, methylglyoxal; TBARS, thiobarbituric acid reacting substance; TNF, tumor necrosis factor; VCAM, vascular cell adhesion molecule.
A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion nous AGEs, and the food’s AGE content is 2006 by the American Diabetes Association.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. DIABETES CARE, VOLUME 29, NUMBER 9, SEPTEMBER 2006 Stirban and Associates
and because of their known susceptibility the fasting state (7:00 A.M.) and 2, 4, and exert different pathological effects (21) in- for cardiovascular complications (30).
6 h following the test meal. Venous blood cluding binding with and activation of re- ipating in any major physical activity. Af- synthesis increases in parallel with hyper- ter the test meal, patients were allowed to glycemia in vivo (23). Postprandially, the rate of 50 ml/h until the last test (6 h) was 3/10, 4 patients had nonproliferative ret- stress. Part of these effects might be coun- At both visits, subjects received a cooked teracted by benfotiamine, a liposoluble vi- tamin B1 with much higher bioavailability A.M.). The meal consisted of 200 g chicken (n ϭ 9/4), aspirin (n ϭ 11), ACE inhibi- neuropathy (25), is a transketolase activa- tors (n ϭ 9), angiotensin receptor block- tor that directs glucose substrates to the ers (n ϭ 1), hydroxymethylglutaryl-CoA inhibitors (n ϭ 6), ␤-blockers (n ϭ 5), provided 54 g protein, 17 g fat, and 48 g diuretics (n ϭ 5), and calcium channel AGEs and dicarbonyls formation (26).
Ն13%, pregnancy, heart failure New (15.100 kU AGE/meal). The AGE content York Heart Association III-IV, history of was calculated according to recently pub- stroke, peripheral arterial vascular disease benfotiamine on in vivo endothelial func- stadium IIB or higher, renal failure (serum (resting blood pressure Ͻ90/50 or Ͼ180/ three antihypertensives, therapy with ni- trates, and severe diabetes complications thelial function was assessed at the mac- (proliferative diabetic retinopathy, mac- of arterial diameter were performed with a derivatives, and diabetic foot syndrome).
written informed consent. The local ethics endothelial function (vascular cell adhe- was carried out according to the principles outlined in the Declaration of Helsinki.
rested for at least 10 min before the first crosis factor [TNF]-␣, C-reactive protein brinogen), oxidative stress (thiobarbituric investigation, but they were kept constant avoid movement artifacts, the subject’s standard diabetes diet for the 9-day study period. On day 4 (n ϭ 6) or 6 (n ϭ 7), we RESEARCH DESIGN AND
assessed the acute effects of a cooked test METHODS — Thirteen adults with
endothelial function and oxidative stress.
without a history of acute cardiovascular orally on days 7, 8 (3 ϫ 350 mg/day), and 9 (1.050 mg, 1 h before the intake of the after the cuff release, followed by contin- cause they represent the majority of indi- terial diameter for 120 s after deflation. At DIABETES CARE, VOLUME 29, NUMBER 9, SEPTEMBER 2006 Benfotiamine prevents endothelial dysfunction
20 min at 4°C. Aliquots of 750 ␮l were and 5 min later, the last data acquisition cose and triglycerides. The level of signif- icance was set at 0.05, and all tests were was defined as the percent change in arterial compared with the baseline diameter.
was the percent increase in arterial diam- eter 5 min following glycerotrinitrate.
to 4.40 Ϯ 0.67 (2 h), 4.15 Ϯ 0.53 (4 h), and 4.89 Ϯ 0.79% (6 h) (P Ͻ 0.01 vs.
skilled investigator (M.N.) blinded to the sequence of investigation. For the reactive were taken 60 s after cuff deflation (max- were 6.21 Ϯ 0.85 (baseline), 5.73 Ϯ 0.71 imal arterial diameter following reactive cals, Gru¨nberg, Switzerland), fibrinogen, 0.85% (6 h) (P Ͻ 0.01 vs. HAGE).
alyzed at the end of the diastole, and ar- u m - i n d e p e n d e n t v a s o d i l a t a t i o n .
0.14 (2 h), 4.11 Ϯ 0.15 (4 h), and 4.29 Ϯ 0.15 (6 h) (P Ͻ 0.01 vs. baseline) and (baseline) to 4.04 Ϯ 0.14 (2 h) (P Ͻ 0.05 m o d i fi e d A r g / m o l B S A , b y h i g h - 4.17 Ϯ 0.14 (6 h) (P Ͻ 0.01 vs. baseline).
There were no differences in postprandial The skin microcirculation was assessed si- changes in maximal arterial diameter fol- LEA Medizintechnik, Giessen, Germany).
the thenar surface of the right hand. The the study during both occasions; values at tissue, where it is scattered and collected and is expressed in arbitrary units. Given the great inter- and intra-assay variability of absolute blood flow values, a reproduc- Student’s t test was used to compare the (baseline) to 1.15 Ϯ 0.09 (2 h), 2.00 Ϯ the increase in blood flow following a 4.5- 0.41 (4 h), and 1.60 Ϯ 0.27 (6 h) (P Ͻ 0.05 vs. baseline). The effect of HAGE was treatment; in this case (HAGEϩBT), reac- ANOVA. If differences reached statistical significance, a two-tailed paired t test was (baseline), 1.93 Ϯ 0.28 (2 h), 2.71 Ϯ 0.81 (4 h), and 2.36 Ϯ 0.44 (6 h) (P Ͻ 0.05 vs.
time periods, with Bonferroni’s correction effects of benfotiamine, changes in vari- Baseline systolic and diastolic blood pres- sure, as well as heart rate, was comparable obtained after centrifugation at 1,500g for models were fitted to assess the relation- DIABETES CARE, VOLUME 29, NUMBER 9, SEPTEMBER 2006 Stirban and Associates
Figure 1—A: Change in FMD following HAGE (E) and HAGEϩBT (F). *P Ͻ 0.05 vs. baseline; †P Ͻ 0.01 and ‡P Ͻ 0.001 vs. HAGEϩBT. B:Change in reactive hyperemia following HAGE (E) and HAGEϩBT (F). *P Ͻ 0.05 vs. baseline; ‡P Ͻ 0.05 vs. HAGEϩBT. C: Change in E-selectinfollowing HAGE (E) and HAGEϩBT (F). *P Ͻ 0.05 vs. baseline; ‡P Ͻ 0.05 vs. HAGEϩBT. D: Change in TBARS following HAGE (E) andHAGEϩBT (F). *P Ͻ 0.05 vs. baseline; ‡P Ͻ 0.05 vs. HAGEϩBT. line, 0.259 Ϯ 0.101 mg/l at 6h; P ϭ NS vs.
All circulating markers of endothelial dys- changes in fibrinogen, TNF-␣, IL-6, and and 53.3 Ϯ 5.2* ng/ml (4 h); ICAM-1,213.9 Ϯ 11.3 (baseline), 231.8 Ϯ 11.5 (2 h), 228.5 Ϯ 12.5* (4 h), and 221.6 Ϯ 9.9 Serum glucose values at baseline and at 2, (baseline), 741.3 Ϯ 95.9* (2 h), 650.1 Ϯ h) (*P Ͻ 0.05 vs. baseline).
line), 8.75 Ϯ 1.49 (2 h), 13.28 Ϯ 1.83* (4 h), and 7.09 Ϯ 1.44 units/ml (6 h) (*P Ͻ 14*, 117 Ϯ 9, and 106 Ϯ 6* mg/dl (*P Ͻ ng/ml (4 h); ICAM-1, 221.3 Ϯ 15.4 (base- (baseline), 12.35 Ϯ 1.93 (2 h), 11.57 Ϯ line), 220.0 Ϯ 11.7 (2 h), 210.2 Ϯ 10.4* 24, 162 Ϯ 26, 176 Ϯ 28*, and 172 Ϯ 25* 1.90 (4 h), and 8.05 Ϯ 2.16 units/ml (6 h) m g / d l , r e s p e c t i v e l y , a n d d u r i n g (P ϭ NS vs. baseline and vs. HAGE for 611.4 Ϯ 65.0* (2 h), 614.6 Ϯ 52.8 (4 h), all). Similarly, MG increased 4 h following 157 Ϯ 17*, and 151 Ϯ 19* mg/dl (*P Ͻ and 650.1 Ϯ 76.3 ng/ml (6 h) (*P Ͻ 0.05 0.05 vs. baseline). No significant differ- line) to 3.04 Ϯ 0.31 (2 h), 4.16 Ϯ 0.42* ences with respect to the baseline and the postprandial blood glucose and triglycer- (*P Ͻ 0.05 vs. baseline) but not after 3.61 Ϯ 0.38 (2 h), 3.51 Ϯ 0.54 (4 h), and 0.273 Ϯ 0.092 mg/dl at 6 h (P ϭ 0.043), 2.68 Ϯ 0.44 nmol/ml (6 h) (P ϭ NS vs.
DIABETES CARE, VOLUME 29, NUMBER 9, SEPTEMBER 2006 Benfotiamine prevents endothelial dysfunction
nantly a measure of NO bioavailability (if glycerotrinitrate is assessed in parallel), 10.53 Ϯ 0.97* (2 h), 7.57 Ϯ 0.61 (4 h), while the regulation of microcirculation is and 7.02 Ϯ 0.49 nmol/ml (6 h) (*P Ͻ 0.05 vs. baseline). The effect was reduced of vasodilatatory prostaglandins has been ally, an increase in oxidative stress prob- 8.62 Ϯ 0.66 (2 h) (P Ͻ 0.05 vs. baseline; ably further promoted NO scavenging.
P Ͻ 0.05 vs. HAGE), 6.37 Ϯ 0.46 (4 h), and 6.12 Ϯ 0.31 nmol/ml (6 h) . All fast- tion can be the result of a combined effect culation, as shown by our finding of max- smaller following HAGEϩBT only at 2 h.
Similar postprandial changes in the arte- HAGE, baseline to 4 h: r ϭ Ϫ0.782, P ϭ HAGE, baseline to 2 h: r ϭ Ϫ0.751, P ϭ vasodilatation (measured as arterial diam- Ϫ0.708, P ϭ 0.010), MG and CML (dur- clearly showed a parallel increase in changed the postprandial response, it diding HAGE, baseline to 2 h: r ϭ 0.726, P ϭ 0.008). A borderline significant, inverse sured in the fasting state: FMD, reactive h y p e r e m i a , E - s e l e c t i n , V C A M - 1 , prandial vasodilatation could be the insu- r ϭ Ϫ0.532, P ϭ 0.075) was found. We also found a positive correlation between baseline values of MG and TBARS (r ϭ production, leading to vasodilation at the duce baseline changes, but an alternative 0.701, P ϭ 0.011). No further correla- macrovascular level (36), and this can be tions existed between either absolute val- only partially counteracted by the super- ues or changes in different parameters.
side “outside” the endothelial cells. Ben- CONCLUSIONS — The main find-
(decrease of TBARS after 2 h), can amelio- rate insulin resistance. Consequently, af- ter benfotiamine pretreatment, a decrease of insulin secretion and of the subsequent AGE-rich test meal in patients with type 2 directly improving endothelial function.
as well as reduction of oxidative stress as sure of NO were not assessed; instead, the curred already after 2 h, reached a maxi- mined. They directly reflect the endothe- correlate with plasma nitrite and nitrate levels, at least under controlled dietary in- FMD, it also showed some differences: re- a decrease in endothelial NO synthesis, an increase in NO scavenging (e.g., by AGEs, vented postprandial increases in circulat- dicarbonyls, reactive oxygen species), or a reach the baseline value after 6 h (similar cle cells. We can exclude the latter mech- reactive hyperemia impairment, and a full anism since the glycerotrinitrate-induced sider the possibility that not only MG but DIABETES CARE, VOLUME 29, NUMBER 9, SEPTEMBER 2006 Stirban and Associates
emia and is related to the lipoprotein(a) tion, effects that are prevented by benfo- level. J Clin Invest 93:50 –55, 1994 dilatation and essential hypertension(Letter). N Engl J Med 331:951, 1994 4. Tack CJ, Ong MK, Lutterman JA, Smits P: thelial function in obesity/insulin resis- tance: effects of troglitazone. Diabetologia A1C (8.5 Ϯ 0.5%) with relatively low lev- 5. Tousoulis D, Charakida M, Stefanadis C: els of fasting glycemia. This could be ex- reducing oxidative stress (it also slows the plained by the postprandial blood glucose coronary artery disease (Review). Heart might further reduce AGE and reactive ox- 6. Zelis R, Mason DT, Braunwald E: A com- ygen species– dependent endothelial cell parison of the effects of vasodilator stimuli on peripheral resistance vessels in normal jects (n ϭ 13) is certainly not large, it al- heart failure. J Clin Invest 47:960 –970, significance to our results. Moreover, this borderline significant inverse correlation DK, Betteridge DJ, Deanfield JE: Impaired pairment. This suggests direct effects of diabetes mellitus is related to disease du- ration and low density lipoprotein choles- regulation of the microcirculation and the terol levels. J Am Coll Cardiol 28:573–579, MG in the regulation of macrocirculation.
ation of other toxic compounds (49).
Therefore, the contribution of substances 8. Tooke JE, Goh KL: Vascular function in type 2 diabetes mellitus and pre-diabetes: the case for intrinsic endotheiopathy (Re- view). Diabet Med 16:710 –715, 1999 9. Ceriello A: The post-prandial state and cardiovascular disease: relevance to dia- function and that these effects can be re- betes mellitus (Review). Diabetes Metab ing a more pronounced oxidative stress.
10. Ling L, Zhao SP, Gao M, Zhou QC, Li YL, disease after a high-fat meal. Clin Cardiol function is reversed by simvastatin treat- cular dysfunction but raises some hypoth- 11. Ceriello A, Taboga C, Tonutti L, Quagli- bring light into these subtle mechanisms.
values after 6 h, an effect that was pre- vented by benfotiamine. Direct or indirect Acknowledgments — M.N. is supported by a
research fellowship offered by the German Ac- long-term simvastatin treatment. Circu- made possible by the excellent work of Marlen Ewald. The study medication was kindly pro- rameters characterized both study days.
12. Schinkovitz A, Dittrich P, Wascher TC: vided by Woerwag Pharma (Boeblingen, Ger- vessel reactivity and on indicators of oxi- crease in blood glucose after 2 h and de- dative stress in healthy volunteers. ClinPhysiol 21:404 – 410, 2001 crease after 6 h (compared with baseline).
References
13. Akbari CM, Saouaf R, Barnhill DF, New- Robinson J, McCredie R, Donald A, Dean-field JE: Passive smoking and impaired blood glucose and triglyceride excursions tion in healthy young adults. N Engl J Med the different results cannot be attributed cemia. J Vasc Surg 28:687– 694, 1998 14. Lee IK, Kim HS, Bae JH: Endothelial dys- poulos D, Hatcher G, Betteridge DJ, Dean- function: its relationship with acute hy- DIABETES CARE, VOLUME 29, NUMBER 9, SEPTEMBER 2006 Benfotiamine prevents endothelial dysfunction
perglycaemia and hyperlipidemia. Int 26. Hammes HP, Du X, Edelstein D, Taguchi J Clin Pract Suppl 59 – 64, 2002 15. Mangoni AA, Sherwood RA, Asonganyi B, terolemic patients. J Clin Endocrinol Metab 38. Deanfield J, Donald A, Ferri C, Giannat- with type 2 diabetes. Am J Hypertens 18: perimental diabetic retinopathy. Nat Med D, Rossi GP, Salvetti A, Schiffrin EL, Tad- 16. Lind L, Fugmann A, Branth S, Vessby B, 27. Pomero F, Molinar MA, La SM, Allione A, Millgard J, Berne C, Lithell H: The impair- dysfunction: part I: methodological issues similar to thiamine in correcting endothe- non-esterified fatty acids can be reversed lial cell defects induced by high glucose.
by insulin. Clin Sci (Lond) 99:169 –174, 28. Beltramo E, Berrone E, Buttiglieri S, Porta 17. Paolisso G, Tagliamonte MR, Barbieri M, the European Society of Hypertension.
and pericytes cultured in high glucose.
administration improves brachial reactiv- Diabetes Metab Res Rev 20:330 –336, 2004 ity and increases intracellular magnesium concentration in type II diabetic patients.
and aged human skin. J Physiol 563:965– J Clin Endocrinol Metab 85:109 –115, 2000 18. Vlassara H, Cai W, Crandall J, Goldberg tions. Diabetes Care 21:1414 –1431, 1998 30. Stamler J, Vaccaro O, Neaton JD, Went- worth D: Diabetes, other risk factors, and risk factor for diabetic angiopathy. Proc Natl Acad Sci U S A 99:15596 –15601, tervention Trial. Diabetes Care 16:434 – early haemodynamic changes. Eur J Clin 19. Vlassara H, Uribarri J: Glycoxidation and 41. Karachalias N, Babaei-Jadidi R, Kupich C, and a warning? (Review). Rev Endocr Boseska L, Ulbrecht J, Hinderliter AL: Bi- ological correlates of day-to-day variation 20. Goldberg T, Cai W, Peppa M, Dardaine V, streptozotocin-induced diabetic rats. Ann with type 2 diabetes: a study of test-retest reliability. Diabetologia 47:1625–1631, commonly consumed foods. J Am Diet As- 32. Celermajer DS, Sorensen KE, Gooch VM, Vlassara H: Oxidative stress-inducing car- 21. Koschinsky T, He CJ, Mitsuhashi T, Bu- Spiegelhalter DJ, Miller OI, Sullivan ID, novel mediators of cellular dysfunction.
Vlassara H: Orally absorbed reactive gly- tection of endothelial dysfunction in chil- dren and adults at risk of atherosclerosis.
43. Nagaraj RH, Oya-Ito T, Bhat M, Liu B: Dicarbonyl stress and apoptosis of vascu- nephropathy. Proc Natl Acad Sci U S A 94: lar cells: prevention by alphaB-crystallin.
Ann N Y Acad Sci 1043:158 –165, 2005 44. Wang X, Desai K, Chang T, Wu L: Vascu- clusion time. J Pharmacol Toxicol Methods velopment of hypertension. J Hypertens 34. Mitsuhashi T, Vlassara H, Founds HW, Li products (Review). J Mol Med 83:876 – 45. Schiekofer S, Andrassy M, Chen J, Rud- 23. Beisswenger PJ, Howell SK, O’Dell RM, J Immunol Methods 207:79 – 88, 1997 ␣-Dicarbonyls increase in the postpran- 35. Oya T, Hattori N, Mizuno Y, Miyata S, dial period and reflect the degree of hy- perglycemia. Diabetes Care 24:726 –732, glyoxal modification of protein: chemical factor ␬B in PBMCs. Diabetes 52:621– 24. Schreeb KH, Freudenthaler S, Vormfelde methylglyoxal-arginine adducts. J Biol parative bioavailability of two vitamin B1 46. Ceriello A, Assaloni R, Da RR, Maier A, Piconi L, Quagliaro L, Esposito K, Giugli- mononitrate (Letter). Eur J Clin Pharmacol skeletal muscle vasodilation is nitric oxide ano D: Effect of atorvastatin and irbesar- increase nitric oxide release. J Clin Invest postprandial endothelial dysfunction, ox- idative stress, and inflammation in type 2 37. Desideri G, Marinucci MC, Tomassoni G, diabetic patients. Circulation 111:2518 – study). Int J Clin Pharmacol Ther 43:71– 47. Monnier L, Lapinski H, Colette C: Contri- DIABETES CARE, VOLUME 29, NUMBER 9, SEPTEMBER 2006 Stirban and Associates
48. Klopotek Y, Otto K, Bohm V: Processing strawberries to different products alters 49. Elvevoll EO, Osterud B: Impact of pro- patients: variations with increasing levels contents of vitamin C, total phenolics, to- of HbA . Diabetes Care 26:881– 885, food items (Review). Forum Nutr 56:337– ity. J Agric Food Chem 53:5640 –5646, DIABETES CARE, VOLUME 29, NUMBER 9, SEPTEMBER 2006

Source: http://hirezz.com/satvenandmer.com/pdf/N4N/Benfotiamine%20Prevents%20Macro-%20and%20Microvascular%20Endothelial%20Dysfunction.pdf