Effects of Flavanol-Rich Dark Chocolate on Visual Function and Retinal Perfusion Measured With Optical Coherence Tomography Angiography: A Randomized Clinical Trial | Ophthalmic Imaging | JAMA Ophthalmology | JAMA Network
[Skip to Navigation]
Figure 1.  Flow of Participants Through the Study
Flow of Participants Through the Study
Figure 2.  Comparison of Effects of Dark Over Milk Chocolate on Primary and Secondary Outcome Measures
Comparison of Effects of Dark Over Milk Chocolate on Primary and Secondary Outcome Measures

No treatment effect of dark chocolate (linear mixed model) could be observed on the secondary end points visual acuity as measured with Early Treatment Diabetic Retinopathy Study (ETDRS) letters (P = .09), Pelli-Robson chart contrast sensitivity (P = .21), and Mars chart contrast sensitivity (P = .83, data not shown). No treatment effect of dark chocolate could be observed on the primary end point retinal perfusion measured as vessel density on optical coherence tomography angiography in the central 3 mm of the posterior pole centered on the fovea (superficial plexus: P = .56, deep plexus: P = .42). Confounding effects of blood pressure on vessel density were ruled out as tertiary outcome measure (P = .21, data not shown). LogCS indicates logarithm of contrast sensitivity.

Table 1.  Baseline Characteristics of Both Groups
Baseline Characteristics of Both Groups
Table 2.  Efficacy End Points
Efficacy End Points
1.
Ried  K, Fakler  P, Stocks  NP.  Effect of cocoa on blood pressure.  Cochrane Database Syst Rev. 2017;4:CD008893.PubMedGoogle Scholar
2.
Decroix  L, Tonoli  C, Soares  DD, Tagougui  S, Heyman  E, Meeusen  R.  Acute cocoa flavanol improves cerebral oxygenation without enhancing executive function at rest or after exercise.  Appl Physiol Nutr Metab. 2016;41(12):1225-1232. doi:10.1139/apnm-2016-0245PubMedGoogle ScholarCrossref
3.
Sorond  FA, Lipsitz  LA, Hollenberg  NK, Fisher  NDL.  Cerebral blood flow response to flavanol-rich cocoa in healthy elderly humans.  Neuropsychiatr Dis Treat. 2008;4(2):433-440.PubMedGoogle Scholar
4.
Flammer  AJ, Hermann  F, Sudano  I,  et al.  Dark chocolate improves coronary vasomotion and reduces platelet reactivity.  Circulation. 2007;116(21):2376-2382. doi:10.1161/CIRCULATIONAHA.107.713867PubMedGoogle ScholarCrossref
5.
Heinrich  U, Neukam  K, Tronnier  H, Sies  H, Stahl  W.  Long-term ingestion of high flavanol cocoa provides photoprotection against UV-induced erythema and improves skin condition in women.  J Nutr. 2006;136(6):1565-1569. doi:10.1093/jn/136.6.1565PubMedGoogle ScholarCrossref
6.
Grassi  D, Necozione  S, Lippi  C,  et al.  Cocoa reduces blood pressure and insulin resistance and improves endothelium-dependent vasodilation in hypertensives.  Hypertension. 2005;46(2):398-405. doi:10.1161/01.HYP.0000174990.46027.70PubMedGoogle ScholarCrossref
7.
Lee  KW, Kim  YJ, Lee  HJ, Lee  CY.  Cocoa has more phenolic phytochemicals and a higher antioxidant capacity than teas and red wine.  J Agric Food Chem. 2003;51(25):7292-7295. doi:10.1021/jf0344385PubMedGoogle ScholarCrossref
8.
Fisher  ND, Hughes  M, Gerhard-Herman  M, Hollenberg  NK.  Flavanol-rich cocoa induces nitric-oxide-dependent vasodilation in healthy humans.  J Hypertens. 2003;21(12):2281-2286. doi:10.1097/00004872-200312000-00016PubMedGoogle ScholarCrossref
9.
Mao  TK, Powell  J, Van de Water  J,  et al.  The effect of cocoa procyanidins on the transcription and secretion of interleukin 1 beta in peripheral blood mononuclear cells.  Life Sci. 2000;66(15):1377-1386. doi:10.1016/S0024-3205(00)00449-5PubMedGoogle ScholarCrossref
10.
Gu  L, Kelm  MA, Hammerstone  JF,  et al.  Concentrations of proanthocyanidins in common foods and estimations of normal consumption.  J Nutr. 2004;134(3):613-617. doi:10.1093/jn/134.3.613PubMedGoogle ScholarCrossref
11.
Guo  XF, Ruan  Y, Li  ZH, Li  D.  Flavonoid subclasses and type 2 diabetes mellitus risk: a meta-analysis of prospective cohort studies.  Crit Rev Food Sci Nutr. 2018;1-13. doi:10.1080/10408398.2018.1476964PubMedGoogle Scholar
12.
Field  DT, Williams  CM, Butler  LT.  Consumption of cocoa flavanols results in an acute improvement in visual and cognitive functions.  Physiol Behav. 2011;103(3-4):255-260. doi:10.1016/j.physbeh.2011.02.013PubMedGoogle ScholarCrossref
13.
Rabin  JC, Karunathilake  N, Patrizi  K.  Effects of milk vs dark chocolate consumption on visual acuity and contrast sensitivity within 2 hours: a randomized clinical trial.  JAMA Ophthalmol. 2018;136(6):678-681. doi:10.1001/jamaophthalmol.2018.0978PubMedGoogle ScholarCrossref
14.
World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.  JAMA. 2013;310(20):2191-2194. doi:10.1001/jama.2013.281053.Google ScholarCrossref
15.
Manach  C, Donovan  JL.  Pharmacokinetics and metabolism of dietary flavonoids in humans.  Free Radic Res. 2004;38(8):771-785. doi:10.1080/10715760410001727858PubMedGoogle ScholarCrossref
16.
Ludovici  V, Barthelmes  J, Nägele  MP,  et al.  Cocoa, blood pressure, and vascular function.  Front Nutr. 2017;4:36. doi:10.3389/fnut.2017.00036PubMedGoogle ScholarCrossref
17.
Alnawaiseh  M, Lahme  L, Treder  M, Rosentreter  A, Eter  N.  Short-term effects of exercise on optic nerve and macular perfusion measured by optical coherence tomography angiography.  Retina. 2017;37(9):1642-1646. doi:10.1097/IAE.0000000000001419PubMedGoogle ScholarCrossref
18.
Yanik Odabaş  Ö, Demirel  S, Özmert  E, Batioğlu  F.  Repeatability of automated vessel density and superficial and deep foveal avascular zone area measurements using optical coherence tomography angiography: diurnal findings.  Retina. 2018;38(6):1238-1245. doi:10.1097/IAE.0000000000001671PubMedGoogle ScholarCrossref
19.
Wellek  S, Blettner  M.  On the proper use of the crossover design in clinical trials: part 18 of a series on evaluation of scientific publications.  Dtsch Arztebl Int. 2012;109(15):276-281.PubMedGoogle Scholar
20.
NHS. Reports that dark chocolate 'improves eyesight' are unconfirmed. https://www.nhs.uk/news/food-and-diet/reports-dark-chocolate-improves-eyesight-are-unconfirmed/. April 27, 2018.
21.
Elliott  DB, Sanderson  K, Conkey  A.  The reliability of the Pelli-Robson contrast sensitivity chart.  Ophthalmic Physiol Opt. 1990;10(1):21-24. doi:10.1111/j.1475-1313.1990.tb01100.xPubMedGoogle ScholarCrossref
22.
Huber  KK, Adams  H, Remky  A, Arend  KO.  Retrobulbar haemodynamics and contrast sensitivity improvements after CO2 breathing.  Acta Ophthalmol Scand. 2006;84(4):481-487. doi:10.1111/j.1600-0420.2006.00687.xPubMedGoogle ScholarCrossref
23.
Spaide  RF, Fujimoto  JG, Waheed  NK, Sadda  SR, Staurenghi  G.  Optical coherence tomography angiography.  Prog Retin Eye Res. 2018;64:1-55. doi:10.1016/j.preteyeres.2017.11.003PubMedGoogle ScholarCrossref
24.
Hwang  TS, Gao  SS, Liu  L,  et al.  Automated quantification of capillary nonperfusion using optical coherence tomography angiography in diabetic retinopathy.  JAMA Ophthalmol. 2016;134(4):367-373. doi:10.1001/jamaophthalmol.2015.5658PubMedGoogle ScholarCrossref
25.
Alnawaiseh  M, Lahme  L, Müller  V, Rosentreter  A, Eter  N.  Correlation of flow density, as measured using optical coherence tomography angiography, with structural and functional parameters in glaucoma patients.  Graefes Arch Clin Exp Ophthalmol. 2018;256(3):589-597. doi:10.1007/s00417-017-3865-9PubMedGoogle ScholarCrossref
26.
Conti  FF, Qin  VL, Rodrigues  EB,  et al Choriocapillaris and retinal vascular plexus density of diabetic eyes using split-spectrum amplitude decorrelation spectral-domain optical coherence tomography angiography.  BMJ Ophthalmol. 2018;103:452-456. doi:10.1136/bjophthalmol-2018-311903Google Scholar
27.
Lauermann  JL, Eter  N, Alten  F.  Optical coherence tomography angiography offers new insights into choriocapillaris perfusion.  Ophthalmologica. 2018;239(2-3):74-84. doi:10.1159/000485261PubMedGoogle ScholarCrossref
28.
Lamport  DJ, Pal  D, Moutsiana  C,  et al.  The effect of flavanol-rich cocoa on cerebral perfusion in healthy older adults during conscious resting state: a placebo controlled, crossover, acute trial.  Psychopharmacology (Berl). 2015;232(17):3227-3234. doi:10.1007/s00213-015-3972-4PubMedGoogle ScholarCrossref
29.
Morris  MC, Tangney  CC.  A potential design flaw of randomized trials of vitamin supplements.  JAMA. 2011;305(13):1348-1349. doi:10.1001/jama.2011.383PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Original Investigation
    September 26, 2019

    Effects of Flavanol-Rich Dark Chocolate on Visual Function and Retinal Perfusion Measured With Optical Coherence Tomography Angiography: A Randomized Clinical Trial

    Author Affiliations
    • 1Department of Ophthalmology, Ludwig-Maximilians-University, Munich, Germany
    JAMA Ophthalmol. 2019;137(12):1373-1379. doi:10.1001/jamaophthalmol.2019.3731
    Key Points

    Question  Does flavanol consumption in dark chocolate improve visual acuity, contrast sensitivity, or retinal perfusion on optical coherence tomography angiography as suggested previously?

    Findings  In a double-blind randomized clinical crossover trial, 22 participants were randomized to consume flavanol-rich dark and regular milk chocolate. Dark chocolate was not shown to increase visual acuity, contrast sensitivity, or retinal perfusion on optical coherence tomography angiography compared with milk chocolate ingestion.

    Meaning  In contrast to a previous trial reporting beneficial effects of dark chocolate flavanol on visual function, this similarly sized trial did not find any effects on subjective (visual function) or objective (retinal perfusion) end points.

    Abstract

    Importance  A recently reported randomized clinical trial suggested beneficial effects of vasodilating flavanols in dark chocolate on visual function without objective quantification of retinal perfusion.

    Objective  To assess the effects of dark chocolate flavanols on subjective visual function and retinal perfusion objectively quantified on optical coherence tomography (OCT) angiography.

    Design, Setting, and Participants  This randomized, masked double-blind crossover clinical trial analyzed 22 healthy participants at the Department of Ophthalmology, Ludwig-Maximilians-University Munich, Germany, in July 2018. Analysis was intention to treat. Analysis began in July 2018.

    Interventions  Participants were randomized to consume 20 g of dark chocolate containing 400 mg of flavanols or 7.5 g of milk chocolate. Two hours later, visual function and retinal perfusion on OCT angiography were evaluated. Systemic blood pressure was measured to rule out artifacts on OCT angiography.

    Main Outcomes and Measures  The primary end point was macular retinal perfusion quantified as vessel density on OCT angiography. The secondary end point was subjective visual function (Early Treatment Diabetic Retinopathy Study visual acuity, Pelli-Robson chart, and Mars chart contrast sensitivity).

    Results  All 22 participants (13 women [59.1%]; mean [SD] age, 27.3 [11.1] years) completed the trial. No relevant differences in baseline parameters between groups were identified. No change in the primary outcome measure, retinal perfusion, could be detected after consumption of dark vs milk chocolate (superficial plexus 48.0% vs 47.5%, treatment effect: −0.59 [95% CI, −2.68 to 1.50], P = .56; deep plexus 54.1% vs 54.0%, treatment effect: −1.14 [95% CI, −4.01 to 1.73], P = .42). No differences in changes in the secondary outcome parameters Early Treatment Diabetic Retinopathy Study visual acuity, Pelli-Robson chart, or Mars chart contrast sensitivity could be detected. Potentially confounding effects of changes in blood pressure were excluded.

    Conclusions and Relevance  In contrast to a previous similarly sized randomized clinical trial reporting beneficial effects on visual function, no short-term effects of flavanol-rich dark chocolate on automatically assessed retinal blood flow on OCT angiography or subjective visual function were observed in this study. As this small trial does not rule out the possibility of benefits, further trials with larger sample sizes would be needed to rule in or out possible long-term benefits confidently.

    Trial Registration  German Clinical Trials Register identifier: DRKS00015065

    Introduction

    The potential health benefits of cocoa have been investigated extensively in the past years.1-9 Of cocoa’s ingredients, the strongest beneficial effect is thought to be conveyed through flavonoids, mainly flavan-3-ol, mostly simply termed flavanol, which has antioxidative properties9 and can induce nitric oxide–dependent vasodilation.6,8

    Flavonoids naturally occur in cocoa beans, tea, wine, fruit, and vegetables in variable concentrations, of which cocoa products serve as the most important source of flavanol in a regular human diet.10 For this reason, the evaluation of flavanol-associated health benefits is usually carried out by means of cocoa-rich dark chocolate consumption. Several companies have meanwhile started to sell chocolate advertised as “rich in flavanols” owing to supposedly brand-specific careful processing.7

    The potential health benefits of flavanol span a wide range of organs and indications, eg, cardiovascular disease, in which flavanols have been reported to lower systemic blood pressure,1 optimize coronary perfusion, and limit thrombocyte aggregation4; type 2 diabetes mellitus, which has been reported to occur less in patients consuming significant amounts of flavonoids11; or brain diseases, in which flavonoids are supposed to improve cerebral perfusion.2,3

    Concerning visual function, 2 research teams,12,13 most recently Rabin et al13 in 2018, have reported that the consumption of flavanol-rich chocolate significantly improves visual contrast sensitivity owing to supposedly improved oxygenation of the retina. The study by Rabin et al13 has been met with an overwhelming media reception despite the small effect demonstrated and despite the absence of any objective parameters linking functional improvement to a retinal morphological correlate. Our study was therefore designed to further investigate the correlation of subjective flavanol effects on visual function and the objective purported explanation for this phenomenon, namely retinal perfusion measured with optical coherence tomography (OCT) angiography.

    Methods
    Trial Design

    The trial was designed as a randomized, controlled, double-blind, masked 2 × 2 crossover study. An overview of the participants’ flow through the study is given in Figure 1. After obtaining institutional review board approval (Ethics Committee of the Medical Faculty of the Ludwig-Maximilians-University Munich, Germany), the trial was registered in the German Clinical Trials Registration Board. The trial protocol and statistical analysis plan are available in Supplement 1. Patient recruitment and the 2 study visits were carried out at the Department of Ophthalmology, Ludwig-Maximilians-University Munich, Germany, in early July 2018. All patients were thoroughly informed about the study procedures, which were explained as an effort to investigate the effect of chocolate on visual function in general, without any specification of the difference studied between milk and dark chocolate. Written informed consent was obtained from all patients. All study procedures adhered to the tenets of the Declaration of Helsinki.14

    Participants

    Participants were eligible to participate in the trial if they had normal binocular function in the absence of ocular disease, were able to fully understand the study’s procedures and grant informed consent, and if their eyes allowed for sufficient imaging quality. Myopia was tolerated up to −8.0 D, hyperopia up to 5.0 D, and astigmatism up to −3.0 D.

    Intervention, Randomization, and Masking

    Each participant either consumed 20 g of flavanol-rich dark chocolate, equaling a total of 400-mg flavanol intake, or 7.5-g milk chocolate with a total flavanol amount of approximately 5 mg. To maintain the double-blind design, the wrapping around the chocolate was removed before it was given to the participant. All participants and study staff besides the 1 study assistant only performing randomization and handing over of the chocolate were masked during the study. No unmasking had to be performed.

    Randomization was performed by allocation of the participants by sequence of inclusion into a prespecified list. On the second visit 1 week later, the opposite chocolate was chosen. Participants were instructed to refrain from consuming caffeine or milk products 24 hours prior to testing. Morphologic and functional changes were assessed after 2 hours, as flavanol is known to have a short plasma half-life15 and as its cardiovascular effects have been shown to peak after 2 hours.16

    Methods

    Before and 2 hours after chocolate consumption, a binocular assessment of subjective visual function was performed. This included binocular visual acuity testing with an Early Treatment Diabetic Retinopathy Study (ETDRS) chart at 4 meters with individually prescribed corrective glasses or contact lenses, if needed. Moreover, contrast vision testing was performed at 1 m on a Pelli-Robson chart (Precision Vision) and at reading distance with a Mars letter chart (Mars Perceptrix).

    To include objective, participation-independent measures, OCT angiography cube scans of 3 × 3 mm and 6 × 6 mm were acquired before and after chocolate consumption on the right eye of each participant using the RTVue XR Avanti spectral-domain OCT (Optovue) at 840 nm with an A-scan rate of 70 000 scans per second, using the split-spectrum amplitude-decorrelation algorithm to visualize blood flow. To exclude confounding effects by changes in blood pressure, eg, due to physical exercise,17 the patient was instructed to rest on the examination chair in front of the OCT angiography device for 5 minutes prior to the examination, and blood pressure was measured before each imaging session to validate repeatability.

    The AngioVue software with AngioAnalytics (Optovue) was used for en face image analysis, offering project artifact removal, automatic segmentation of the superficial and deep retinal plexus, and automatic quantification of vessel density on the acquired en face image. The automatic software results were then checked for imaging quality and segmentation errors by 2 masked experienced readers (N.M. and J.S.). Results of vessel density are expressed as percentage obtained by dividing the total sum of flow covered areas in millimeters squared by the whole image area.

    Primary and Secondary Outcome Measures

    The prespecified primary outcome measure was the change in retinal perfusion attributable to dark chocolate consumption as measured on OCT angiography. Recent findings18 define a significant perfusion change on OCT angiography as more than 8% difference in vessel density in the superficial and more than 10% in the deep plexus. Based on a mean vessel density of approximately 50% in the superficial and 55% in the deep plexus, an absolute change of 4% in either plexus was set as cutoff for a significant difference. The resulting sample size for a power of 80% at a 2-sided α error of <.05 was calculated to be n = 18. Target enrollment was planned to be 20 participants.

    Secondary outcome measures included ETDRS visual acuity and Pelli-Robson and Mars charts contrast sensitivity. Arterial blood pressure was analyzed as tertiary outcome measure to ensure the absence of confounding effects on vessel density on OCT angiography.

    Statistical Analysis

    All data were gathered in Microsoft Excel spreadsheets (version 15.39 for Mac; Microsoft). Statistics and figures were done using SPSS version 25 (SPSS Inc). Baseline demographics were analyzed for normal distribution using the Kolmogorov-Smirnov test. An independent-samples t test and a Mann-Whitney U test were used to test for significant baseline differences between the randomized groups. Concerning the study’s end points, all patients were analyzed according to their randomization in a 2-sided fashion. First, data were analyzed using a crossover trial–specific t test approach taking carryover and period effects into consideration.19 To verify the results, a linear mixed model with fixed-effect terms considering multiplicity was applied. The level to indicate statistical significance was defined as 2-tailed P < .05.

    Results
    Participants

    Twenty-two healthy participants were recruited (Table 1). There were 13 women (59.1%) and 9 men (40.1%) with a mean (SD) age of 27.3 (11.1) years (range, 20-62 years). Apart from 4 participants (18.2%) who were medical students, no participant was a health care professional. Eleven participants (50%) were randomized to milk chocolate first and dark chocolate second; the other 11 (50%) were randomized to dark chocolate first and milk chocolate second. All 22 patients completed the study and were fully available for analysis. There were no adverse events.

    An overview of baseline parameters stratified by randomization group is given in Table 1, confirming the healthy ocular status of the participants involved. Both groups did not significantly differ at baseline concerning age (milk chocolate: 95% CI, 21.95-28.45 vs dark chocolate: 95% CI, 20.79-39.81; P = .29), sex (milk chocolate: 95% CI, 7.93-64.79 vs dark chocolate: 95% CI, 16.03-74.88; P = .34), baseline visual acuity using ETDRS (milk chocolate: 95% CI, 60.28-65.72 vs dark chocolate: 95% CI, 56.62-64.78; P = .35), or Pelli-Robson (milk chocolate: 95% CI, 1.80-1.94 vs dark chocolate: 95% CI, 1.74-1.92; P = .53) or Mars (milk chocolate: 95% CI, 1.84-1.90 vs dark chocolate: 95% CI, 1.73-1.81; P = .73) chart contrast sensitivity. On OCT angiography, no baseline differences concerning vessel density at 3 mm (milk chocolate: 95% CI, 48.25-52.74 vs dark chocolate: 95% CI, 48.14-51.46; P = .10) and 6 mm (milk chocolate: 95% CI, 45.19-48.61 vs dark chocolate: 95% CI, 47.54-50.26; P = .67) in the superficial or at 3 mm (milk chocolate: 95% CI, 51.97-56.23 vs dark chocolate: 95% CI, 54.51-56.29; P = .33) and 6 mm (milk chocolate: 95% CI, 52.15-58.06 vs dark chocolate: 95% CI, 52.31-59.29; P = .78) in the deep retinal plexus could be detected.

    Exclusion of Carryover and Period Effects

    Carryover effects were excluded for every outcome measure (ETDRS visual acuity: 95% CI, −8.8 to 14.5; P = .61; Pelli-Robson chart: 95% CI, −0.12 to 0.23; P = .51; Mars chart: 95% CI, −0.04 to 0.13; P = .28; vessel density 3 mm and 6 mm at the superficial plexus: 95% CI, −4.40 to 4.51; P = .98 and 95% CI, −1.58 to 12.90, P = .11, respectively; and vessel density 3 mm and 6 mm at the deep plexus: 95% CI, −1.26 to 9.46; P = .12 and 95% CI, −2.42 to 16.52, P = .13, respectively). Also, period effects were ruled out for every outcome measure (ETDRS visual acuity: 95% CI, −2.03 to 2.78; P = .75; Pelli-Robson chart: 95% CI, −0.15 to 0.03; P = .36; Mars chart: 95% CI, −0.25 to 0.17; P = .28; vessel density 3 mm and 6 mm at the superficial plexus: 95% CI, −2.90 to 1.28; P = .43 and 95% CI, −3.50 to 2.52; P = .74, respectively; and vessel density 3 mm and 6 mm at the deep plexus: 95% CI, −3.01 to 2.73; P = .92 and 95% CI, −1.78 to 6.77; P = .24, respectively).

    Primary Outcome Measures

    Compared with milk chocolate, there was no significant effect of dark chocolate on the primary end point of changes in retinal vessel density at the posterior pole (Table 2 and Figure 2). Imaged at 3 mm, vessel density was 48.0% vs 47.5% after dark and milk chocolate in the superficial plexus with a treatment effect of −0.59 (95% CI, −2.68 to 1.50; P = .56) and 54.1% vs 54.0% in the deep plexus with a treatment effect of −1.14 (95% CI, −4.01 to 1.73; P = .39). Imaged at 6 mm, vessel density was 51.0% vs 51.0% after dark and milk chocolate in the superficial plexus with a treatment effect of 2.20 (95% CI, −0.82 to 5.21; P = .15) and 54.6% vs 55.8% in the deep plexus with a treatment effect of −0.57 (95% CI, −4.84 to 3.70; P = .95).

    Secondary Outcome Measures

    Compared with milk chocolate, there were no differences noted regarding the effect of dark vs milk chocolate on changes in secondary end points of subjective visual function, comprised of ETDRS visual acuity or Pelli-Robson or Mars chart contrast sensitivity (Table 2 and Figure 2). ETDRS visual acuity (approximate Snellen equivalent) letter score was 61.2 (20/63) vs 62.9 (20/63) letters after dark and milk chocolate with a treatment effect difference of 1.93 (95% CI, −0.47 to 4.33; P = .09). Pelli-Robson chart contrast sensitivity was 1.87 vs 1.89 logarithm of contrast sensitivity (logCS) after dark and milk chocolate with a treatment effect difference of 0.06 (95% CI, −0.03 to 0.15; P = .21), while Mars chart contrast sensitivity was 1.83 vs 1.83 logCS with a treatment effect difference of 0.03 (95% CI, −0.01 to 0.07; P = .83).

    Exclusion of Confounding by Changes in Blood Pressure

    Compared with milk chocolate, there was no effect of dark chocolate identified on systemic arterial blood pressure. Mean arterial pressure was 99.2 mm Hg vs 99.2 mm Hg after dark and milk chocolate with a treatment effect of 6.04 (95% CI, −5.65 to 17.74; P = .21). Moreover, there was no difference in mean arterial pressure directly before and after OCT angiography on both examination days (day 1 pre– vs post–OCT angiography: mean [SD], 99.2 [8.6] vs 101.2 [11.6] mm Hg; P = .50; day 2 pre– vs post–OCT angiography: mean [SD], 96.8 [10.0] vs 97.3 [7.8] mm Hg; P = .73).

    Discussion

    In 2018, a small randomized clinical trial by Rabin et al13 reported improvement in visual function 105 minutes after ingestion of a 72% cacao dark chocolate bar containing 316.3 mg of flavanols in 30 healthy participants. While the study did not show a difference in distance visual acuity or conventional large-letter contrast sensitivity, small-letter contrast sensitivity was reported to have improved by 0.15 logCS 2 hours after flavanol-rich chocolate ingestion.13 Although Rabin et al13 have acknowledged that the effect shown was most likely of little clinical relevance, their statement that dark chocolate improves vision was nonproportionally augmented by a media echo that urged the United Kingdom National Health Service to publish an article qualifying the study’s statement.20

    To challenge these findings, the present randomized clinical trial was created to thoroughly reproduce the design by Rabin et al13 mentioned above, testing the participants after the same time span of 2 hours with a comparable, even slightly higher dosing of flavanol (400 μm vs 316 μm). In addition to the subjective secondary end points concerning visual function, the assessment of retinal vessel density as measured on OCT angiography was introduced as the primary end point, as the effects of flavanol on visual function are purported to derive from flavanol-induced nitric oxide–dependent vasodilation.8

    Despite administering a great amount of flavanol, in neither respect investigated was our small crossover trial able to demonstrate any effects of flavanol-rich dark chocolate on ocular function or anatomy, be it the primary objective end point of retinal perfusion in the central 3 mm and 6 mm of the posterior pole centered on the macula, both in the superficial and deep retinal plexus, or the secondary end points concerning subjective visual function, including ETDRS distance visual acuity or contrast sensitivity measured by both the established Pelli-Robson and Mars charts. Although a single study, especially when the sample size is small, cannot exclude a possible treatment difference, this stands in harsh contrast to the study by Rabin et al.13

    Generally, most scientific reports on the benefits of chocolate consumption have certain limitations that should be considered when their results are being interpreted. First, most studies were conferred using a small sample size, and second were only able to demonstrate, if any, weak cocoa-related effects. For example, a Cochrane meta-analysis reviewing 20 studies including 856 participants found a significant mean blood pressure level reduction of 2 to 3 mm Hg attributable to flavanol ingestion,1 which does not justify any recommendation for the use of cocoa as a treatment for arterial hypertension. The same applies for the study by Rabin et al,13 which was only able to show a significant improvement in small-letter contrast sensitivity of 0.15 logCS after ingestion of dark chocolate; large-letter contrast sensitivity as well as distance visual acuity remained unchanged. Even if small-letter contrast sensitivity testing seems to be more sensitive than large-letter testing,13 repeatability of contrast sensitivity testing has been shown to be within 0.15 logCS, and it has been suggested that significant differences should be interpreted as a delta of 0.30 logCS and above.21 This makes the clinical significance of a delta of 0.15 logCS questionable.

    An important problem in designing prospective studies on this subject is the discernibility of chocolate type by taste, making it hardly possible to perform testing in a double-blind fashion. Studies only assessing subjective parameters, such as Rabin et al,13 might therefore be very susceptible to placebo artifacts. To reassess and objectify the effects demonstrated in the study by Rabin et al,13 retinal vessel density on OCT angiography was therefore chosen as objective primary end point that is expected to be less vulnerable to placebo effects. Furthermore, the problem of discernibility between chocolate by taste was approached by taking care to inform the participants that the effects of chocolate in general, and not dark flavanol-rich chocolate, were studied.

    As shown by Huber et al,22 improved retrobulbar hemodynamics induced by hypercapnia can improve contrast vision, presumably by carbon dioxide–driven vasodilation resulting in optimized oxygenation of the retina. Similarly, the potential health benefits of flavanol, including improvements in contrast sensitivity, should be detectable as vasodilation.8 Optical coherence tomography angiography is a new imaging modality that allows for the noninvasive detection of ocular blood flow and has been successfully established as new important diagnostic tool for a multitude of retinal diseases.23 In past years, automated quantification of vessel density has additionally entered the clinic.24 Using OCT angiography, the present study now for the first time, to our knowledge, offers in vivo imaging of human microvasculature under the influence of flavanol, quantified as vessel density. Vessel density has been established as most widely used quantifiable OCT angiography parameter and is calculated as percentage of vessel-covered area compared with the whole area imaged. Chronic pathological conditions such as glaucoma or diabetes mellitus have been shown to introduce persistent reproducible decreases in vessel density.25,26 Moreover, vessel density reacts quickly to stimuli to the cardiovascular system, for example to physical exercise, which leads to vasoconstriction detectable on OCT angiography.17 Therefore, this study’s rationale was that vasodilation secondary to flavanol intake should be detectable on OCT angiography. However, in accordance to the absence of significant subjective effects of dark chocolate on visual function, OCT angiography was not able to demonstrate any flavanol-derived effects on vessel density, while baseline intra-individual repeatability was excellent. However, as a caveat, our study lacks data on the quantification of choroidal blood flow, which has been proving difficult to measure reliably with today’s tools.27

    Limitations and Strengths

    Despite the negative results reported in this study, further research examining higher dosing, different times after ingestion, and moreover the possible effects of a flavanol loading phase for a couple of days are warranted. As our study also deviates from the standard procedure of testing visual acuity and contrast sensitivity monocularly, interpretation of the tests involved might be limited. However, as flavanol shows systemic action,1,3,4 a symmetrical involvement of both eyes can be expected.28

    Although comparable with the study by Rabin et al,13 the present study is mainly limited by its small sample size. Moreover, a normal, healthy human diet consists of a handful of foods that naturally contain flavanol.10 A potential confounder in studying flavanol and dietary supplements therefore consists in of the fact that healthy participants might already present with high baseline levels of the substance studied and thus not show an additional benefit by supplementation.29 Owing to the short plasma half-life of flavanol in the human body, this limitation might however be less pronounced.15 Besides, the effect of flavanol on patients with retinal disease cannot be estimated from our study. This is especially crucial for contrast sensitivity, which in both the study by Rabin et al13 as well as the present study showed such excellent baseline results that a ceiling effect might have diminished possible flavanol effects.

    As a strength, our study is the first to add OCT angiography as an objective parameter to the evaluation of possible flavanol effects on visual function, and the first study to offer imaging of human microvascular under the influence of flavanol.

    Conclusions

    In conclusion, no short-term effects of flavanol-rich dark chocolate on automatically assessed retinal blood flow on OCT angiography or subjective visual function could be observed in this small trial, in contrast to a previous similarly sized randomized clinical trial. As this small study does not rule out the possibility of benefits, further trials with larger sample sizes would be needed to rule in or out possible long-term benefits confidently.

    Back to top
    Article Information

    Corresponding Author: Jakob Siedlecki, MD, Department of Ophthalmology, Ludwig-Maximilians-University Munich, Mathildenstrasse 8, 80336 Munich, Germany (jakob.siedlecki@med.uni-muenchen.de).

    Accepted for Publication: July 14, 2019.

    Published Online: September 26, 2019. doi:10.1001/jamaophthalmol.2019.3731

    Author Contributions: Drs Siedlecki and Priglinger had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

    Concept and design: Siedlecki, Priglinger.

    Acquisition, analysis, or interpretation of data: All authors.

    Drafting of the manuscript: Siedlecki, Priglinger.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: Siedlecki, Mohr, Luft.

    Administrative, technical, or material support: Schworm.

    Supervision: Keidel, Priglinger.

    Conflict of Interest Disclosures: Dr Siedlecki received previous speaker fees and travel expenses from Novartis Pharma GmbH, Carl Zeiss Meditec AG, Oculentis OSD Medical GmbH, and Pharm-Allergan GmbH and received personal consultation fees from Bayer AG. Nikolaus Luft received income from honoraria as a lecturer from Alcon Laboratories, Nidek Co Ltd, and CenterVue SpA. Dr Schworm received previous speaker fees and travel expenses from Novartis Pharma GmbH and Topcon. Dr Priglinger received previous speaker fees and/or travel expenses from Novartis Pharma GmbH, Oertli Instrumente AG, Bayer AG, Alcon Pharma GmbH, and Pharm-Allergan GmbH. No other disclosures were reported.

    Data Sharing Statement: See Supplement 2.

    References
    1.
    Ried  K, Fakler  P, Stocks  NP.  Effect of cocoa on blood pressure.  Cochrane Database Syst Rev. 2017;4:CD008893.PubMedGoogle Scholar
    2.
    Decroix  L, Tonoli  C, Soares  DD, Tagougui  S, Heyman  E, Meeusen  R.  Acute cocoa flavanol improves cerebral oxygenation without enhancing executive function at rest or after exercise.  Appl Physiol Nutr Metab. 2016;41(12):1225-1232. doi:10.1139/apnm-2016-0245PubMedGoogle ScholarCrossref
    3.
    Sorond  FA, Lipsitz  LA, Hollenberg  NK, Fisher  NDL.  Cerebral blood flow response to flavanol-rich cocoa in healthy elderly humans.  Neuropsychiatr Dis Treat. 2008;4(2):433-440.PubMedGoogle Scholar
    4.
    Flammer  AJ, Hermann  F, Sudano  I,  et al.  Dark chocolate improves coronary vasomotion and reduces platelet reactivity.  Circulation. 2007;116(21):2376-2382. doi:10.1161/CIRCULATIONAHA.107.713867PubMedGoogle ScholarCrossref
    5.
    Heinrich  U, Neukam  K, Tronnier  H, Sies  H, Stahl  W.  Long-term ingestion of high flavanol cocoa provides photoprotection against UV-induced erythema and improves skin condition in women.  J Nutr. 2006;136(6):1565-1569. doi:10.1093/jn/136.6.1565PubMedGoogle ScholarCrossref
    6.
    Grassi  D, Necozione  S, Lippi  C,  et al.  Cocoa reduces blood pressure and insulin resistance and improves endothelium-dependent vasodilation in hypertensives.  Hypertension. 2005;46(2):398-405. doi:10.1161/01.HYP.0000174990.46027.70PubMedGoogle ScholarCrossref
    7.
    Lee  KW, Kim  YJ, Lee  HJ, Lee  CY.  Cocoa has more phenolic phytochemicals and a higher antioxidant capacity than teas and red wine.  J Agric Food Chem. 2003;51(25):7292-7295. doi:10.1021/jf0344385PubMedGoogle ScholarCrossref
    8.
    Fisher  ND, Hughes  M, Gerhard-Herman  M, Hollenberg  NK.  Flavanol-rich cocoa induces nitric-oxide-dependent vasodilation in healthy humans.  J Hypertens. 2003;21(12):2281-2286. doi:10.1097/00004872-200312000-00016PubMedGoogle ScholarCrossref
    9.
    Mao  TK, Powell  J, Van de Water  J,  et al.  The effect of cocoa procyanidins on the transcription and secretion of interleukin 1 beta in peripheral blood mononuclear cells.  Life Sci. 2000;66(15):1377-1386. doi:10.1016/S0024-3205(00)00449-5PubMedGoogle ScholarCrossref
    10.
    Gu  L, Kelm  MA, Hammerstone  JF,  et al.  Concentrations of proanthocyanidins in common foods and estimations of normal consumption.  J Nutr. 2004;134(3):613-617. doi:10.1093/jn/134.3.613PubMedGoogle ScholarCrossref
    11.
    Guo  XF, Ruan  Y, Li  ZH, Li  D.  Flavonoid subclasses and type 2 diabetes mellitus risk: a meta-analysis of prospective cohort studies.  Crit Rev Food Sci Nutr. 2018;1-13. doi:10.1080/10408398.2018.1476964PubMedGoogle Scholar
    12.
    Field  DT, Williams  CM, Butler  LT.  Consumption of cocoa flavanols results in an acute improvement in visual and cognitive functions.  Physiol Behav. 2011;103(3-4):255-260. doi:10.1016/j.physbeh.2011.02.013PubMedGoogle ScholarCrossref
    13.
    Rabin  JC, Karunathilake  N, Patrizi  K.  Effects of milk vs dark chocolate consumption on visual acuity and contrast sensitivity within 2 hours: a randomized clinical trial.  JAMA Ophthalmol. 2018;136(6):678-681. doi:10.1001/jamaophthalmol.2018.0978PubMedGoogle ScholarCrossref
    14.
    World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.  JAMA. 2013;310(20):2191-2194. doi:10.1001/jama.2013.281053.Google ScholarCrossref
    15.
    Manach  C, Donovan  JL.  Pharmacokinetics and metabolism of dietary flavonoids in humans.  Free Radic Res. 2004;38(8):771-785. doi:10.1080/10715760410001727858PubMedGoogle ScholarCrossref
    16.
    Ludovici  V, Barthelmes  J, Nägele  MP,  et al.  Cocoa, blood pressure, and vascular function.  Front Nutr. 2017;4:36. doi:10.3389/fnut.2017.00036PubMedGoogle ScholarCrossref
    17.
    Alnawaiseh  M, Lahme  L, Treder  M, Rosentreter  A, Eter  N.  Short-term effects of exercise on optic nerve and macular perfusion measured by optical coherence tomography angiography.  Retina. 2017;37(9):1642-1646. doi:10.1097/IAE.0000000000001419PubMedGoogle ScholarCrossref
    18.
    Yanik Odabaş  Ö, Demirel  S, Özmert  E, Batioğlu  F.  Repeatability of automated vessel density and superficial and deep foveal avascular zone area measurements using optical coherence tomography angiography: diurnal findings.  Retina. 2018;38(6):1238-1245. doi:10.1097/IAE.0000000000001671PubMedGoogle ScholarCrossref
    19.
    Wellek  S, Blettner  M.  On the proper use of the crossover design in clinical trials: part 18 of a series on evaluation of scientific publications.  Dtsch Arztebl Int. 2012;109(15):276-281.PubMedGoogle Scholar
    20.
    NHS. Reports that dark chocolate 'improves eyesight' are unconfirmed. https://www.nhs.uk/news/food-and-diet/reports-dark-chocolate-improves-eyesight-are-unconfirmed/. April 27, 2018.
    21.
    Elliott  DB, Sanderson  K, Conkey  A.  The reliability of the Pelli-Robson contrast sensitivity chart.  Ophthalmic Physiol Opt. 1990;10(1):21-24. doi:10.1111/j.1475-1313.1990.tb01100.xPubMedGoogle ScholarCrossref
    22.
    Huber  KK, Adams  H, Remky  A, Arend  KO.  Retrobulbar haemodynamics and contrast sensitivity improvements after CO2 breathing.  Acta Ophthalmol Scand. 2006;84(4):481-487. doi:10.1111/j.1600-0420.2006.00687.xPubMedGoogle ScholarCrossref
    23.
    Spaide  RF, Fujimoto  JG, Waheed  NK, Sadda  SR, Staurenghi  G.  Optical coherence tomography angiography.  Prog Retin Eye Res. 2018;64:1-55. doi:10.1016/j.preteyeres.2017.11.003PubMedGoogle ScholarCrossref
    24.
    Hwang  TS, Gao  SS, Liu  L,  et al.  Automated quantification of capillary nonperfusion using optical coherence tomography angiography in diabetic retinopathy.  JAMA Ophthalmol. 2016;134(4):367-373. doi:10.1001/jamaophthalmol.2015.5658PubMedGoogle ScholarCrossref
    25.
    Alnawaiseh  M, Lahme  L, Müller  V, Rosentreter  A, Eter  N.  Correlation of flow density, as measured using optical coherence tomography angiography, with structural and functional parameters in glaucoma patients.  Graefes Arch Clin Exp Ophthalmol. 2018;256(3):589-597. doi:10.1007/s00417-017-3865-9PubMedGoogle ScholarCrossref
    26.
    Conti  FF, Qin  VL, Rodrigues  EB,  et al Choriocapillaris and retinal vascular plexus density of diabetic eyes using split-spectrum amplitude decorrelation spectral-domain optical coherence tomography angiography.  BMJ Ophthalmol. 2018;103:452-456. doi:10.1136/bjophthalmol-2018-311903Google Scholar
    27.
    Lauermann  JL, Eter  N, Alten  F.  Optical coherence tomography angiography offers new insights into choriocapillaris perfusion.  Ophthalmologica. 2018;239(2-3):74-84. doi:10.1159/000485261PubMedGoogle ScholarCrossref
    28.
    Lamport  DJ, Pal  D, Moutsiana  C,  et al.  The effect of flavanol-rich cocoa on cerebral perfusion in healthy older adults during conscious resting state: a placebo controlled, crossover, acute trial.  Psychopharmacology (Berl). 2015;232(17):3227-3234. doi:10.1007/s00213-015-3972-4PubMedGoogle ScholarCrossref
    29.
    Morris  MC, Tangney  CC.  A potential design flaw of randomized trials of vitamin supplements.  JAMA. 2011;305(13):1348-1349. doi:10.1001/jama.2011.383PubMedGoogle ScholarCrossref
    ×