Long-term effects of blue-blocking spectacle lenses on color perception

Resea ch A icle Vol. 30, No. 11 / 23 May 2022 / Op ics Exp ess 19757 Long- e m e ec s o blue-blocking spec acle lenses on colo pe cep ion MARÍA SANTANDREU, EVA M. VALERO,*LUIS GÓMEZ-ROBLEDO, RAFAEL HUERTAS, MIGUEL-ÁNGEL MARTÍNEZ-DOMINGO,AND JAVIER HERNÁNDEZ-ANDRÉS Depa men o Op ics, Uni e si y o G anada, G anada 18071, Spain * ale ob@ug .es Abs ac : The use o blue-blocking il e s is inc easing in spec acle lens use s. Despi e he low abso p ion in he blue ange, some use s complain abou hese il e s because hey a ec hei colo pe cep ion. In a pilo s udy we ha e e alua ed how he long- e m use o 8 di e en blue-blocking il e s impac he colo pe cep ion du ing mo e han 2 weeks on a g oup o 18 no mal colo ision obse e s, compa ed wi h a con ol g oup o 10 obse e s. The e alua ion was done using he FM100, he Colo Assessmen and Diagnosis (CAD) and an ach oma ic poin measu emen . Ou esul s show ha he e is a end o wo sen wi h he il e s on. © 2022 Op ica Publishing G oup unde he e ms o he Op ica Open Access Publishing Ag eemen 1. In oduc ion The blue ligh emi ed by di e en de ices (LEDs, sc eens, e c.) has ecei ed a lo o media a en ion in ecen yea s. This has c ea ed a con o e sial ala m among he popula ion abou i s dange wi hou su icien scien i ic suppo o jus i y i . In 2018 wo Eu opean Commission scien i ic commi ees issued a join comp ehensi e epo ocusing on he assessmen o po en ial isks o human heal h associa ed wi h he use o LEDs, [1] concluding ha “ he e is no e idence o di ec ad e se heal h e ec s om LEDs emission in no mal use (lamps and displays) by he gene al heal hy popula ion”. They also ecommended ha u he esea ch is needed o asce ain he speci ic impac o LED ligh s on he ci cadian sys em because nowadays “i is no clea i his e ening dis u bance o he ci cadian sys em leads o long- e m ad e se heal h e ec s”. Despi e he absence o s ong scien i ic e idence highligh ing he isk o blue ligh , some comme cial companies decided, a ound en yea s ago, o manu ac u e il e s o block blue ligh , isualizing a comme cial niche. Fil e s o abso b blue ligh a e cu en ly ma ke ed as addi ional ea men in oph halmic lenses, con ac lenses o in aocula lenses, as well as in he o m o speci ic il e s as sc een p o ec o s o compu e s, mobile phones o able s. Elec onic de ices also in eg a e unc ions ha allow he use o change he in ensi y and colo empe a u e o he display by dimming he blue LEDs in he RGB displays. An example o his unc ion is he popula nigh mode p esen in mos cell phones, able s and compu e s. We ha e no ed he po en ially con using na u e o e minology o “blue-blocking” il e s. He e we do no deal wi h he s udy o il e s ha ac ually abso b almos all he blue con en , which ha e been analyzed in o he wo ks [2–4], bu wi h il e s ha , e en named as blue-blocking o comme cial pu poses, hei blue abso p ion is no ema kable. Today, i ually all manu ac u e s o oph halmic lenses o e hese op ions and a la ge majo i y o use s op o selec hem when buying hei glasses because hey ha e been wa ned abou he po en ial isks o blue ligh in some ma ke ing campaigns. Howe e , op ome is s some imes encoun e use s ha ejec his ype o il e s because hey a ec hei ch oma ic disc imina ion and colo pe cep ion. Consequen ly, his may a ec he pe o mance in hei p o essional ac i i ies pa icula ly when hese a e ela ed o colo as, o example, g aphic designe s, pho og aphe s, ligh ing echnicians, pain e s, a oois s, e c. #455209 h ps://doi.o g/10.1364/OE.455209 Jou nal © 2022 Recei ed 31 Jan 2022; e ised 26 Ma 2022; accep ed 27 Ma 2022; published 19 May 2022 Resea ch A icle Vol. 30, No. 11 / 23 May 2022 / Op ics Exp ess 19758 In 2017 Leung e al. [5] e alua ed, among o he isual unc ions, he colo disc imina ion o wo blue-ligh il e ing spec acle lenses using he Fa nswo h Munsell 100 hue es (FM100) on 80 pa icipan s wea ing he lenses o one mon h o a minimum o 2 hou s pe day. The colo disc imina ion was e alua ed a e ha mon h, bu au ho s only conside ed he o al e o sco e om he FM100. Pa icipan s also assessed subjec i ely hei pe o mance h ough a ques ionnai e. In he ecen s udy in 2021 by Baldasso e al. [6] hey e alua ed how h ee comme cial blue-blocking lenses il e s change colo disc imina ion on en young subjec s, jus a e pu ing on he glasses, wi h he ollowing h ee compu e -based colo ision es s: Colo Assessmen and Diagnosis (CAD), Camb idge Colo Tes , and FM100. They concluded ha no s a is ically signi ican e ec s on colo pe cep ion we e ound. In none o hese Re s. [5,6] he au ho s ackled he analysis o long- e m con inuous changes in colo pe cep ion due o he use o di e en comme cial “blue-blocking” lenses. How does ch oma ic adap a ion a ec he ch oma ic pe cep ion h ough “blue-blocking” il e s? How does he colo pe cep ion change wi h long- e m usage o hese lenses? Long- e m ch oma ic adap a ion has been expe imen ally induced by exposing obse e s o long pe iods o a i icially colo ed lenses, showing ha ch oma ic expe ience was al e ed [7,8]. In he wo k by Eisne and Enoch [7], h ee subjec s wo e a ed il e o e one eye o one week and hei ch oma ic pe cep ion was e alua ed using he unique-yellow expe imen and an anomaloscope. Thei esul s demons a ed ha “long- e m exposu e o a long-wa eleng h wo ld can induce ela i ely p olonged (a leas hou s) pos ecep o al adap a ion”. The au ho s s a ed ha he changes in sensi i i y could no be a ibu ed o p eneu al changes in he eye, as he anomaloscope se ings appea ed o emain unchanged. Nei z e al. [8] subjec ed ou obse e s o ch oma ically al e ed condi ions (colo ed con ac lenses, glasses wi h colo ed il e s, o a oom wi h colo ed ligh ing) o e se e al days. The ch oma ic al e a ion pe iod changed om 4 hou s o 12 hou s and he unique yellow was measu ed. Au ho s ound ha each subjec ’s unique yellow shi ed p og essi ely, being he shi size no cons an o e ime and a ied be ween he subjec s. The change in unique yellow induced by wea ing colo ed lenses was la ges du ing he i s ew days, and appea s o change mo e slowly a he end o a 2-week ch oma ic al e a ion pe iod. Nei z e al. claimed ha hei expe imen s o ch oma ic al e a ion demons a e ha “long- e m changes in ch oma ic expe ience induce a change in he weigh ing o he L and M cone inpu s o he ch oma ically opponen ed/g een channel”. T egillus e al. [9] claimed ha he mechanisms and iming ha con ol he compensa o y adjus men s inhe en o ch oma ic adap a ion “a e s ill poo ly unde s ood”. In hei pape hey e alua ed ch oma ic adap a ion by ha ing 12 young obse e s wea ing glasses, 8 hou pe day o 5 days du ing dayligh hou s, wi h yellow il e s ha app oxima ed he lens ansmi ance o an olde obse e (a ound 70 yea s old). The obse e ’s whi e poin was measu ed on a calib a ed moni o along he “blue–yellow” axis unde di e en s a es o adap a ion and wi h he lenses ei he on o o . A e a 5 minu es da k-adap ion, he whi e poin was de e mined a he beginning and end o each 8 hou s in e al, bo h wi h and wi hou he yellow lenses. Au ho s ound la ge indi idual di e ences in he magni ude o he colo se ings. Resul s showed a la ge shi owa ds yellow o all obse e s, ha occu ed in he i s 2 hou s and ha was educed by up o 50% by he end o he day. Au ho s ound ha , a e emo ing he glasses a he end o he day, he expec ed bluish a e e ec was no obse ed. They also concluded ha “signi ican compensa ion occu s wi hin each day bu did no ans e ac oss days o lead o a p olonged a e e ec a he end o he day”. In 2020 We ne e al. [10] e alua ed he e ec s o long- e m usage du ing wo weeks o a model o EnCh oma glasses (a b oad-band spec ally selec i e il e ) in eigh anomalous ich oma s and nine no mal ich oma s, on luminance and ch oma ic con as esponse. Resul s showed Resea ch A icle Vol. 30, No. 11 / 23 May 2022 / Op ics Exp ess 19759 ha hei ch oma ic pe cep ion changed as hey wo e he glasses o e se e al days. Mo eo e , his colo modi ica ion pe sis ed days a e he glasses we e emo ed, demons a ing an adap i e isual esponse. Webs e [11] s a es ha ision is a plas ic p ocess and “how he b ain in e p e s he ecep o signals o c ea e ou conscious imp essions o colo is s ill deeply mys e ious”. Ishe wood e al. [12] ecen ly claimed ha unde s anding compensa o y adjus men s can se e as an impo an insigh in o he gene al p inciples o adap a ion and plas ici y in he human isual sys em. The main objec i e o his pilo s udy is o in es iga e how he long- e m use o “blue-blocking” il e s could modi y he colo pe cep ion o no mal colo ision obse e s, and o desc ibe he empo al e olu ion o he changes. To do so we ha e in es iga ed 8 blue-blocking il e s om i e di e en b ands, whose ansmi ances ha e been measu ed, on a g oup o 18 no mal colo ision obse e s. A con ol g oup o 10 addi ional obse e s who did no use any il e was also included. To analyze he changes in colo pe cep ion h ee di e en es s we e used: he FM100 [13], he Colo Assessmen and Diagnosis (CAD) [14,15], and an ach oma ic poin measu emen simila o he one used by T egillus e al. [9]. The compa ison be ween hese h ee es s was no an objec i e o ou s udy, al hough he esul s ob ained allowed us o compa e which one was mo e sensi i e o desc ibe he po en ial changes in colo pe cep ion. This pape is o ganized as ollows: Sec ion 2desc ibes he il e s used, he obse e s and he es s employed; Sec ion 3p esen s and analyses he esul s ob ained; and in Sec ion 4we d aw he conclusions. 2. Me hods 2.1. Fil e spec al ansmi ances In his pilo wo k, ying o co e many o he di e en b ands a ailable in he ma ke , we ha e used 16 blue-blocking lenses, all nominally plano a ocal, om i e di e en b ands, inco po a ed o neu al lenses and moun ed on eigh con en ional ames (see Table 1). The il e ansmi ances, shown in Fig. 1, we e measu ed wi h a Spec o adiome e PR-745 (Spec ascan L d. USA) using a Sphe e Op ics calib a ed whi e e e ence a ge . Hal o he il e s a e made using ma ix-based echnology ( he lenses a e buil wi h ma e ials which p esen blue-blocking p ope ies), and he o he hal using su ace deposi s on lenses wi h no blue-blocking p ope ies. Fig. 1. Spec al ansmi ances o he eigh blue-blocking il e s. (a) in he 380-780 nm ange; (b) in he ange 380-500 nm ange. See Da a File 1, Re . [18] o unde lying alues. We ha e calcula ed he ch oma in CIELAB as he Euclidean dis ance be ween he il e ed e e ence whi e and he cen e o he colo space. These da a we e used o di ide he il e s in o Resea ch A icle Vol. 30, No. 11 / 23 May 2022 / Op ics Exp ess 19760 Table 1. Fil e s echnical speci ica ions: ac onym, b and, model, numbe o subjec s, ab ica ion echnology, a e age ansmi ance in he ange 380-1080 nm, a e age ansmi ance in he ange 380-500 nm, il e g oup(FG1 o FG2), and ch oma in CIELAB colo space space Ac onym B and Model Numbe o subjec s ha used his il e Fab ica ion echnology A e age o al ansmi ance (380-1080 nm) A e age ansmi ance (380-500 nm) FG1 o FG2 Ch oma √a2+b2 Zss Zeiss Du a ision, BlueP o ec UV +Du a ision, Pla inum UV 2 Su ace deposi s 83.66% 66.97% FG2 6.28 Ess1 Essilo C izal Sap- phi e +UV35 3 Su ace deposi s 81.87% 63.67% FG2 4.29 Ess2 Essilo C izal P e encia 35 1 Su ace deposi s 84.87% 67.08% FG1 8.30 HV Hoya Vision Hilux 1.6 Blue con ol 3 Su ace deposi s 83.20% 69.58% FG2 5.62 CSR1 CSR1 CSR 1A 2 Ma ix- based 83.56% 64.84% FG2 6.47 CSR2 CSR CSR 1A 2 Ma ix- based 80.15% 61.24% FG2 7.01 Re 1 Re ica e Re ica e London 2 Ma ix- based 85.55% 59.95% FG1 16.08 Re 2 Re ica e Re ica e London 3 Ma ix- based 83.24% 61.39% FG1 12.41 wo g oups o con enience o analysis: he i s g oup (FG1, o med by Re 1, Re 2 and Ess2, comp ising h ee il e s) has a shi abo e 8 CIELAB uni s, while he second g oup (FG2, o med by HV, Ess1, Zss and CSR1 and CSR2, comp ising i e il e s) ha e an ach oma ic poin shi below 8 CIELAB uni s (see Table 1). As expec ed, he yellowish in o he lenses in he i s g oup is much mo e no iceable han in he second g oup. Al hough hese il e s a e named as “blue-blocking” by manu ac u e s and p ac i ione s, i is clea om Fig. 1 ha he abso p ion o blue ligh is qui e mode a e. 2.2. Obse e s A o al o 28 obse e s ha e pa icipa ed in he expe imen s. All subjec s had no mal colo ision (as assessed by he Ishiha a and FM-100 es s) and did no wea glasses, ei he because hey we e emme opes o con ac lens use s. None o hem had used blue-blocking lenses p io o his s udy. Ten obse e s comp ised he con ol g oup, and eigh een obse e s he g oup ha pe o med he ch oma ic adap a ion ba e y o es s (CA g oup). The age ange o he con ol g oup was 21 o 37 yea s, wi h an a e age age o 23.5 yea s (s anda d de ia ion: 4.99 yea s). The subjec s in he CA g oup had ages anging om 19 o 49 yea s, wi h an a e age age o 27.16 yea s (s anda d de ia ion: 10.11 yea s). The e we e 16 emales, 7 in he con ol g oup, and 9 in he CA g oup. Each obse e in he CA g oup wo e one ype o blue-blocking il e s in hei co esponding ames o 15 days wi h a maximum o 3 h allowance o aking o he glasses while he subjec was awake, al hough hey we e encou aged no o ake he glasses o excep o showe ing o p ac icing spo s. Th ee o he obse e s pe o med he adap a ion expe imen s wice, wi h di e en blue-blocking il e s. In hese h ee pa icula cases he in e al be ween bo h expe imen s was la ge han a mon h. A o al o 12 subjec s we e using il e s wi h lowe ach oma ic poin shi (FG2 g oup), and 6 subjec s we e using il e s wi h highe ach oma ic poin shi (FG1 g oup). Resea ch A icle Vol. 30, No. 11 / 23 May 2022 / Op ics Exp ess 19761 All subjec s signed hei in o med consen o inclusion be o e hey pa icipa ed in he s udy. The s udy was conduc ed in acco dance wi h he Decla a ion o Helsinki, and he p o ocol was app o ed by he E hics Commi ee o he Uni e si y o G anada (1804/CEIH/2020). 2.3. Expe imen s 2.3.1. Tes s pe o med in each session Each session included ou es s: he ach oma ic poin measu emen , he FM100, and wo ounds o CAD h eshold measu emen s: one wi h he s anda d ach oma ic cen e (CIE 1931 ch oma ici y coo dina es 0.305, 0.323), and one wi h a cus om cen e (CIE 1931 ch oma ici y coo dina es 0.2175, 0.2142) co esponding o he blue egion in he CIE 1931 diag am. The FM100 es was ca ied ou in a cabin boo h Ve iVide CAC 60-5 wi h D65 illumina ion, and he s anda d se ing o he es ecommended by he manu ac u e . The To al E o Sco e (TES) gi en by he manu ac u e ’s so wa e was used o analyze pe o mance. Supe io disc imina ion co esponds o a TES alue anging be ween 0 and 20 [16]. We ha e designed an ach oma ic poin measu emen es simila o he one used by [9]. The s imuli we e p esen ed in a calib a ed colo display LG Fla on E1940. The me hod o 2AFC ou -in e lea ed s ai case was used o h eshold de e mina ion in ou di e en di ec ions: ed, g een, yellow and blue. The subjec ’s ask was o iden i y which o he wo squa e a eas p esen ed side by side on he sc een con ained a non-ach oma ic colo , p esen ed du ing 500 ms. In case he subjec pe cei ed he wo colo s as equal, he/she should p o ide a andom answe . The ch oma ici y coo dina es o he ach oma ic cen e we e L* = 46.53, a* = -0.031, b* = -0.55 wi h a luminance o 15.03 cd/m 2 . The e we e 11 s eps o app oxima ely equal size in he CIELAB space be ween he cen e and he u hes ends o he s ai case in each di ec ion, which we e loca ed a coo dina es (L* = 46.49, a* = 10.78, b* = -0.12) o he ed, (L* = 49.60, a* = -13.97, b* = - 0.023) o he g een, (L* = 46.47, a* = 0.24, b* = -9.71) o he blue, and (L* = 49.28, a* = 0.14, b* = 13.00) o he yellow. The backg ound colo was se o black (ch oma ici y coo dina es L* = 1.52, a* = 0.32, b* = -3.63) and luminance 0.17 cd/m 2 . The h eshold was de e mined as he mean o h ee e e sals o he s ai case in each colo di ec ion. Be ween he o se o he s imulus and he subjec s’ esponse, an adap a ion sc een wi h he black backg ound colo was p esen ed. Besides, a he beginning o he session, he subjec was adap ed o he da k oom condi ions and he black backg ound o he moni o o wo minu es. The ach oma ic cen e was no used o adap a ion because i would b eak he subjec ’s adap a ion s a e, and so condi ion his/he answe o he ach oma ic se ing. The ach oma ic poin was ob ained as he a e age o he ou h eshold se ings, and he h eshold was compu ed as he CIELAB colo di e ence wi h he cen e ach oma ic colo . The du a ion o his es was ypically a ound 3 minu es. In Fig. 2we p esen a ial example and he empo al dis ibu ion o p esen a ion and adap a ion sc eens. The CAD was ca ied ou using he s anda d ins uc ions o he ach oma ic cen e (x = 0.3050, y = 0.3230) and he blue cen e (x = 0.2175, y = 0.2142). The du a ion o each es was a ound 9 minu es. Eigh di e en di ec ions in he CIE 1931 ch oma ici y diag am along ed-g een o blue-yellow axes we e es ed, wi h wo colo s imuli in each di ec ion, and opposi e places om he e e ence colo cen e . The luminance o each s imulus was changed each 50-80 ms, and he subjec ’s ask was o de e mine he di ec ion o mo emen o he colo s imuli agains an ach oma ic backg ound. Apa om he h esholds compu ed by he CAD es so wa e in he -g and b-y axes, an ellipse was i ed and he ch oma ici y coo dina es o he 16 h eshold colo s ound du ing each CAD measu emen . The se o (x,y) ch oma ici y coo dina es was i ed o he equa ion o an ellipse, ob aining he pa ame e s a,b(semi-mino and semi-majo axis o he ellipse) and φ(o ien a ion). A e , he ellipse was plo ed using he i ed pa ame e s. Resea ch A icle Vol. 30, No. 11 / 23 May 2022 / Op ics Exp ess 19762 Fig. 2. Le : ypical sc een o he ach oma ic se ing es . Righ : empo al scheme o he ach oma ic se ing ials. 2.3.2. Expe imen al pa adigm desc ip ion Fo he con ol g oup, he e we e ou expe imen al sessions spanning 15 days, including he es s desc ibed in sec ion 2.3.1. Each session las ed 30-40 minu es, and he e we e 4-5 days be ween sessions. Fo he CA g oup, he es s in each session we e he same as o he con ol g oup. The CA g oup subjec s comple ed a o al o 10 sessions. Da a we e ga he ed be o e each subjec s a ed wea ing he il e ed lenses, and ollowing each 2-3 days wi hin he span o 15 days o using he lenses. The 15 h day o use, he subjec s pa icipa ed in wo sessions: one jus be o e aking o he lenses, and one jus a e aking hem o . Finally, one mo e session was comple ed, 24 h a e s opping he use o he il e s. Since he de-adap a ion was no comple e a e 24h o all he obse e s, we decided o pe o m an ex a session 7 days a e s opping he use o he il e s only o some subjec s. Fo hose who did, esul s showed ha he de-adap a ion was comple e a e one week o he end o using he il e s, and hence he colo ision es s esul s we e simila o hose a he beginning o he s udy (be o e s a ing using he il e s). 3. Resul s 3.1. Con ol g oup The esul s ob ained by he con ol g oup obse e s we e used o de e mine app oxima e anges o ole ance o in a-obse e a iabili y o each o he pa ame e s measu ed in he expe imen s (FM100 TES, he ach oma ic se ing h esholds, and he CAD h esholds). The ole ances we e de e mined in no malized uni s o he FM100 and CAD, di iding each o he sco es o h esholds by he esul ob ained in he i s session. Fo he ach oma ic poin se ings, he ole ance was de e mined in CIELAB uni s. Fo each pa ame e ( h eshold o sco e), he ole ance was compu ed as he a e age o he s anda d de ia ion in he pa ame e ac oss he ou sessions o each o he en obse e s in his g oup. In e -obse e a iabili y was also s udied by calcula ing he s anda d de ia ion o no malized h esholds ac oss obse e s in each session, and a e aging he alues co esponding o he ou sessions. In Table 2, we show he a e age no malized sco es, he in a-obse e a iabili y ( ole ance) and he in e -obse e a iabili y. These alues will be used o compa ison wi h CA g oup esul s. In Fig. 3(a) he FM100 a e age TES and s anda d de ia ion a e shown. The a e age TES alue ob ained is below 20 in all sessions, which acco ding o he manu ac u e co esponds o supe io pe o mance in colo disc imina ion. The a ia ion in a e age sco e be ween sessions is less han 5 uni s, and he ole ance is 1.1 uni s (see Table 2). Resea ch A icle Vol. 30, No. 11 / 23 May 2022 / Op ics Exp ess 19763 Table 2. No malized a e age sco es, in e -obse e a iabili y and ole ance (in a-obse e a iabili y) o he con ol g oup. The da a o he ach oma ic poin h eshold se ings a e in CIELAB uni s A e age sco es Tole ance In e -obse e a iabili y FM100 No malized TES 1.8 1.1 3.0 Ach oma ic poin h eshold 2.8 0.9 1.7 CAD ach oma ic cen e a0.97 0.08 0.1 b1.01 0.23 0.3 - g h eshold 0.97 0.09 0.10 y - b h eshold 0.96 0.11 0.12 CAD blue cen e a0.97 0.09 0.12 b1.03 0.18 0.23 - g h eshold 0.97 0.08 0.11 y - b h eshold 0.98 0.11 0.15 Fig. 3. Con ol g oup esul s. (a) Mean TES and s anda d de ia ion ac oss obse e s o each session. (b) A e age ach oma ic poin se ings in he CIELAB space o each session. (c) A e age CAD -g and b-y h esholds and s anda d de ia ions o each session wi h he ach oma ic cen e (d) A e age CAD -g and b-y h esholds and s anda d de ia ions o each session wi h he blue cen e . Resea ch A icle Vol. 30, No. 11 / 23 May 2022 / Op ics Exp ess 19764 In Fig. 3(b) we show he a e age ach oma ic se ing poin s o he con ol g oup in each session. As shown in Table 2, he a e age h eshold alue ac oss obse e s and ac oss sessions was 2.8 CIELAB uni s, wi h an in e -obse e a iabili y o 1.7 uni s and a ole ance (in a-obse e mean s anda d de ia ion ac oss sessions) o 0.9 uni s. These esul s sugges ha no mal obse e s a e no able o ob ain e y p ecise measu emen s o he ach oma ic poin , since bo h he a e age h eshold and he ole ance a e abo e wha is usually conside ed isually pe cep ible o colo di e ences in he CIELAB space. In Fig. 3(c) and (d) we can see he CAD -g and b-y a e aged h esholds o each session, wi h hei co esponding s anda d de ia ion alues o he ach oma ic and blue cen e s. These plo ed da a a e no no malized o isualiza ion pu poses. No e ha he e e ence no mal h esholds a e p o ided by he manu ac u e o he CAD o he ach oma ic cen e , bu no o he blue colo cen e . The con ol g oup alues a e wi hin he ange conside ed no mal by he CAD manu ac u e o ach oma ic cen e . Fo he bule cen e , he ole ances ound a e e y simila o ach oma ic one. The e is a sligh end owa ds lowe h esholds in he -g axis in bo h colo cen e s, and o sligh ly highe h esholds o he ach oma ic han o he blue colo cen e . Fo all he es sco es and esul s p esen ed in Table 2, he in e -obse e a iabili y is abo e he ole ance (in a-obse e a iabili y). The ole ance and in e -obse e a iabili y a e close o he CAD esul s han o he FM100 and ach oma ic poin esul s, which sugges s ha he e is less in insic a iabili y in he CAD esul s ac oss obse e s when compa ed wi h he a iabili y o each indi idual measu emen . 3.2. CA g oup wea ing il e s om FG1 The FG1g oup is o med by he wo Re ica e il e s (Re 1 and Re 2) and he Ess2 il e , all o hem p oducing ach oma ic poin shi s highe han 8 CIELAB uni s. Then, we should expec mo e no iceable e ec s on he colo ision o he obse e s wea ing hese il e s han o he FG2 g oup. A o al o 6 subjec s used hese il e s and comple ed he 10 expe imen al sessions. Fo each es , we ha e compa ed he esul s ob ained du ing he pe iod when he obse e s we e wea ing he il e s o he esul s o he i s and he las sessions, in which he obse e s we e no wea ing he il e s. A e he analysis o sessions-wise esul s (wi h he il e s), no clea end was ound. Hence, he a e ages ac oss all sessions wi h he il e s a e conside ed in he ollowing discussion. In Fig. 4, we show he no malized FM100 TES sco e in he wo condi ions (wi h and wi hou he il e s), o FG1 and FG2. In he ba g aph, he esul s o he i s and las sessions a e pooled oge he , while in Fig. 4 igh we can see he di e ence be ween he ini ial and he inal sessions. The TES sco es end o be highe when wea ing he il e s, and he a iabili y be ween obse e s’ esul s is also highe . Mo eo e , i can be obse ed how he subjec s end o e e o he o iginal TES sco e in he inal session. The di e ence be ween he wo condi ions is highe han he ole ance es ablished o he con ol g oup (1.1 o he TES sco es, see Table 2). In he igu e, he da a a e no malized o acili a e compa ison be ween obse e s. I we analyze he o iginal (no no malized) da a, he a e age TES sco e wi hou he il e is 13.6 o his g oup, and i ises o 17.02 when he il e s a e on. Bo h da a a e below 20, which could be conside ed as supe io disc imina ion, bu he e is a end owa ds wo sening esul s when he il e s a e used. In Fig. 5, we show he esul s o he a e age ach oma ic poin se ings in CIELAB colo space o FG1 and FG2, in he ini ial and las sessions, and on a e age o he pe iod when he obse e s we e wea ing he il e s. The e is e y li le a ia ion in he ach oma ic se ings, and no a clea end owa ds e e ing o he ini ial se ing in he inal session. The a e age h eshold in CIELAB uni s o he six obse e s wi hou il e s (ini ial and inal sessions) is 0.692, and he a e age h eshold wi h he il e s on is 0.987 uni s. Gi en ha he in a-obse e a iabili y o his g oup Resea ch A icle Vol. 30, No. 11 / 23 May 2022 / Op ics Exp ess 19765 Fig. 4. CA g oup esul s wi h FG1 and FG2 il e s o he FM100 es . Le : a e age no malized TES and s anda d de ia ion wi h il e s and wi hou il e s ( i s and las sessions). Righ : a e age no malized TES and s anda d de ia ion o he ini ial session, pe iod wi h glasses and las session. in he ach oma ic se ings h eshold is 0.91 uni s, he end owa ds a sligh ly highe h eshold wi h he il e s on canno be conside ed as consis en enough. Fig. 5. A e age ach oma ic poin se ings o he CA g oup esul s wi h FG1 and FG2 il e s, in CIELAB colo space. Ci cle: ini ial session. T iangle: a e age o he sessions wi h he il e s on. Squa e: las session. In Fig. 6we show wo examples o he i ed ellipses ob ained o he blue and ach oma ic cen e in he CAD es esul s. Fo he obse e in Fig. 6le , he e is a end owa ds inc easing he angle alue a e wea ing he glasses, and a conside able change in he ellipse size be ween he days 7 and 9 o wea ing he glasses. The ellipse ob ained in he inal session is simila o he one ob ained in he ini ial session, showing a end owa ds e e ing o he ini ial esul s. The subjec shown in Fig. 6 igh has a end owa ds an a ea inc ease o he ellipses du ing he pe iod wi h he il e on, ha ends o e e a e aking o he il e . These wo examples show ha he e is no clea end in he CAD ellipses o obse e s wea ing FG1 il e s, al hough he e a e ce ain a ia ions in he pa ame e s o he ellipses om he ini ial and inal sessions o he in e media e sessions in which he obse e s we e wea ing he il e s. In Table 3, we show he global a e age no malized sco es o he CAD pa ame e s (ellipse semiaxis leng h in he CIE 1931 space, and -g and y-b h esholds), as well as he o iginal a e age da a (no no malized), and he ole ance alues de e mined om he con ol g oup esul s. The e is a end o inc ease bo h he ellipse semiaxis leng h and h eshold alues in bo h -g and y-b di ec ions, when he obse e s a e wea ing he il e s. The end is he same o bo h colo cen e s, and o he mino axis (a) h eshold alues i is sligh ly abo e he ole ance o he