EVALUATION OF LONG-TERM EFFECTS OF EARLY AUDITORYDEPRIVATION: AN ANIMAL MODEL FOR OTITIS MEDIA.byTrudy AdamB.Sc., University of Calgary, 1990A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THEREQUIREMENTS FOR THE DEGREE OF MASTER OF ARTSinTHE FACULTY OF GRADUATE STUDIES(Department of Psychology)We accept this thesis as conforminghe required standardTHE UNIVERSITY OF BRITISH COLUMBIASeptember 1992©^Trudy Adam, 1992In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)Department of^PsychologyThe University of British ColumbiaVancouver, CanadaDate October 15, 1992DE-6 (2/88)ABSTRACTThis project aimed to establish a developmental animal model for the long-term effects of Otitis Media experienced in youth. The procedure employed wasbased on a model that mimmics the acute physiological symptomology of thesyndrome. It consisted of the infusion of lipopolysaccharides (derived fromthe plasmolemmae of Klebsiella pneumoniae, a pathogen known to cause theinfection in nature) into the middle ear cavity. In the young (12 day old)animal, it was found that a dosage 200 times that employed in previous modelswas necessary to induce like symptomology. This is thought to be due to theboosted immunity of the present rat pups stemming from clostrum in themother's initial milk production. Rat pups were raised to an age ofapproximately 35 days with their respective effusion conditions. Groups ofrats were used to study the effects of monolateral versus bilateral infections,severe versus mild hearing reductions, and in steady-state versus fluctuatingconditions. During and after the experience of middle ear effusion producedby lipopolysaccharide injection, audiograms were constructed to examinereductions in hearing level associated with the presence of the mucoid mass.Upon adulthood, prior to any other testing, open-field behavior was examined;focussing on exploration and re-exploration tendancies, reaction to novelty,and habituation to new environments. No consistent differences were notedfor emotionality, reaction to novelty, or exploratory behavior in general.Subsequently, the spatial navigational abilities of rats being studied wereexamined. Testing was conducted in both the visual (non-deprived) andauditory (deprived) modalities, employing the standard spatial water maze task.As expected, no group differences were observed in the visual version of thetask. However, the auditory task also failed to detect group differences.Implications and directions for future study of Otitis Media are discussed, andalternative behavioral paradigms suggested.111TABLE OF CONTENTSABSTRACT^ iiTABLE OF CONTENTS^ i vLIST OF TABLES vLIST OF FIGURES v iINTRODUCTION 1Characterization and Oetiology of Otitis Media^ 1Possible Long-Term Effects of Early OM Experience inHumans^ 6Methodological Problems Associated with Clinical Studyof OM 9Animal Study of Effects of Early Auditory Deprivation^ 10Development of the Rat Auditory System^ 14PURPOSES^ 15METHOD 18Subjects 18Suspension and Surgery.^ 18Behavioral Tests^ 21Early Audiometric Testing^ 21Open Field Behavior. 25Spatial Competency. 26Distal Visual Cues^ 28Distal Auditory Cues 29RESULTS^ 29Surgical Manipulation 29Audiogram Data^ 31Open Field Data 40Visual Water Maze Data^ 41Auditory Water Maze Data 45DISCUSSION^ 48Success of the Surgical Manipulation^ 48Relation of Effusion to Hearing Level Reduction^ 50Methodological Problems in Determining Thresholds 52Open Field Exploration^ 56Visual Spatial Water Maze 57Auditory Spatial Water Maze 58Methodological Problems with the Water Maze^ 59Conclusions and Recommendations^ 60REFERENCES^ 65APPENDICES 73Appendix 1. 74Appendix 2. 76Appendix 3.^ 78Appendix 4. 80Appendix 5. 82Appendix 6. 88Appendix 7.^ 91Appendix 8. 93Appendix 9. 99ivLIST OF TABLESTable 1.^Summary of results from the Tukey's HSD tests performed onaudiogram data.^ 37VLIST OF FIGURESviFigure 1. Depiction of the gross anatomy of the peripheral auditorynervous^system. 4Figure 2 Depiction^of the simple interaction between effusion^state^andfrequency,^as^they^affect^threshold^for^each^group. 33Figure 3. Mean daily escape latencies for each group of rats over ten daysof testing in the visual spatial water maze. 42Figure 4. Mean daily escape latencies for each group of rats, over thetwenty-five days of testing in the auditory spatial water maze. 45INTRODUCTIONCharacterization and Oetiology of Otitis MediaOtitis Media (OM) with effusion is one of the most common diseaseprocesses experienced by children.^Several demographic reports state thatover 67 percent of children have had at least one episode by the age of 3 years(Berko-Gleason, Ventry, Gray, McWilliams, & Gates, 1983; Bluestone, Klein,Paradise, Eichenwald, Bess, Downs, Green, Todd, 1986; Jenkins, 1986; Klein,1986). Past the age of 7, this disorder is reported to occur much less frequently(Giebink, 1986; Teele, Klein, Rosner, and The Greater Boston Otitis Media Group,1984).OM with effusion is characterized as an inflammation of middle earepithelial tissue and an accumulation of mucoid fluid in the tubotympanum(Bluestone, et al., 1983). It is caused by a number of pathogens, most of whichare known to occupy the soft palate (eg. Streptococcus pneumoniae,Haemophillus influenzae). These are thought to migrate from thenasopharynx to the tubotympanum via the Eustachian tube (Doyle, 1989;Kurono, Tomonaga, & Mogi, 1988). Normally, it would be cleared, but anexcessive influx of bacteria and/or a malfunctional Eustachian tube wouldcertainly facilitate infection.^Indeed, neonatal characteristics of theEustachian tube contribute to the high incidence of OM in young children. Atthese ages, this structure is relatively narrow, and horizontal in orientation.This not only contributes to the likelihood of onset, but also to the difficulty inclearing the cavity of OM effusion (Roland, & Brown, 1990; Todd, 1986).1There are two forms of OM. In purulent OM, the effusion is a result ofinterference with mucocilliary function of tubotympanic and Eustachian tubeepithelium by the antigenic surface of the bacterial cell membrane (Ohashi,Nakai, Ikeoka, Koshimo, Esaki, & Kato, 1988). This interference is thought toinduce inflammation of the Eustachian tube epithelium, resulting in blockageof the lumen. This results in the creation of negative pressure in thetubotympanum, and in turn, the production of exudate (termed an effusion)from submucosal vasculature in middle ear epithelium (Hellstrom, Salen,Stenfors, & Soderberg, 1982;. Kuijpers, van der Beek , & Willart., 1979).Production of an effusion may also be induced in the absence of any apparentpathogenic component through obstruction of the Eustachian tube.^In thiscase, the disorder is referred to as being "serous", and is absent of anypathogenic component (Bluestone, et al., 1983; Hellstrom, Salen, & Stenfors,1983; Stenfors, Hellstrom, & Salen, 1982; Hellstrom, Hermansson, Johansson, &Prellner, 1988; Kuijpers, et al., 1979; Proud, & Odoi, 1970). As the infection(whether purulent or serous) progresses, the effusion thickens, approachingthe consistency of glue. As this mass grows, some degree of conductivehearing loss generally occurs (Friel-Patti,1990; Friel-Patti, & Finitzo, 1990;Webster, Bamford, Thyer, & Ayles, 1989). During the final stages, the tympaniccavity is drained and apparently normal hearing restored.^This processappears to occur spontaneously after having persisted for some weeks ormonths; treatment with antimicrobial agents appearing to have little or noeffect on the time course of recovery (Bluestone, et al., 1983).One property that is particularly characteristic of the infectionpertains to the intermittent nature of the blockage. The volume of the2effusion, and therefore the extent of the accompanying hearing loss waxesand wanes for one to three months (Teele, Klein, Rosner, & The Greater BostonOtitis Media Group, 1984). The hearing loss accompanying the effusion variesin degree, but appears to average 25 decibels (dB); ranging from 10 to 40 dBrelative to hearing level (HL) (Bess, 1986; Friel-Patti, 1990). This is usually aconductive loss, but sensorineural involvement has also been suspected insome instances where the effusion has persisted for a significantly long time(Widemar, Sten, Hellstrom, Nordling, & Johansson, 1988).General audiometric measures taken from OM patients in the range of125 to 8 000 Hertz (Hz) yield profiles demonstrating that the hearing level shiftis parallel across the frequency spectrum examined (Bess, 1986; Bluestone etal., 1983). The hearing loss is more or less constant, regardless of thefrequency tested suggesting that the effusion serves to attenuate auditoryinput in much the same manner as earplugs do. This loss is generally termedto be conductive, where ossicular transmission of acoustic stimuli is impeded,while the inner ear (sensorineural) functions are left unaffected (Moore,1982). However, several studies conducted on humans as well as the animalpreparation have suggested that sensorineural loss may occur, due to diffusionof toxic immune products resulting from the infection into the inner ear(Goycoolea, Paparella, Goldberg, Schlievert, & Carpenter, 1980; Widemar,Hellstrom, Nordling, & Johansson, 1988). It has also been observed that theseverity of this loss is directly correlated with frequency (Moore, 1982). Thisis consistent with the notion that immunoreactive byproducts pass into theinner ear, as high frequencies are transduced along the basilar membraneproximal to the round window (see Figure 1).3Figure 1. Depiction of the gross anatomy of the peripheral auditory system. Thisincludes the outer ear, middle ear (or tubotympanum), and inner ear. Note thetube leaving the ventral portion of the middle ear. This is the Eustachian tube.The pars flaccida is the dorsal portion of the tympanic membrane overlyingthe ossicular chain (from Moore, 1982., p. 14).4Relative to the rest of the inner ear, concentration of immunity-related toxinsshould be highest here during diffusion from the tubotympanum.OM sequelae commonly involve more than experience with periods ofattenuated auditory input.^Effects noted primarily involve structuresassociated with the tubotympanum. For example, the tympanic membrane mayperforate as a result of tension produced by the effusion. A large rupture isusually associated with a hearing loss of approximately 30 dB, which is amarked deficit (Bluestone et al., 1983). Tympanosclerosis is also reportedfrequently. This is characterized by the presence of opaque white plaques inthe tympanic membrane, and by nodules in the mucosal lining of thetubotympanum (Kuijpers, Wielinga, Tonnaer, & Jap, 1988). It is generally notconsidered to have a significant effect on hearing, unless middle ear ossiclesbecome embedded in the nodules (Ohashi, et al., 1988). The epithelial tissue ofthe tubotympanum may also become thickened, and fibrous tissue mayaccumulate within the cavity (Kuijpers, Wielinga, Tonnaer, & Jap, 1988).Again however, conductive hearing losses are not experienced unlessossicular movement is impeded (Bluestone, et al., 1983).Possible Long-Term Effects of Early OM Experience in HumansAside from anatomical changes, the existence of long-term behavioraleffects of early OM have been the subject of much controversy. Reports ofdecreased cognitive ability and deficient scholastic performance have oftenbeen attributed to that history of long-term hearing impairments in childhood(e.g. Friel-Patti, 1990;).Much evidence has accumulated favoring the presence of long-termdeficits stemming from early experience with OM. Of these, auditory6processing deficits are the most frequently cited disorders, and these arethought to be mediated through hearing; not through some other sequela ofOM (Friel-Patti, & Finitzo, 1990). Included are problems with attention,discrimination, and sequential memory in the auditory modality. Theseimpairments are thought to interfere with the development of notably speech,reading, and mathematics, as these rely quite heavily on auditory input(Rourke, & Finlayson, 1977; Zinkus, 1986). It is commonly during these earlyyears of schooling that OM effects first surface, as impaired hearing mayquite easily occur in the absence of other overt symptomology.Research has been reported suggesting that the mild fluctuatinghearing loss experienced with OM may well produce similar deficits in later(even adult) language functioning.^These impairments seem especially likelywhen OM was experienced between 6 months and 3 years of age (Clopton, 1980;Jenkins, 1986; Menyuk, 1986; Paradise, 1981; Roberts, Burchinal, Koch, Collier,& Henderson, 1988; Strange, 1986; Teele, et al., 1984; Wallace, Gravel, McCarton,& Ruben, 1988). These are held to be the formative years for languagedevelopment, a process that relies critically on listening to speech producedby others, attaching meaning to this input, and formulating a vocabulary andgrammar on this basis (Friel-Patti, Finitzo, Meyerhoff, & Hieber, 1986; Roberts,& Sanyal, Burchinal, Collier, Ramey, & Henderson, 1986).When one considers that the the average hearing loss experienced byOM patients would represent a significant degradation of normalconversational speech, it is not surprising that language development wouldbe detrimentally affected by early OM effusion (Matkin, 1986; Nozza, 1988).Further, the claim that a 25 dB HL hearing loss is mild is based on data gathered7from older, more easily tested children. While this may hold true for them, theequivalent reduction in a younger child (of 3 years, for example) may prove tobe severe, as more reliable input may be required at this linguistic stage(Menyuk,1986; Nozza, 1988).It is plausible that inconsistent or unstable decrements in the speechsignal input would impair the ease with which an individual acquiresprinciples of speech processing. A mild and fluctuating hearing loss couldresult in the reception of an inconsistent auditory signal, and subsequentconfusion of important information upon which to base languagedevelopment (Roberts, & Schuele, 1990). Some researchers even purport thatthe unstable input received by the OM child is more detrimental to laterlanguage functioning than a non-fluctuating sensorineural loss.^Chronicbouts of OM are thought to be especially damaging for the acquisition of finephonetic discriminations (eg. "place" versus "plays") necessary for normallanguage development. This would hinder the formulation of linguisticcategories such as tenses and plurals (Bluestone, et al., 1983; Clarkson, Elimas,& Marean, 1989; Dobie, & Berlin, 1979; Feldman, & Gelman, 1986; Friel-Patti,1990; Glasberg, & Moore, 1986; Landis, 1990; Leviton, & Bellinger, 1986; Menyuk,1986; Steig-Pearce, Saunders, Creighton, & Sauve, 1988; Strange, 1986; Watkins,1990).Many of those opposing the notion that early OM affects later languagefunctioning claim that recovery from acute OM is complete, without long-termconsequence.^Evidence for this appears to rely primarily on the observationthat OM children obtain normal audiograms after recovery from the effusion(Friel-Patti, 1990).^In addition, a longitudinal study of children attending the8same daycare demonstrated no reliable deficit in the utterance of specificphonemes, but did report that OM children did grow out of making phonemicerrors more slowly than controls (Roberts, Burchinal, Koch, Footo, &Henderson, 1988). However, these results are equivocal, and requirereplication.Methodological Problems Associated with Clinical Study of OMData supporting the existence of long-term OM effects are also equivocaland subject to alternative interpretations.^Deficits noted have been attributedto OM sequelae unrelated to reduced auditory input (Bishop, & Edmundson,1986; Jenkins, 1986). For example, an OM child who is ill much of his/her timecertainly stands a strong chance of having difficulty with the acquisition ofmany perceptual, linguistic and cognitive capabilities.^Indeed, all long-termOM sequelae may not be attributable to the peripheral auditory restriction, andthis should be kept note of in studying the potential long-term effects of OM-related auditory deprivation.Clearly, the connection between OM and language development isconceptually compelling, but tenable at best. The delineation of the exacteffect of OM on later language functioning has proven to be a formidable task.Not only may OM patients generally be able to compensate largely for theirdeficit, but symptoms may only be obvious when conditions are difficult forthe normal listener too (Bess, 1986; Friel-Patti, 1990).In addition, clinical investigation of OM tends to be plagued by designweaknesses. Many problems described appear to center around vaguedefinitions and criteria for the severity of the language deficit at hand, as wellas extreme difficulties in measurement.^Indeed, even information about the9severity of the OM infection itself and resulting auditory deprivation are oftenelusive to the researcher (Shurin, Johnson, & Wegman, 1986; Ventry, 1980). Asof yet, research conducted tends to employ between-subject designs thatgenerally fail to control for the highly variable nature of the disorder withinand across individuals. This is exacerbated by the lack of descriptive datapertaining to an individual's OM history, and further compounded by theretrospective and anecdotal nature of studies conducted (Bluestone et al., 1983;Matkin, 1986). Clearly there is a need for more rigorous and longitudinalclinical studies. For example, higher order linguistic processes may well needto be addressed in order to detect any subtle effects of early OM. Further,emphasis needs to be expanded from language capacities in particular, toinclude examination of basic auditory processing skills, such as theappreciation of spatial relations among incoming auditory stimuli.Animal Study of Effects of Early Auditory DeprivationIn outlining some of the methodological difficulties involved in theclinical study of OM, it appears that some benefit should emerge from the studyof changes in auditory competences with measured changes in early auditorystimulation history. The nature and degree of prior auditory experienceshould be important in perceptual development of the organism, givenappropriate controls.^Much information has been gained from the animalstudy of early visual deprivation and adult functioning, and parsimonynecessitates that at least some of the same developmental principles must holdtrue for the auditory modality.With respect to vision, changes noted not only include behavioraldeficits but also dramatic alterations in central nervous system structure1 011underlying these abilities (Kyle, 1978).^Although this is still controversial,visual acuity appears to remain relatively unaffected by restricted visualexperience such as dark-rearing, provided normal visual input is obtained fora time thereafter.^Conversely, an appreciation (or responsiveness) to complexvisual input seems to be deficient (Tees, 1990). Similar findings have begun toemerge in the clinical study of auditory deprivation, but they tend to becontradictory and sparse.^Case studies where early auditory deprivationexperienced was fully overcome in terms of peripheral sense organ damageare difficult to achieve as much of the organ is relative inaccessible toexamination at the time of testing. However, such an approach is certainlyappealing, as any deficits observed may be ascribed to the abnormaldevelopment of structures, (and their respective functions) of the centralauditory nervous system, rather than the peripheral auditory system (Kyle,1978).Through the development of the animal preparation, it is presentlyhoped that some of the questions raised thusfar may be addressed using amodel of OM. Using animals, one is able to manipulate the auditory experienceof subjects, and verify that peripheral structure is undamaged in the adult, anadvantage unavailable to the clinician, and in this manner avoid some of theproblems described. Several notable attempts have been made, but these tendto fall short in the requirements of modelling OM.Early work entailing the rearing of animals in an acoustically deficientenvironment have found that the adult animal is relatively insensitive tosound in general, some cases not even developing a reliable pinna reflex(Batkin, DiCarlo, & Sission, as cited in Kyle, 1978). This was later supported byBatkin, Grath, Watson, & Ansberry, (1970) in the rat where mature auditorythresholds were consistently 25 dB relative to sound pressure level (SPL)higher than the normally reared animal (as cited in Kyle, 1978).Anatomically, deprivation results in decreased neuronal soma size, cellcounts, and less extensive dendritic arbors in structures such as the cochlearnucleus, inferior colliculus, and cortical areas.^These effects are associatedwith animals that have experienced ear plugs throughout development (forreview, see Moore, Hutchings, King, & Kowalchuk, 1989; Webster, & Webster,1977). In addition, broad tuning curves are observed to persist throughadulthood following a deprived stimulation history. This is attributed in largepart to retarded development of cochlear analysis abilities which serve toprovide the mechanism of curve sharpening in development. Increases in thefiring latency and alterations in the response patterns of central neurons arealso reported (Clopton, 1980; Clopton, & Silverman, 1978).Behavioral findings stemming from animal study appear to indicate thatdeficits involve more complex auditory functioning, such as temporal durationand pattern discrimination.^Frequency and intensity discrimination appear^tobe unaffected by auditory deprivation (Tees, 1967a; Tees, 1967b). Thus, theeffects of early stimulation history may be said to be more complex thansubsequent sensory retardation.^Appreciation of auditory information, andthe high order processing of it are clearly deficient following restrictedsensory experience.Evidence such as that outlined here has sparked much researchconducted of late in the clinical study of OM attempting to demonstrate a causalrelationship between early OM and delayed or abnormal language12development. Indeed, a developmental model of OM in the animal preparationwould not only entail more control over possible extraneous factors thanclinical investigation, but would present a model more directly related to thehuman condition. Thusfar, only models addressing the pathology andphysiology of the disease in the adult rat have been established.Developmental aspects of OM have yet to be examined, as do perceptual andcognitive facets of adult auditory behavior in these animals.In the physiological models established, purulent OM is most frequentlyinduced by infusing viable bacteria directly into the tubotympanum.^In thesecases, the mechanism through which an effusion is produced is thought to beincreased vascular permeability in the mucosal lining through a chain ofevents starting with the presence of a pathogen (Ohashi, et al., 1988).Normally, the mucocilliary system and Eustachian tube would drain this intothe pharynx but lipopolysaccharides (LPS) present in the plasmolemmae ofthe bacteria interfere with this action by ceasing or desynchronizing cilliarysweeping, causing inflammation of the mucosal lining, and secretion ofabnormal mucoid fluid. This prevents effective drainage of the cavity for 7 to10 days (Ohashi, et al., 1988).Obstructing or ligating the Eustachian tube is generally the meansadopted in producing serous OM with effusion (Hellstrom, et al., 1988; Jung,Hwang, Poole, Olson, Miller, Lee, Yoon, & Juhn, 1988; Kuijpers, et al., 1979;Proud, & Odoi, 1970). In this case, occlusion of the Eustachian tube createsnegative pressure in the middle ear cavity, which in itself results in therelease of fluid from the mucosal membrane lining it (Kuijpers, & van derBeek, 1984).13Despite these models, there has been no account of short- or long-termeffects of OM in basic auditory functioning, communicational abilities, orspatial competences. There has been no successful attempt to induce thisinfection in neonatal animals, nor has any behavioral data been collected.^Inthe animal preparation, OM completely lacks characterization, barring short-term physiological effects. As a result, it is presently being forwarded thatinfusion of LPS derived from the cell membranes of pathogens known to causethe disorder will generate a syndrome very similar to that induced usingviable bacteria. Effusion should ensue as readily as in the natural occurenceof the disease, as should interference with the mucocilliary system andEustachian tube. Experimentally induced effusions are reported to inducehearing losses of approximately 15 to 30 dB HL, a level more realistic in termsof human clinical occurrences than the drop of 40 to 55 dB HL accompanyingossicular removal or canal blockage (Proud, & Odoi, 1970). Deficits observedunder this less severe condition of deprivation provides more realistic,applicable, and compelling results, and allows for more valid generalization tothe human condition.Development of the Rat Auditory System In addition, the existence of relatively well delineated data regardingthe ontogeny of the rat auditory system may allow selective deprivation ofparticular aspects of auditory function, as sensitive periods have in large part,been identified in this animal. At birth, the rat external auditory canal existsbut is filled with connective tissue. Around day 10 this begins to subside, andthe canal becomes apparent upon visual examination by approximately day 12or 13. The tubotympanum at birth is completely filled with connective tissue14similar to that present in the canal. By day 9, this also begins to clear,becoming totally air-filled by day 21 (Hellstrom, Cerne, & Stenfors, 1984).Observable functionality in the auditory system is apparent at 10 to 15 days,developing rapidly during this time. Cochlear potentials in response to soundfirst appear at around 10 days, gradually improving in sensitivity andfrequency range until adult levels are reached at approximately 20 days of age(Crowley, & Hepp-Reymond, 1966). Complex auditory functioning (such asspatial localization) continues to develop, and remains modifiable untiltermination of head growth (at approximately 30 days) (Clopton, 1980; Clopton,& Silverman, 1978).As a result, this model may provide a convenient alternative to methodspreviously employed to study auditory development while circumventingmany of the problems associated with their lack of reversibility. Even if thismodel does not prove to be free from these problems, it is sure to aid in thecharacterization of long-term deficits stemming from OM.PURPOSESThe first purpose of this experiment was to attempt to establish LPSinfections in young (12 day old) rats, prior to the onset of auditorydevelopment, and to maintain effusion conditions until the basic auditoryprocessing abilities have been established in the normal animal (atapproximately 30 days of age). The rat provides an excellent model for OM asits Eustachian tube is very narrow and of horizontal orientation. In addition,the tympanic cavity of the rat is relatively large (Hermansson, Emgard,Prellner, & Hellstrom, 1988). As a result of these two characteristics, OM should15be relatively easy to induce in the rat, and symptomology should closelyresemble that noted in humans. The effusion should bring on a reduction inauditory sensitivity of about 10 to 25 dB SPL, and should fluctuate in severity.In an attempt to mimic the human condition, and to compare it to moresevere conductive losses, a comparison group of rats raised with constant,severe effusions were studied. In addition, effects of monaural versus binauraleffusions were explored, as monaural deprivation would be expected tointerfere with spatial aspects of auditory perception.^Both of these parameters(strength and laterality) are of great concern to the clinical study of OM, asthey may contribute significantly to the diversity in severity of effects noted.The efficacy and reversibility of the effusion induction method wasexamined through the measurement of rat auditory thresholds during variousstages of effusion and again, after sufficient time has passed for recovery (asindicated by clear otoscopic exams). Should immunoreactive agents producesome sensorineural loss in the cochlea, a frequency-dependence in hearingloss should be apparent in that persistent hearing loss will be evident forhigher frequencies than lower ones Very low frequencies (<= 1 kHz) areconducted through the skull as vibrations, and therefore may be heard evenin the presence of a conductive hearing loss (Buser, & Imbert, 1992; Yost, &Nielsen, 1985). As a result, any impedance in the transmission of acousticstimuli by the middle ear ossicles should interfere with the sensation of allfrequencies higher than 1 kHz, but not with those at or below it (a frequency-dependance of sorts, independent of sensorineural loss).Because there are little behavioral data on rats that have experiencedOM, it was necessary to examine their general behavior as adults. In order to16accomplish this, adult OM rats and controls were studied in an open fieldsituation intended to disseminate possible differences in anxiety or timidity, aswell as aspects of exploratory behavior, and reactions to novelty. Thistechnique provides a convenient means of obtaining indices for emotionality,exploratory tendencies, habituation, and memory for previously encounteredobjects and their spatial relations (Ennaceur, Cavoy, Costa, & Delacour, 1989;Thinus-Blanc, Bouzouba, Chaix, Chapuis, Durup, & Poucet, 1987; Poucet,Chapuis, Durup, & Thinus-Blanc, 1986). In this manner, it is believed that OMeffects on more general aspects of behavior may be obtained, and support maybe lent to the notion that any later effects noted are due to early hearing loss,and not other OM sequelae.Subsequently, the preliminary analysis of spatial ability warrantedexamination using signals of both visual and auditory modalities. A variant ofthe Morris water maze task was used in this regard (Morris, 1981; Sutherland, &Dyck, 1984). This paradigm not only allows one to test spatial competence, butprovides information about the type of strategy subjects adopt in attempting tosolve the problem (Sutherland, & Dyck, 1984). Investigations of the effects ofvisual deprivation on spatial ability report that dark-reared rats are impairedin their ability to acquire and remember spatial locations using distal visualcues (Tees, Burhmann, & Hanley, 1989). Should OM also have long-termperceptual and/or cognitive effects on spatial navigation, differences shouldarise between the performances of OM and normal rats on an auditory, but nota visual version of the task. Monaural versus binaural effects will also beexamined, to determine if monolateral deprivation plays a role in theestablishment of auditory spatial competence in the adult animal.17METHODSubjects. The subjects were 50 male Long-Evans rats bred in colony facilities atthe University of British Columbia. They were maintained in a 12/12 hourhour light/dark cycle, as well as fed and watered ad libitum throughout theentire experiment. Rats were assigned to one of five groups: Severe BinauralEffusion (SBE), Severe Monaural Effusion (SME), Mild Binaural Effusion(MBE), Mild Monaural Effusion (MME), and Saline Control (SC). All effusionswere produced by the injection of a lipopolysaccharide suspension into thetubotympanum of the rat (see Hellstrom et al., 1988; Ohashi, et al., 1988). Ratsin the binaural groups received such treatment to both ears, while monauralgroups had only one ear treated. In the latter case, left ears and right earswere counterbalanced within each group. All untreated ears were subjected toa sham procedure entailing the injection of sterile saline at the same timesthat ears of experimental rats were infused. Two concentrations (mild andsevere) of LPS suspension were employed in this study, and are describedbelow. Adult behavioral testing in the open field and water maze tasks beganwhen the rats were 70 days of age.Suspension and Surgery. The LPS suspension used is similar to that utilized by others examiningpossible models of OM in an adult animal preparation (Hellstrom, et al., 1988;Ohashi, et al., 1988). Lipopolysaccharides extracted from the cell membrane ofKlebsiella pneumoniae^using the trichloroacetic acid procedure (protein18content being 1-10%) were employed (Sigma). A sample of 15 mg of LPSpowder was suspended in 500 ml of sterile saline yielding a suspension ofapproximately 30 gg/11.1 concentration. This solution was infused in ears ofexperimental rats belonging to severe effusion groups. For mildly effusedrats, 7 mg of the LPS was suspended in 500 ml of sterile saline, and infused intotheir ears.Rats were anesthetized using halothane at the age of 12 days. Usingblunt dissection techniques, the ear canal was opened, and the tympanicmembrane exposed. Using forceps to hold the canal open, the appropriatesolution to be injected was introduced using a 10 ml syringe passing throughthe membrane, into the tubotympanum, at a point against the rostroventralportion of the ear canal. The needle on this syringe was of a fine enoughdiameter as to minimize damage to the eardrum area adjacent to the malleus,through which injections were performed. The volume of solution injectedwas sufficient to fill the cavity, and the infusion stopped when the solutionbegin to overflow from around the needle into the ear canal. In older rats,this involved a second injection as the cavity volume grew to exceed 10 gl. Thissecond injection was performed at the same site as the initial injection.For 30 days following LPS infusion, the ear canal and tympanicmembrane were examined otoscopically for changes in the appearance of thetympanic membrane, particularly the pars flaccida.^This portion of themembrane surrounds the head of the malleus and bulges or retracts inresponse to changes in attic space (or superior middle ear) pressure. Indeed, itis crucial in the regular maintenance of the attic space air pressure(Hellstrom, Cerne, & Stenfors, 1984). In normal ears, the tympanic membrane19appears to have a sharp peak at the malleus head, but during middle earinfections, it is flat or bulged due to pressure from the middle ear effusion(Stenfors, Hellstrom, & Salen 1982). In addition, the tympanic membranegenerally appears cloudy or opaque and less mobile than normal.Otoscopic examination of rat ear drums was conducted while the animalwas lightly anesthetized using halothane.^During these exams, the followingscheme was utilized for grading the severity of an effusion stemming from OM,and the rats were maintained in the desired severity of their infectionthroughout the time period important for auditory development (Marrow,Whitley, & Fulghum, 1988; Karma, Penttila, Sipila, & Timonen, 1988):0 - No visible signs, therefore no apparent infection.1 - Very mild inflammation, mild bulge of eardrum.2 - Prominent bulge.3 - Large bulge, and eardrum translucent.4 - Very large bulge, eardrum opaque.5 - Rupture (discard subject).This type of diagnostic measure has been demonstrated to be morereliable than other, more prevalent methods such as tympanic membranecolor, and is consistent across both patients, and raters of acute OMsymptomology (Karma, Penttila, Markku, Sipila, & Timonen, 1986). Anyexudate observed in the ear canal was noted and described. This is indicativeof Otitis Externa, which often accompanies severe OM, the fluid leakingthrough the tympanic membrane (Marrow, et al., 1988). For ears in the SBEcondition, the reappearance of a slight peak in the tympanic membrane(infection grading of 2) indicated the time for reinfusion of LPS. The animal2021was anesthetized using halothane, and the surgical procedure repeated.^Thisallowed for a total auditory deprivation time of approximately 30 days (untilthe animal was about 40 days of age) for SBE and SME animals. For mildlyeffused ears, reinfusion was not performed until a prominent peak in thetympanic membrane was visible for 2 days (a grading of 1). In this manner,the mild groups were subjected to a fluctuating loss in auditory inputthroughout the system's development. (At the dosage used for mild effusions,the infection was less severe at its worst stage than that used for severeeffusion production; being graded only as a 3 rather than 4 or 5). Any animalnoted to have a ruptured eardrum at any point was eliminated from the study.Control ears were subjected to the same procedure, with sterile salinebeing injected into the cavity rather than the LPS solution. It was observed inpilot animals that this solution is cleared from the cavity within 24 hrs.Control ears were checked for evidence of effusion at the same times as theexperimental ears, to ensure that no natural cases of Otitis Media were evidentand to expose control animals to similar experiences with handling andanaesthetic. This was also done to ensure that the control ears ofmonolaterally effused rats did not develop symptomology.Behavioral TestsEarly Audiometric TestingStimuli and ApparatusBoth during the acute phase of OM effusion and after sufficient time forrecovery, pilot rat pups were tested for hearing sensitivity.^Fifteen rats (5saline controls, 5 mildly effused, and 5 severely effused) were subjected to thesurgical procedure described above. Due to the nature of the testingprocedure, examination of hearing sensitivity did not begin until the ratswere 20 days of age. At this age, thresholds begin to stabilize at adult levels,and the rat is capable of eliciting reflexive behaviors such as orienting of thehead and pinnae reliably (Crowley, & Hepp-Reymond, 1966; Rudy, & Hyson,1984). Audiograms were constructed during the severe phases of infection andbetween effusions (for mildly effused rats), just prior to the subsequentinjection. Rats tested during these phases were 20 to 30 days of age. Uponrecovery from the final effusion period, as determined otoscopically (atapproximately 35 to 45 days of age) rats were tested again to determine post-effusion thresholds (Hyson, & Rudy, 1984).Auditory stimuli were comprised of trains of 10 tonepips each, at 1, 8,and 16 kHz. Each tonepip was 400 ms long, with 50 ms rise and fall times toavoid artifactual frequencies being emitted at onset and offset of the stimulus.Ten tonepips separated by 100 ms interstimulus intervals were presentedduring each stimulus train. Sound pressure levels at the centre of the testingarena were calibrated for each frequency using a microphone orientedtowards each speaker. Intensity readings were taken with a spectrumanalyzer (Data Precision 6 000, Model 610) in order to obtain true intensityreadings at this point for each stimulus train (1, 8, and 16 kHz).All threshold testing was conducted in a pseudo-anaechoic chamber(Industrial Acoustics Co. Ltd.). The floor of the testing arena was circular, and50 cm in diameter. This area was subdivided into eight equally-sized, andnumbered sections. The rat was restrained by a cage (15 cm in diameter), of22sorts, constructed of wire mesh standing 5 cm high, with a wire mesh ceiling.Stimulus trains were produced by an IBM Personal computer (XT). Tones weregenerated by a 2 MHz function generator (Dynascon Corp., BK Precision, model3011), and passed through a Technics Stereo Integrated Amplifier (SU-700).Two speakers (KEF Model T27) through which stimuli were presented werelocated at the periphery, symmetrically arranged on either side of the startingpoint. The distance between the speakers was 45 cm. Rats were tested underred light conditions produced through the use of red light bulbs.Procedure.Measurement of acoustic thresholds in these (15) rats employed theunconditioned head orientation response used by others studying auditorydevelopment in this species (e.g. Kelly, & Masterton, 1977). Rats were allowedto move about freely within the testing cage. A five minute adaptation periodinitiated a test session, wherein rats became accustomed to the quietsurroundings. Following this time period, each rat was observed until his headand body were oriented perpendicular to the plane of the two speakers, at thecenter of the arena. At this point, a stimulus train was delivered through aspeaker at either the - 90 or + 90 degree azimuth. The frequency of the signal(1, 8, 16 kHz), and speaker (left versus right) were determined randomly, bythe Gellerman sequence (Fellows, 1967; Gellerman, 1933; Goldstein, 1967). Theintensity of tones presented was varied in an attempt to obtain thresholdmeasures using the Von Bekesy method of 'high-low approach' (Yost, &Nielsen, 1985). Stimuli of a given frequency were presented in a trial serieswhere their intensity was lowered until orientation responses were no longerapparent.^Presentation was then in ascending order until detection was2324reliable.^Near-threshold intensities were emphasized during testing, andpresented in random order until each intensity step has been presented 3times. For the random-order trials, threshold was defined as the intensity atwhich two out of three stimulus presentations resulted in an appropriateorientation response.During testing, mock trials were interspersed with test trials to ensurethat orientation responses were not being elicited in response to someacoustical artifact emitted in conjunction with stimulus presentation.^Inaddition, this procedure allowed for control over the occurence of spontaneousmovements and over possible group differences in general activity levels.Stimulus presentation conditions during a mock trial were identical to those ofa test trial except that the intensity with which a given frequency waspresented was set at zero. The experimenter was blind as to the groupassignment of each rat being tested at the time, and to the actual intensity atwhich stimuli were presented.As described by Kelly, et al., (eg. 1987), a response was defined as beingan observable lateral movement of the head (a minimum of 30 deg. to eitherside), or orientation of the pinna(e). One of these responses had to occurwithin 3 s of stimulus onset. Non-directional responses and failures to elicitresponses within the 3 s were also noted. Responses such as these (and thePreyer reflex) are related to the absolute audiogram of the rat, and are anindication of the behavioral threshold of the rat (Francis, 1979). Thus, a ratwith effusion and related hearing loss should exhibit these behaviors athigher intensities than normal rats.^Psychophysical functions (or simpleaudiograms) were then constructed on the basis of data obtained.Open Field Behavior. The testing arena was comprised of a square field (120 x 120 cm),enclosed by walls standing 30 cm high. This field contained four objects. Thefirst of these was a piece of wood (23 x 5 cm) with course wire mesh (30 x 23cm) attached along the midline, perpendicular to the board. The second objectwas comprised of a tube (8 cm in diameter, and 13 cm long), with wire mesh (13x 30 cm) attached. The third object was a piece of wire mesh (47 x 4 cm) thatwas bent every 15 cm or so, creating peaks and troughs along its length. Thefinal object was comprised of a relatively large cable spool (13 cm in diameter)with a 25 x 15 cm piece of wire mesh attached. In all cases, the wire meshattached to objects was arranged such that it would provide a sloping grid thatrats could climb on or go under.All rats were tested on an individual basis, at approximately the sametime of day to avoid possible differential effects of circadian rhythms on therats' open field behavior (Thinus-Blanc, Bouzouba, Chaix, Dump, & Poucet,1987). Three of the four objects described above were located at the center ofeach of three quadrants. The fourth was left empty. Rats were placed in thearena facing the outside corner of the empty quadrant, and allowed to explorethe area and objects therein for 8 min. This comprised the first test session.Rats were then returned to their home cages until the second 8-min testsession, 24 hrs later. During this phase, the arrangement of the objects wasaltered on order to study effects of total object novelty versus a novel spatialarrangement of a previously experienced object. One object was exchangedfor a novel (the fourth) one. Another object was moved to another locationwithin the same quadrant, and its orientation changed. The final object was25left unchanged. Assignment of objects to a rat's particular test situation wasrandom. The apparatus and objects were cleaned between trials. Data collectedduring the sessions consisted of 1) the amount of time a rat spent in eachquadrant, 2) the number of contacts with each object, 3) the number ofdefecations, and 4) incidents of rearing.^Criteria for scoring behaviors wereselected from those employed in anumber of open-field studies investigatingdifferent aspects of rat benavior (see Deyo, et al., 1990; Poucet et al., 1986;Thinus-Blanc, et al., 1987). A rat was considered to have entered a particularquadrant when his forepaws crossed over the line delineating the partitionbetween the two. A contact was counted only when the rat touched an objectwith his nose, vibrissae, or paws. A subsequent contact was only considered tobe a separate one, and scored as such if it followed a physical separationbetween the rat and the object equal to approximately half of the rat's bodylength. A rear was scored if the rat's forepaws were raised off of the arenafloor.Spatial Competency. Examination of spatial ability in rats with and without early auditorydeprivation was conducted using a procedural variation of the task developedby Morris (1981). Abilities in both the visual and auditory modalities werestudied. All rats were tested in both modalities, and the order of task performedcounterbalanced within each group.The water maze consisted of a circular, white fiberglass pool of diameter180 cm. This was filled with water to a depth of 30 cm, and made opaque usingwhite paint (Tempera). At testing, the water was maintained at room26temperature (21 deg. C). A jar filled with stones for ballast, and topped with agrid of wire mesh served as the escape platform. This jar was 28-29 cm high,allowing the platform to be submerged approximately cm below the water'ssurface. The pool was surrounded by curtains hanging from the ceiling andreaching beyond the pool rim. Cues denoting the position of the platformwere placed at three of the four compass points around the pool, and theirspatial arrangement with respect to each other maintained throughout thetesting period. The platform was located in the center of a given quadrant,between two of the three cues, and the cue-platform relationship wasmaintained throughout testing. Suspended above the center of the pool was avideo camera that served to collect information regarding the rats' behaviorin the pool during a given trial. The entire surface of the pool could be viewedby the experimenter via a VP112 monitor connected to the camera. Dataregarding escape latency was gathered using a HVS Image Tracking System.This system recorded a particular animal's position in the pool at a sample rateof 10 traces per second.The general procedure was conducted as follows: The location of theplatform for a given trial was determined randomly, and cues placed inpositions appropriate for that trial. For both modalities, these cues were placedat three of the four compass points around the pool, and their spatialarrangement with respect to each other maintained throughout the testingperiod. The platform was located in the center of a given quadrant, betweentwo of the three cues, and the cue-platform relationship was maintainedthroughout testing. With each new trial, the platform was located in aquadrant determined by random sequence. The cues were moved with the2728platfrom to maintain their spatial relationship with it.^Over the trialsconducted in one day, the rat was released from each of the 4 compass pointsan equal number of times. With this method, it was believed that the rat wouldbe discouraged from using irrelevant cues in platform search strategiesThe rat was released in the pool facing the wall, through a gap in thecurtains at one of the four compass points chosen in random order. Amaximum of 90 s. was allowed for the rat to attain the escape platform. If thistime expired without the rat escaping, the rat was "led" to the platform by theexperimenter. Once the rat climbed onto the platform, 10 sec. elapsed beforethe rat was returned to the holding cage. Rats were tested with an intertrialinterval of approximately 10 minutes. Eight trials were conducted per day. Foreach trial, information regarding latency and search strategy were recorded.Rats were tested for 10 days, by which time the average control animalhad attained criterion (escape latencies of less than 10 sec., 7/8 trials over 2days). This constituted the acquisition phase. The following day, rats weretested in the normal manner, but without the platform being present in thepool. This constituted the probe phase designed to establish the searchstrategy being employed by rats in the task. Subsequently, rats were testedwith the platform to bring them back to pre-probe performance.Distal Visual Cues Three cues providing information as to the location of the escapeplatform were mounted on the curtains 150 cm above the floor of the testingroom (127 cm above the surface of the water). Each cue then subtended avisual angle of approximately 10 deg. from the centre of the pool. The first ofthese was comprised of a large (20 x 25 cm) 'XX'. The second cue was a large '0'(20 cm in diameter). The third was comprised of three parallel lines (3 x 25 cmeach). The visual cues were placed at three of the four compass points aroundthe pool, with the platform being located between two of them. From trial totrial, the platform's location changed, but its relation to the visual cues wasconserved throughout testing.Distal Auditory CuesAt 3 of the 4 compass points around the pool, distinctive auditory cueswere placed. The first of these was a metronome set to produce 92 beats permin. The second was a sonalert providing a constant narrow-bandwidthsignal. The final cue was a digital alarm clock that produced trains of pulses.These cues were covered with a layer of foam to dampen their intensities to amoderate sound level. From the center of the pool, these cues produced stimuliof intensities 59, 58, and 59 dB HL (with C-weighting). These cues were locatedout of visible range for the rats, 60 cm above the floor of the testing room,beyond the rim of the pool. Curtains surrounding the pool reduced theinfluence of irrelevant visual cues to aid in the focussing of attention onauditory stimuli.^All stimuli were presented concurrently, and continuouslythroughout the testing session. The platform was located between the sametwo cues for any given rat during any given trial.RESULTSsurgical ManipulationThe LPS technique employed for inducing middle ear effusion readilybrought on acute symptomology in the young rat. Mucoid effusions wereapparent within 24 hrs of injection, and saline was drained from control earsby this time. Generally, the effusions persisted for 6 to 7 days, although the29mild effusions tended to wane somewhat more quickly (3 to 5 days), and weremore variable in longevity. With a consistent dosage, very few (4 of the 62prepared, or 6.5 %) surgeries failed, and few rats (4 of the 62) died as a result ofear infection symptomology. Further, only 2 effused rats became relativelysmall, or stunted with respect to other rats in the same, and other litters of thesame age. In addition, only 4 rats developed ruptures in the tympanicmembrane due to induced effusion. None was included in the study.In order to induce OM symptomology in the rats used in this study, muchhigher dosages of LPS solution were required than what was expected on thebasis of work performed by Ohashi et al., (1986), and Hellstrom (1986). Whiletheir adult rats only required a suspension of 1 ptg per p,1 to induce severesymptomology using K pneumoniae (Sigma), adult rats in our laboratoryrequired 10 1.t.g/p,1 to induce similar signs. These adult pilots appeared to havelow auditory sensitivity, and had dramatic problems with balance, indicative ofsemicircular canal involvement.^Young rats required an even higher dosagefor elicitation of OM symptoms (30 lig/R1 for severe conditions, and 14 Ii.g/g1 formild). Even still, OM sequelae were not as severe as those reported. Forexample, the death rate observed here was a fifth of that cited elsewhere foradult animals. Further, no rats were observed to be suffering from diarrhea,lack of appetite, dehydration, or fever; symptoms demonstrated in LPS animalsprepared previously (Ohashi, et al., 1988).Otoscopically, effusions induced bulging of the eardrum (particularlythe pars flaccida), and rendered them opaque and yellowish in color. Inaddition, circulation to the area was greatly increased, as indicated by thedilation of blood vessels there. Otitis Externa was not commonly observed30amongst all effused rats. The external ear canal did not fill with this mucoidfluid, but wax was generally built up over the eardrum, and had to be removedin order to inspect the tympanic membrane. Only one of the MBE ratsdemonstrated this symptom. Rats with infections administered only to one earappeared to be somewhat less likely to show Otitis Externa. None of the mildlyeffused rats in this group demonstrated signs of Otitis Externa, while 8 of theseverely effused ones did.Audiogram DataThreshold data collected from the rats during 3 representative phases ofeffusion history were used in the determination of threshold shiftsexperienced between the ages of 20 and 45 days. Longitudinal data spanningthe entire effusion treatment phase (days 12 to 45, approximately) could not bestudied due to several practical difficulties.^Firstly, the orienting responseselected to indicate detection of a stimulus was extinguished relatively quickly.In addition, rats did not elicit orienting responses reliably enough forthreshold estimation outside the age range of 20 to approximately 45 days. Forthese reasons, only data gathered between the ages of 20 and 45 days wereincluded in this study, and testing sessions were kept short (7 to 10 trials each)and minimal in number (about 3). Data examined were gathered at the heightof an effusion episode occuring between 20 and 25 days of age, just prior to thesubsequent injection of LPS (and therefore between effusions for MBE rats),and upon recovery (as determined by otoscopic examination) at about 45 daysof age.The otoscopic severity grading for the effusion state during whichthresholds were determined was 4 for SBE rats, 3 for MBE, and 0 for controls.31Thresholds were obtained for rats between effusions when severe effusionswere graded at 3 or 4, mild effusions were graded as 1, and controls were 0.After apparent recovery from effusion, rats were tested again, when theireardrums appeared to be normal (0). In all effused rats, however, there werewhite plaques visible on the tympanic membrane surface at the time oftesting. These did fade with time, but responding to acoustic stimuli ceased asthe animal aged past approximately 45 to 50 days. Data from all animals testedwere included in the formulation of audiograms (see Appendix 1 fordescriptive statistics).The data were subjected to a mixed model analysis of variance (ANOVA).Specifically, there was one, 3-level, between subjects factor (group), and two,3-level, within subjects factors (effusion state, and frequency).^Theassumptions underlying the univariate approach to repeated measures ANOVAwere upheld, as indicated by the nonsignificance of the test for sphericity,and of Bartlett's tests for homogeneity of variances. On this basis, theunivariate approach was adopted in favor of the less powerful multivariateapproach (see Appendix 2) (Harris, 1985).A significant main effect of group was obtained, F(2,12) = 60.61 p.< 0.01(see Appendix 2). The univariate repeated measures F-test for effusion state(during, between, and after effusion episodes) was also significant, F(2,24) =38.820, a< 0.01. The effect of frequency was also significant, F(2,24) = 293.762,p.< 0.01. Interaction effects were obtained for group by effusion state atsignificance (E(4,24) = 19.282, p.< 0.01), group by frequency (F(4,24) = 79.605, 11.<0.01), and effusion state by frequency (F(4,48) = 12.310, a< 0.01). An interaction32effect for group by effusion state by frequency was also significant (F(8,48) =9.933, p.< 0.01) (see Figure 2).33Figure 2. Depiction of the simple interaction between effusion state andfrequency, as they affect threshold for each group of animals. Simple simplemain effects of effusion state at each frequency can also be seen in the spreadof data points at each frequency. Note that "dB SPL" denotes decibels measuredrelative to sound pressure level.3410 200 0^200 10FREQUENCY (kHz)—0-- EFFUSION—A-- IN BEPNEENAFTER10 20FOR CONTROL ANIMALS807060Fg soa 401.1rri 30V)ad114 2051: 1003580 -.„,„ 70 -^ —0-- EFFUSIONINBETVVEENIN 60 AFTERFOR SEVERELY EFFUSED ANIMALS0FOR MILDLY EFFUSED ANIMALS—0-- EFFUSIONINBETVVEENAFTERPQ• 50a 40c2 30 -20111:1▪^1080700▪ 4 600:1 505 400:1 30g4 20tclF., 10Due to the presence of this three-way interaction, simple interactioneffects of effusion state and frequency were examined for each group in orderto begin addressing questions regarding the effects of these variables onthreshold for each group of rats (Howell, 1991; Winer, Brown, & Michels, 1991).Significance was obtained for the simple interaction of effusion state andfrequency for MBE (F(4,48) = 9.88, p.< 0.01), and SBE (F(4,48) = 20.70, a< 0.01) ratsonly (see Appendix 3 for summary of results from the analysis for these simpleinteraction effects).^The simple interaction of effusion state and frequencywas nonsignificant for SC animals (see Figure 2).For the significant simple interactions obtained (ie effusion state byfrequency for MBE and SBE), the possible existence of any simple simple maineffects was then determined (Howell, 1991; Winer, et al., 1991). A significanteffect of effusion state was obtained at 1 kHz (F(2,48) = 9.27, a < 0.01), and at 16kHz (F(2,48) = 20.73, a < 0.01) for MBE rats. This effect was not significant at 8kHz for this group (see Figure 2). Significance was also obtained for the maineffect of effusion state at 8 kHz (F(2,48) = 32.96, a < 0.01), and 16 kHz (F(2,48) =59.95, a < 0.01) for SBE rats (Figure 2). A significant main effect of effusionstate at 1 kHz was not obtained for this group. Control.animals showed nodifferences with effusion state at any frequency (Figure 2, and Appendix 4).It is evident that saline control (SC) rats demonstrated no differenceswith effusion state (or time, as it were) at any frequency. Mild bilaterallyeffused (MBE) rats were less able to hear acoustic stimuli of 1 kHz betweeneffusion states than during an effusion episode, or after recovery. They werealso less able to detect 16 kHz tones at the height of an effusion episode thanbetween injections, or after the recovery period.^Further, they were more3637sensitive after recovery than between effusions.^For severe bilaterallyeffused (SBE) rats, an acute effusion episode yielded higher thresholds at 8 kHzand 16 kHz than those observed after recovery. Thresholds determined at thesame time as those between effusions for MBE rats were also much higher thanthe post-infection thresholds at 8 kHz and 16 kHz. Effusion and "between"effusion thresholds did not differ significantly from each otherIn order to detect specific differences between effusion stages atdifferent frequencies for each group, post hoc comparisons were conducted onthe significant simple simple main effects. Because a number of post hoccomparisions were being performed, Tukey's HSD was employed as aconservative measure, controlling for increased family-wise error rate, andthe resultant high incidence of Type I errors. Results are summarized in Table1.During an acute effusion episode, there were significant differencesbetween all 3 groups' thresholds at 8 and 16 kHz. No effect of effusion state wasdetected at 1 kHz. At the 2 higher frequency stimuli, SC rats had the lowestthresholds, followed by MBE, and then SBE rats. These differences persisteduntil the subsequent injection of LPS (during the "between" phase). Resultsare less straightforward for group differences remaining after the recoveryperiod.Table 1. Summary of results from the Tukey's HSD tests performed on audiogramdata. Results for control, mildly effused, and severely effused rats areincluded.381 kHz 8 kHz 16 kHzSTATE DIFFERENCEDUELING SC-MBE * *SC-SBE - * *EFFUSIONMBE-SBE * *BETWEEN SC-MBE * *EPISODES SC-SBE * *MBE-SBE * * *SC-MBE * *AFTERSC-SBE * *RECOVERYMBE-SBE - *- nonsignificant result of Tukey's HSD test.* significant result of Tukey's HSD test (12 < 0.01).39SC animals had significantly lower thresholds than MBE rats at 8 kHz only, andwere less sensitive at 1 kHz. SC differed from SBE rats at 8 kHz (having lowerthresholds) and 16 kHz (higher thresholds) while the 2 experimental groupsdiffered only at 16 kHz.Open Field DataThe data consisted of the number of rears, and defecations (ie. fecalboli), the amount of time spent in each quadrant of the arena, and the numberof contacts made with each object. These data were collected during each ofthe two 8-minute testing sessions conducted 24 hours apart, from 60 rats.Session 2 differed from the first in that one object was exchanged for a newone, and another was changed in location within a quadrant, and inorientation. Each data class was dealt with separately, and subjected to a oneway repeated measures ANOVA to detect significant differences in behaviorbetween groups over sessions 1 and 2. Descriptive statistics for eachparameter examined are presented in Appendix 5.There were no group differences between sessions for the number ofdefecations or rears. With respect to the latter, rats made significantly fewerrears in the second test phase of the experiment than in the initial session(F(1,55) = 5.28, a < 0.05). For number of defecations, rats left more fecal bolibehind in phase 1 than 2 (F(1,55) = 6.15, a < 0.05). Further, there were nodifferences in the amount of time spent in each quadrant of the testing arena.All rats appeared to spend similar amounts of time in each quadrant duringphases 1 and 2. See Appendix 5 for descriptive statistics.Contacts made with each object during phases 1 and 2 did yield somedifferences between groups. A main effect of group was found for the object40that was changed in orientation between phases 1 and 2 (F(4,45) = 4.37, a <0.01). Tukey's HSD test indicated that this difference was found betweencontrol rats and the 2 mildly effused groups; SC rats making an average of 3more contacts than MME or MBE rats. No differences were noted betweensessions 1 and 2 for this object for all rats. The object that was replaced inphase 2 with a novel one also yielded differences (F(4,45) = 2.63, a < 0.05). Witha Tukey's HSD test, it was found that this difference was between controlanimals and MME rats; the former making 5 more contacts than severelyeffused animals, on average. Again, no changes were noted between the twotest sessions. Descriptive statistics are presented in Appendix 5.Visual Water Maze DataThe data for the visual water maze task were averaged over 8consecutive trials so that a daily average escape latency was obtained for eachof the 50 rats tested (see Appendix 6) (Sutherland & Dyck, 1984). Although thisinvolved trials where rats were released from different points, and theposition of the platform and cue arrangement was different, these werecounterbalanced within a day. These data were then analyzed using a mixedmodel repeated measures ANOVA, with one 5-level, between-subjects factor,group and one 10-level, within-subjects factor, day.^The multivariateapproach to analysis of the within subjects factor was adopted for this analysissince Mauchly's test for sphericity was significant (W=0.00233; a < 0.01) Thisindicates that the assumption of sphericity of the variance/covariancematrices underlying the univariate approach was not met (Tabachnik & Fidell,1983; Winer, et al., 1991).41All five groups of rats acquired the visuospatial water maze task readily(see Figure 3). Further, the results of the ANOVA indicated that there were nosignificant group differences (7.1 2 = 0.135) in average performance for thevisuospatial maze (see Appendix 7). However, it was found that, on average, allrats improved in maze performance over the training period (L = 0.479, a <0.01. 1 . This was evident in the decreased average escape latencies achievedtoward the end of testing, relative to the beginning and confirmed by thesignificance of the multivariate test of the repeated measure for day (2 =0.3 8 8).There was no significant group by day interaction (,1 2 = 0.211). An ANOVAsummary table for this analysis is presented in Appendix 7. To determine ifrats were different in performance from controls at either the beginning orthe end of training, t-tests were performed between groups on days 2„ 5, and10, while protecting the a - level. No significant differences were found.421 Wilks' A was employed in this and all subsequent multivariate analyses asrecommended by Tabachnik, and Fidell, (1981).Figure 3. Mean daily escape latencies for each group of rats over ten days oftesting in the visual spatial water maze task.430^2^4^6^8^10^1250 --Ar-- sc—4)--- MME--0--- MBE--*-- SME---1,— SBE^I403020100DAYAuditory Water Maze DataAs with visual water maze data, daily averages were obtained and thenanalyzed for group differences (see Appendix 8 for summary descriptivestatistics). As with the visual modality, the data were subjected to a mixedmodel repeated measures ANOVA, with one 5 level between subjects factor (i.e.,group) and one 25 level within-subjects factor (i.e., day). For this analysis, theassumption of sphericity required for adoption of the univariate approach torepeated measures ANOVA was not met (W = 1.35 E - 10 , a <0.01). Thus, themultivariate approach was adopted for the analysis of the within-subjectsfactor (Harris, 1985).There was no significant effect of group in the performance of rats inthis water maze task (q 2 = 0.071). All the animals learned to escape the wateron the basis of auditory cues available, as indicated by decreased dailyaverages of escape latencies (see Figure 4). All rats did improve significantlyin average maze performance over days, a = 0.17, a < 0.01). The group by dayinteraction was found to be nonsignificant (1 2 = 0.415). An ANOVA summarytable for the analysis of the auditory data appears in Appendix 9. Again,results from days 2, 13, and 25 were subjected to t-tests.^No differences werefound.45Figure 4. Mean daily escape latencies for each group of rats, over the twenty-fivedays of testing in the auditory spatial water maze task. Data from all groups isincluded.46DAYSCMMEMBESMESBEDISCUSSIONSuccess of the Surgical ManipulationOn the basis of otoscopic criteria, the manipulation employing LPSappears to be effective, inducing symptomology similar to that experiencedwith naturally occuring OM. The mucoid effusion appears to completely fillthe tubotympanum, producing a bulge and decreased translucency in theeardrum.^This bulge was particularly pronounced with the pars flaccida,consistent with other reports attempting to establish an animal model of OM(Stenfors, Hellstrom, & Salen, 1982). Accompanying this is increasedcirculation to the area, and the appearance of wax in the external ear canal,with more extreme cases of the OM model syndrome.The use of LPS infusion presents several advantages over viablebacteria. Firstly, the death rate of subjects undergoing LPS proceduresappears to be much lower than that encountered with viable bacteria^Pilotstudies conducted in this laboratory to find an optimal dosage of LPS, resultedin only 15 % (6) of the animals dying subsequent to the infusion procedurewithin 24 hours of surgery. Presumably, some of these may have been due toprolonged experience of halothane anaesthesia (up to 1 hr per injection, insome cases). Other reports using viable bacteria state loss rates of up to 67 %(Marrow, Whitley, & Fulgam, 1988). Although the literature cited employed theadult rat preparation rather than pups, the reduced death rate likely indicatesthat the LPS model more closely resembles the human condition in terms ofseverity, as descriptions of OM obtained in the bacterial model resemble onlythe most serious cases of acute OM. Generally, effused rats appear to be48relatively free of other OM sequelae noted in literature that describes animalmodels of OM (for example, dehydration, diarrhea, and dramatic decreases inappetite). Whether or not these symptoms are present in animals used herebut to a lesser extent, such that they are not easily observable is unknown.However, the animal models previously used are extreme in relation to thehuman condition, involving more severe symptoms, and the probable additionof new ones in worse infections. As with naturally occuring OM, it is verylikely that the less extreme manipulation forwarded here involves some of themild symptoms, but do not impact overall health and development of theanimal. On this basis, effects noted in adult testing are held to be attributableto the ear infection itself, and not to the experience of being ill for themajority of auditory development.The lack of severe symptomology is interesting since the dosagerequired for even the mildest of effusive effects was upwards of 200 times thatcited in other experiments (e.g. Hellstrom, et al., 1988; Ohashi, et al., 1988).Although adult rats in our laboratory were also more difficult to induceeffusions with, the very young rats' dosages are remarkably high. Becausethe source of LPS and the surgical protocol were the same across theseexperiments, the reason for the difference is likely with the rats themselves.At the ages during which effusion was experienced in this study, all animalswere still nursing, and were apparently not weaned by their mother until theend of the effusion period. This could play a very important role in dosedetermination, at least with the initial infusions. The presence of clostrum inthe milk (for at least 24 hours after birth) serves to greatly enhance theyoung rat's immunity to common pathogens. As a result, the mucosal lining of49the tubotympanum is very likely to be more robust with the appearance of LPSin the cavity. This situation may be exacerbated by the reproduction ofpathogens in the middle ear.^In natural cases, and most animal models of OM,the disease process involves the introduction of viable bacteria, which canquickly increase in number, partly compensating for any quick immunereaction. With LPS, the maximal effect is constant over time, and is determinedby the concentration and volume of solution injected. As a result, the LPSsituation would be much easier to remedy, and symptoms experienced milderin comparison to other forms of OM.Relation of Effusion to Hearing Level ReductionThe ear effusion produced by LPS injection lasts for approximately 6days in severely effused rats, clearing from the middle ear relatively suddenly.Up until this point, the severity of the effusion appears to be relativelyconsistent in terms of otoscopic criteria, being graded as a 4 to within a day ofevident clearing.^During this time, severely effused rats experiencedreductions in hearing sensitivity of approximately 32 dB SPL at 8 kHz, and 29dB at 16 kHz, compared to control animals. Approximately 5 days later, at thepoint when mildly effused rats were reinjected, SBE rats were still harder ofhearing by 28 and 27 dB SPL at 8 and 16 kHz, respectively. This represents adecrease in attenuation of approximately 3 dB SPL over a period of almost aweek. As these rats were reinfused with LPS just subsequent to their secondaudiogram session, the LPS protocol followed here provided the successfulmanipulation of hearing threshold in a chronic manner. That is, SBE ratswere subjected to an auditory deprivational experience of approximately 29 dBSPL (depending on frequency) from 12 to about 45 days of age.50The effusions produced in mildly infected rats were not so predictable.Not only did they appear to vary more in extremity than rats effused usinghigher doses, but the length of time that the effusion persisted variedsomewhat more too. This is problematic in the present context, since the effectof effusion was more difficult to control. As a result, hearing thresholdsestimated for MBE rats were more difficult to interpret, and contributions ofeffusion to possible deficits in adult functioning more elusive.^During theheight of an effusion, they required 8 kHz stimuli presented at an intensity 10dB SPL higher than controls did. For 16 kHz, they needed an additional 18 dB.Just prior to reinfusion of LPS, MBE thresholds were still 13 and 8 dB SPL(respectively) higher than controls.^At the same time, their eardrums hadbeen judged as being at or very near normal, these remaining discrepanciesare of potential research interest.^They indicate that manipulating hearingsensitivity in the short term is problematic for the LPS model. During the"between" infusion phase, mildly effused rats regained only 10 dB SPL of theirnormal hearing level at 16 kHz. At 8 kHz, these animals were 7 dB lesssensitive than at the height of effusion. Since audiograms were not obtaineduntil the eardrums of these rats appeared normal upon otoscopic exam, itseems that the hearing reduction persists beyond obvious signs of effusion. Asa result, it may be difficult to produce a truly fluctuating condition of effusionin the rat. It is first necessary for threshold estimates to be obtainedperiodically throughout the period of effusion. In addition, it may benecessary that an animal with a slower time course of auditory development beemployed. The observation that the dB reduction persists for more than acouple of days, and the belief that basic auditory abilities develop quite rapidly51in the rat suggest that even one mild effusion episode may in fact last formuch of the rat's auditory development.It should be noted that threshold changes obtained primarily reflect thepresence of the effusion, and not ongoing developmental processes.Frequency thresholds of the normally developing rat are reported to becomparable to that of the adult rat's by about postnatal day 20 (Crowley, &Hepp-Reymond, 1966), and present threshold data collection did not beginuntil the animals were 20 days old. Moreover, controls' thresholds did notchange from the time of effusion measures to the time that recoverythresholds were determined. As a result, data gathered when animals were 40to 45 days old after recovery, are not likely a reflection of lower thresholds dueto maturation of the auditory nervous system.Methodological Problems in Determining ThresholdsThe fact that rats did not respond to acoustic stimuli in the orientationparadigm after the age of 45 days presents a serious problem for establishingreversibility of the LPS manipulation. The data obtained from severely effusedrats, along with the small variances observed within each cell (see Appendix1) tend to support the reversibility of the effusion, but experimental ratthresholds did not consistently reach the level of controls by 45 days of age. Atthis time, SBE rat thresholds were 5 dB lower than the control rat thresholds at8 kHz. At 16 kHz, SBE rats were 7 dB less sensitive than SC rats. Similarly, MBErats were 11 dB more sensitive than controls at 8 kHz, and were not differentfrom them at 16 kHz.However, both of the effused groups' thresholds decreased markedlyfrom measures taken during the effusion phases. SBE rats attained thresholds52of 35 dB SPL for the 2 higher frequencies examined after the recovery period;a gain of 33 and 22 dB SPL for 8 and 16 kHz, respectively. Mildly effused ratsgained 15 dB SPL in sensitivity for 8 and 16 kHz from the height of theireffusion to the day of reinfusion. These observations provide an indicationthat the LPS protocol for both severe chronic and mild fluctuating effusions isreversible, but on a larger time scale than is optimal for the rat preparation.Perhaps with better characterization of the dB loss associated with mildeffusions, one could be able to produce effusions of different severities thatwould result in the experience of different dB reductions in hearing. Further,with more information regarding the time course of effusions, one should beable to produce hearing decrements of a given level for predictable amountsof time.Clearly, the adoption of alternative behavioral testing methods isnecessary for testing hearing thresholds in older rats.^Some promisingavenues present themselves in studies employing the effects of long-termdeprivation. In the past, methods such as conditioned suppression have beenused to estimate acoustic thresholds in rats raised with earplugs, and withenvironmental silencing.^In conditioned suppression, the rat is placed in adeprivational state (e.g. water deprivation).^During a testing session, the rat isallowed access to the restricted resource for some period of time. While the ratis engaged in consummative behavior (e.g. drinking) an acoustic stimulus ispresented.^Interruption of drinking behavior is interpreted as a positiveresponse, indicating that the animal heard the sound (Heffner, 1980; Kelly, &Masterton, 1977; Heffner, & Masterton, 1980). In running several trials of53differing intensities during a session, one should be able to obtain thresholdestimates.This method seems promising, but as with other protocols using positivereinforcement practical problems exist, where rats adapt to the presentationof stimuli fairly quickly, and cease responding to stimuli they hear. Inaddition, the motivated behavior stops as the effect of resource deprivationwanes. These necessitate the employment of very short testing sessions, andadequate spacing of sessions (Heffner, 1980; Kelly,& Masterton, 1977).Another possibility that may circumvent adaptation difficulties utilizesthe conditioned avoidance response, where a tone predicts the presentation ofa shock stimulus to the grid that the animal is standing on. To avoid the shock,the animal may move to the other (safe) grid of a two-sectioned chamber.Trials where the tone was not heard, result in the delivery of a shock to theanimal (Ryan, 1976). This technique is advantageous to other forms of learnedbehavior because it requires very little shaping or training, and maintains anextremely high level of performance over time. In addition, it is very robustto treatments administered to the animals (Ryan, 1976). All three of thesecharacteristics are optimal in methods determining an animal's threshold.With them, one is able to employ the method of limits for estimating hearingsensitivity. This technique allows for a high level of confidence in thresholdsobtained in that it involves several trials run at near threshold intensities.However, this entails the use of many trials within a given session. As a result,it is ideal to have a protocol employing a technique that is resistant toexhaustion within a test session, and over many of them. Conditioned shock54avoidance appears to satisfy these requirements, and therefore shows promisein characterizing effusion severity in older animals.An alternative to behavioral testing of hearing level presents itself inthe monitoring of brainstem recordings, and tracing of the animal's EEG.Through the analysis of EEG recordings obtained, one may monitor thesensitivity to auditory stimuli (Buser, & Imbert, 1992). These techniques arevery attractive, as they provide indices of absolute threshold, rather than arelative index that relies on a compliance of sorts, from animals being studied.However, brainstem recordings are also difficult (if not impossible) to obtainin the rat because the head is too small to allow for sufficient spacing of scalpelectrodes over the brainstem area. This method would have to be employedwith larger animals such as the chinchilla in order to be useful.Although it will take one of these techniques to verify, it is notunreasonable to suppose that some extraneous reversibility problems exist.^Anexample is tympanosclerosis.^Here, changes primarily include an increase incollagenous fibers, and damage to hyaline tissue. In more severe cases,calcification can also occur (Kuijpers, Wielinga, Tonnaer, & Jap, 1988).Although this has yet to be demonstrated in the LPS model, it is commonlynoted in the viable bacterial models (Widemar, Hellstrom, Nordling, &Johansson, 1988). Effects stemming form tympanosclerosis would likely besomewhat frequency-specific, as it effects different areas of the tympanicmembrane (Buser, & Imbert, 1992; Yost, & Nielsen, 1985). An overall thresholdshift may also occur, particularly if calcification occurs, interfering withossicular vibration. This would impede the the transmission of acousticinformation, regardless of frequency.55Ototoxicity is another probable source of difficulty. It too, is noted toaccompany middle ear effusion (Goycoolea, et al., 1980). It is believed that theround window membrane is or becomes permeable to the immunoreactiveagents produced in reaction to antigenic irritation of the mucosal membranein the middle ear. The bacteria causing the disorder are too large to pass thisbarrier. However, there is no evidence indicating whether or not LPS derivedfrom the fractionation of the bacterial membrane can pass. Further, theimmunoreaction would be similar to that obtained in a natural infection. Theeffect of any of these factors on inner ear functioning can be quite marked.This would include the death of hair cells responsible for sensory transductionof particular frequencies, in turn resulting in decreased sensitivity to thesetones.^In addition, frequency and intensity discrimination would bedetrimentally effected.^However, this possibility cannot be addressed with thepresent data, as such discriminatory abilities were not examined, andthresholds did not return to the baseline expected on the basis of performanceby controls.Open Field Exploration Turning to preliminary results obtained during the adulthood of OMrats, one finds that, as with the clinical literature, long-term effects of OM aresubtle and elusive, if existent at all. Data gathered in the open field indicatethat effused rats (whether mild or severe, monolateral or bilateral) are normalin their emotionality, reaction to novelty, and activity levels as control rats.This is reflected in the general lack of differences found between groups fornumber of defecations, rears and contacts with objects present in the arenaduring the initial testing phase. The absence of differences in their behavior56with habituation and a repeated experience in the field indicates that they arenot more or less able than controls to remember the field situation over aperiod of 24 hours. In addition, with the manipulations involving the objectspresent, their reactions to novelty are not deviant from controls. They did notspend more or less time with the new object that was introduced, replacing onepresent during phase 1. Nor did they appear to change their interest with anobject present in the initial phase, that had been rotated within its quadrantfor the second session. On this basis, OM did not have a dramatic effect ongross indices of activity, reaction to novelty and emotionality. Were it truethat the effusion is accompanied by a more general sequela of symptoms, onewould fully expect that these behavioral indices could be detrimentallyeffected such that they would be more apprehensive in the open field, andspend less time interacting with objects.Visual Spatial Water Maze As was predicted at the outset of these experiments, the ability tonegotiate a water-based spatial maze, relying on visual cues was notsignificantly deficient given the history of an OM-like syndrome of any type.On the basis of escape latency, all rats acquired the task successfully withinthe ten- day period usually allowed for completion (e.g. Sutherland, & Dyck,1984). The deprivational experience of 10 to 30 dB SPL (depending on groupand frequency) is not expected to produce deficits in the ability to form andutilize spatial maps of the external environment using visual cues. Thedeprivation incurred was purely auditory, and therefore should not havepossessed carry-over effects into the visual modality.57This claim is supported in part, by the investigation of the role of distalvisual and auditory cues in place navigation, conducted by Sutherland, & Dyck(1984).^In this study, experimenters examined the effect of visual deprivationon performance in a visual and an auditory spatial water maze. Bilateral eyeenucleation was conducted on neonatal rats, and on adults. Both of thesegroups of rats were much slower than controls in escaping the water, neverreaching the asymptotic level that controls did. However, these animals stilldemonstrated significant decreases in latency with training.^Upon probetesting for search strategies employed in finding the escape platform, it wasfound that the enucleated rats had adopted an alternative, (likely nonspatialand nonvisual strategy) for finding the platform.Auditory Spatial Water MazeWith testing in the auditory modality, it was found that normally sightedrats were able to use auditory beacons in the absence of visual ones, in solvingthe maze.^Further, enucleated rats elicited longer escape latencies during theprobe trials where the platform-cue relation was changed. This indicates thatthey were able to utilize the auditory cues effectively in establishing a spatialstrategy to solve the task. Indeed, this group was found to be superior tocontrols in their abilities to acquire spatial navigation strategies on the basisof auditory cues (Sutherland, & Dyck, 1984). Although this does not involve areversible manipulation of vision, the findings are never the less compelling.Even without the use of vision for the entirety of development and adulthood,these rats have preserved auditory spatial competence, in the absence of anability to solve the maze using visual cues.58Given the findings of Sutherland, and Dyck, (1984) one would expectthat depriving animals of auditory experience (even if reversible) shouldreveal some differences in auditory maze performance compared to controls.In analogy to visual deprivation, there should be deficits in spatial navigationon the basis of auditory cues, but not when using visual cues. This keyhypothesis was not supported by the data collected in the presentinvestigation. All effused rats were as competent as controls in solving thetask, and acquired asymptotic escape latencies that were as short as those ofcontrol animals. Further, they did not demonstrate differences in the second,or last days of training. Thus, it was not a matter that effused rats wereinitially deficient in solving the task, but later compensated for theirinability. Nor was it the case that differences did not appear until later intesting, when asymptotic levels were reached.Methodological Problems with the Water MazeThere are a number of possible methodological reasons that differenceswere not found. Firstly, Sutherland, and Dyck (1984) completely enclosed thewater maze using walls above the pool rim, and a roof. It was arranged so thatthe movements of the rats were still observeable through a piece of solar filmin the ceiling, but the rats could not see out of the apparatus. These visualbarriers were devoid of contours that could serve to distract the rats fromattending to the desired auditory cues. The platform was kept in one locationof the pool, which was signaled by the auditory beacons. Rat release pointswere varied randomly. In the present study, curtains were erected around thepool with the aim of reducing visual cues available. It is possible that this wasnot a sufficient measure, and rats were using visual stimuli that are not59readily apparent to the investigator. The situation could be that it is verydifficult for a rat to resist the use of visual cues in solving spatial tasks. As aresult, all rats (whether effused or not) persist in the attempt to solve the mazeusing visual cues. If this were so, one would not expect to find differencesbetween the groups examined. One possible solution to this problem is to blindall rats performing the auditory task so that visual cues are completelyinaccessible. Since enucleation would not allow rats to perform in the visualversion of the task, and testing rats in the dark would not permit the trackingof their movements in the pool, the preferred technique for this would likelybe through the use of goggles.Conclusions and RecommendationsClearly, one can not refute the existence of a long-term effect of OM onspatial competences. The spatial version of the water maze task is renownedfor its use in studies where effect sizes are apparently large (for example, thedramatic loss of spatial ability associated with hippocampal damage), and isvery useful in allowing for the examination of strategies being adopted byspatially deficient rats (Sutherland, & Dyck, 1984; Sutherland, Whishaw, &Kolb, 1980; Sutherland, Whishaw, & Kolb, 1982). However, the water maze maynot be a sufficiently sensitive medium for the elucidation of spatial deficitsassociated with middle ear effusion history. The effect size of the auditorydeprivational experience turned out to be much smaller than was expected forsuch a manipulation, and group differences were not found in the acquisitionof a spatial navigational escape response using this paradigm. Much care mustbe taken in choosing behavioral approaches as sensitivity to deficientperformance is crucial.60There are several other spatial tasks that show promise in this respect.These may well provide more precise measures of performance (for example,the scoring of errors, rather than latencies), and thereby prove to be moresensitive to OM effects. Further the successful completion of these tasks maybe more difficult such that the demands entailed are at or near the limits ofability for control rats. One of these paradigms is the spatial version of theradial arm maze. This paradigm would presumably be run in a manneranalogous to the water maze, but the dependent measures taken would includethe number of errors (previously baited arms entered a second time) made in aparticular trial. This task has previously been demonstrated to draw outdeficits in spatial ability of dark-reared rats (Tees, Midgley, & Nesbitt, 1981), aswell as with other somewhat subtle manipulations (e.g. Meck, Smith, &Williams, 1988).Another approach that has been alluded to in the review of humanliterature, is to test animals in a noisy environment. This does not involve achange in paradigm, but perhaps a noisy version of the water maze task couldprove to be more demanding. A rat would have to separate relevant auditorycues from the background, or mask, and then solve the problem on the basis ofthese cues. If rats need to pay more attention dissecting relevant stimuli froma backdrop of noise, they will presumably be less able to solve the taskefficiently, and will therefore take longer to master the maze. If effused ratshave a less stable, or reliable auditory signal to begin with, picking outappropriate acoustic stimuli in noisy conditions should prove to be even moredifficult than with normals, and would therefore demand more of theirattention.61Another aspect of spatial perception that bears investigation is soundlocalization. This ability is more psychophysical in nature, than thosepreviously discussed, and may well demonstrate differences between effusedrats and controls as there is likely less opportunity for compensationalbehaviors.^Accurate localization abilities rely on the capability of relatingmonolateral, and interaural difference cues in both time and intensity topinpoint a sound in space (Blauert, 1982; Kelly, & Potash, 1986; Oldfield, &Parker, 1984). On the basis of these cues, the adult rat is normally able tolocalize a sound source along the horizontal azimuth to within about 10 or 12de.g. across the midline, and more than approximately 60 de.g. within ahemifield (Kavanaugh, & Kelly, 1986; Knudsen, 1984).The neural system mediating this ability is very sensitive to changes indevelopment. As the animal matures, localization behavior becomes muchmore refined as the head increases in size, the ears reach adult structure, andas the cochlea and central auditory pathways mature. This requires thefrequent recalibration of the localization substrate, while still providingnormal localization abilities in the adult (King, & Moore, 1991). Physiologicalchanges appear when binaural interaction is prevented in youth via thedeprivation of one ear (asymmetrical deprivation).^Neural firing latencyincreases that ensue on the deprived side dramatically effect binaural time-difference and intensity-difference encoding in a detrimental manner(Clopton, & Silverman, 1978). In addition, inhibition of activity in the inferiorcolliculus units ipsilateral to the stimulus is weak or absent in the deprivedstructure. This would suggest a competitive process underlying thematuration of binaural interaction.^Thus, distortion of binaural interaction62would severely disrupt localization abilities at all frequency bandwidths(Clopton, & Silverman, 1978).One would suspect that early unilateral experience with OM woulddisrupt normal development of the competence, and this would be reflected indeficient behavioral perfomance of a sound localization task.^Indeed, inguinea pigs raised for three weeks with unilateral ear blocks and then testedwithout the plugs, localization is dramatically disrupted for at least 21 days(Clements, & Kelly, 1978). In the clinical literature, there has been a furtherindication that this relationship does occur with unilateral OM experience.During an effusion episode, human infants were less able to localize anacoustic stimulus than after recovery (Morrongiello, 1989). Were this effusionto persist, or recurr for a large part of the child's auditory developmentpermanent effects may well ensue. Clearly, this aspect of spatial auditoryfunction deserves some attention in the near future.Should one of these other approaches yield data supporting theexistence of long-term OM effects, one may begin to characterize the relationof OM, later behavioral competences and their neural substrate.To conclude, the LPS developmental model established provides muchpromise as a powerful tool for the experimental study of OM. It has beendemonstrated to produce hearing decrements of 10 to 30 dB SPL depending onfrequency, and this is related to the presence of middle ear effusion. Withfurther characterization of the time course and reversibility of the technique,one may be able to produce deprivational states of varying severities anddurations.^This model already possesses several advantages over other possiblemodes of Eustachian tube obstruction in that it does not involve mechanical63plugging of the tube lumen which is very difficult in the neonatal animalpreparation. It also does not entail the use of viable OM pathogens which couldwell induce a sequela of symptoms unrelated to auditory function. Possibleavenues of research for the future include the utilization of alternativebehavioral paradigms in determining the development of frequencysensitivity, and the reversibility of the LPS manipulation, as well as theinvestigation of possible deficits in spatial navigation on the basis of distalauditory cues.^In addition, psychophysical study of frequency, intensity, andduration discrimination are recommended as well.64REFERENCESBess, F. (1986). Audiometric approaches used in the identification of middle eardisease in children. In J. F. Kavanagh (Ed)., Otitis Media and ChildDevelopment, York Press: Maryland, .pp.70-82.Bishop, D., & Edmundson, A. (1986). Is otitis media a major cause of specificdevelopmental language disorders? British Journal of Disorders ofCommunication, 21, 321-338.Blauert, J. (1982). Binaural localization. Scandinavian Audiology, 15(Suppl.),7-26.Bluestone, C.D., Klein, J.0., Paradise, J.L., Eichenwald, H., Bess, F.H., Downs, M.P.,Green, M., Berko-Gleason, J., Ventry, I.M., Gray, S.W., McWilliams, B.J., &Gates, G.A. (1983). Workshop on effects of Otitis Media on the child.Pediatrics 21(4), 639-652.Buser, P., & Imbert, M. (1992). Audition. MIT Press: Cambridge, Mass.Clarkson, R., Eimas, P., & Marean, G. (1989). Speech perception in childrenwith histories of recurrent otitis media. J. Acoustical Soc. of America,$5.(2), 926-933.Clements, M., & Kelly, J. B. (1978). Auditory spatial responses of young guineapigs (Cavia porcellus) during and after ear blocking. J. of Comp. andPhysiol. Psychology. 2_2(D, 34-44.Clopton, B. M. (1980). Neurophysiology of auditory deprivation. Birth Defects:original article series., 1_€(1), 271-288.Clopton, B., & Silverman, M. (1978). Plasticity of binaural interaction. II.Critical period and changes in midline response. Journal ofNeurophysiology, 4Q(6), 1275-1280.Crowley, D. & Hepp-Reymond, M. ( 1966). Development of cochlear function inthe ear of the infant rat. Journal of Comparative and Physiological Psychology, 51(3), 427-432.Deyo, R., Straube, K., Moyer, J., & Disterhoft, J. (1990). Nimodipine amelioratesaging-related changes in open-field behaviors of the rabbit.Experimental Aging Research, il(4), 169-175.Dobie, R. A., & Berlin, C. I. (1979). Influence of otitis media on hearing anddevelopment. Annals of Otology. Rhinology. and Laryngology,60(Suppl.), 48-53.65Doyle, W. (1989). Animal models of otitis media: other pathogens. Pediatr. Infect Dis J., li, S45-S47.Ennaceur, A., Cavoy, A., Costa, J., & Delacour, J. (1989). A new one-trial test forneurobiologcal studies of memory in rats. II: effects of piracetam andpramiracetam. Brain Research, 33., 197-207.Feldman, H., & Gelman, R. (1986). Otitis media and cognitive development:Theoretical perspectives. In J. F. Kavanagh (Ed)., Otitis Media and ChildDevelopment, York Press: Maryland, .pp.27-41.Fellows, B. (1967). Chance stimulus sequences for discrimination tasks.Psychological Bulletin, 5.j(2), 87-92.Francis, R. (1979). The preyer reflex audiogram of several rodents, and itsrelation to the "absolute" audiogram in the rat. The Journal of AuditoryResearch, 12, 217-233.Friel-Patti, S. (1990). Otitis media with effusion and the development oflanguage: A review of the evidence. Topics in Language Disorders,fl(1), 11-22.Friel-Patti, S., & Finitzo, T. (1990). Language learning in a prospective study ofotitis media with effusion in the first two years of life. Journal ofSpeech and Hearing Research, 31, 188-194.Friel-Patti, S., Finitzo, T., Meyerhoff, W., & Hieber, J. P. (1986). Speech-language learning and early middle ear disease: A procedural report.In J. F. Kavanagh (Ed)., Otitis Media and Child Development, York Press:Maryland, .pp.129-137.Gellermann, L. (1933). Chance orders of alternating stimuli in visualdiscrimination experiments. Journal of Genetic Psychology, 4/, 206-208.Giebink, G. S. (1986). Prevention and medical treatment of Otitis Media. In J. F.Kavanagh (Ed)., Otitis Media and Child Development, York Press:Maryland, .pp.176-182.Glasberg, B., & Moore, B. (1986). Psychoacoustic abilities of subjects withunilateral and bilateral cochlear hearing impairments and theirrelationship to the ability to understand speech. S c andin avi anAudiology, 32, 1-25.Goldstein, M. (1967). A method for constructing sequences in contingentdiscrimination designs. P.ychological Bulletin, 5_2(5), 346-348.Goycoolea, M. Paparella, M., Goldberg, B., Schlievert, P., & Carpenter, A. (1980).Permeability of the middle ear to staphylococcal pyrogenic exotoxin in66otitis media. International Journal of Pediatric Otorhinolaryngology, 1,301-308.Harris, R. J. (1985). A Primer of Multivariate Statistics. Academic Press, NewYork, N. Y.Heffner, H. (1978). Effect of auditory cortex ablation on localization anddiscrimination of brief sounds. Journal of Neurophysiology, 41(4), 963-976.Heffner, H., & Masterton, B. (1980). Hearing in glires: domestic rabbit, cottonrat, feral house mouse, and kangaroo rat. J. Acoust. Soc. Am., 6_13.(6),1584-1599.Hellstrom, S., Cerne, A., & Stenfors, L. (1984). The attic space--its developmentand some species differences. Acta Otolaryngol (Stockh), 414, 31-33.Hellstrom, S., Hermansson, A., Johansson, U. & Prellner, K. (1988).Experimentally induced mucoid effusion in rat middle ear--a completemodel for otitis media research? In D. Lim, C. Bluestone, J. Klein, and J.Nelson (Eds.), Recent Advances in Otitis Media, B.C. Decker Inc.: Toronto,pp. 462-464.Hellstrom, S., Salen, B., & Stenfors, L. (1982). The site of initial production andtransport of effusion materials in otitis media serosa. Acta Otolaryngol.(Stockh) (93), 435-440.Hellstrom, S., Salen, B., Stenfors, L., & Soderberg, 0. (1982). Appearance ofeffusion material in the attic space correlated to an impaired eustachiantube function. Acta Otolaryngol. (Stockh) (93),Hermansson, A., Emgard, P., Prellner, K., & Hellstrom, S. (1988). A rat model forpneumococcal otitis media. American Journal of Otolaryngology, 2.(3),97-101.Howell, D. C. (1991). Statistical Methods for Psychology, (2nd ed.) DuxburyPress. Boston, Mass.Hyson, R. & Rudy, J. (1984). Ontogenesis of learning: II. Variation in the rat'sreflexive and learned responses to acoustic stimulation. DevelopmentalPsychobiology, 17(3), 263-283.Jenkins, J. (1986).^Cognitive development in children with recurrent otitismedia: Where do we stand? In J. F. Kavanagh (Ed)., Otitis Media andChild Development, York Press: Maryland, .pp.211-221.Jung, T., Hwang, S., Poole, D., Olson, D. Miller, S. Lee, J., Yoon, T., & Juhn, S.(1988). New animal models of otitis media in chinchillas. In D. Lim, C.67Bluestone, J. Klein, and J. Nelson (Eds.), Recent Advances in Otitis Media,B.C. Decker Inc.: Toronto, pp. 450-453.Karma, P., Penttila, M., Sipila, M., & Timonen, M. (1987). Diagnostic value ofotoscopic signs in acute otitis media. In D. Lim, C. Bluestone, J. Klein,and J. Nelson (Eds.), Recent i i Media, B.C. Decker Inc.:Toronto, pp. 44-49.Kavanaugh, G. L., & Kelly, J. B. (1986). Midline and lateral field soundlocalization in the albino rat (Rattus norvegicus). Behavioral Neuroscience 100(2.), 200-205.Kelly, J. & Masterton, B. (1977). Auditory sensitivity of the albino rat. Journalof Comparative and Physiological Psychology, 21(4), 930-936.Kelly, J., & Glazier, S. (1978). Auditory cortex lesions and discrimination ofspatial location by the rat. Brain Research, 145, 315-321.Kelly, J. & Potash, M. (1986). Directional responses to sounds in young gerbils(Meriones unguiculatus). Journal of Comparative Psychology, 100(1),37-45.King, A. J., & Moore, D. R. (1991). Plasticity of auditory maps in the brain.T.I.N.S. 14(1), 31-37.Klein, J. (1986). Risk factors for otitis media in children. In J. F. Kavanagh(Ed)., Otitis Media and Child Development, York Press: Maryland, .pp.45-51Knudsen, E. I. (1984). The role of auditory experience in the development andmaintenance of sound localization. T.I.N.S. 1.Sept.), 326-330.Kuijpers, W., & van der Beek, M., (1984). The role of microorganisms inexperimental eustachian tube obstruction.^Acta Otol aryngol (S t o ckh )(Suppl 414), 58-66.Kuijpers, W., van der Beek, M., & Willart, E. (1979). The effect of experimentaltubal obstruction on the middle ear. Acta Otolaryngol,  132_, 345-352.Kuijpers, W., Wielinga, E., Tonnaer, E. & Jap, P. (1988). Experimentally inducedtympanosclerosis. (1987). In D. Lim, C. Bluestone, J. Klein, and J. Nelson(Eds.), Recent Advances in Otitis Media, B.C. Decker Inc.: Toronto, pp.468-471.Kurono, Y., Tomonaga, K., & Mogi, G. (1988). Bacteriology in otitis media witheffusion: are staphylococcus aureus and staphylococcus epidermidiscausative agents? In D. Lim, C. Bluestone, J. Klein, and J. Nelson (Eds.),Recent Advances in Otitis Media, B.C. Decker Inc.: Toronto, pp. 338-340.68Kyle, J. (1978). The study of auditory deprivation from birth. British Journalof Audiology, 12, 37-39.Landis, L. (1990). Comment on "Otitis media in early childhood and itsrelationship to later phonological development".^Journal of Speech andHearing Disorders, 55., 172.Leviton, A., & Bellinger, D. (1986). Is there a relationship between otitis mediaand learning disorders. In J. F. Kavanagh (Ed)., Otitis Media and ChildDevelopment, York Press: Maryland, .pp.99-106.Marrow, H.,otitisIn D.OtitisWhitley, D., & Fulghum, R. (1988). Experimental otitis externa andmedia in the gerbil model with pseudomonas aeruginosa. (1987).Lim, C. Bluestone, J. Klein, and J. Nelson (Eds.), Recent Advances inMedia, B.C. Decker Inc.: Toronto, pp. 458-461.Matkin, N. (1986).^The role of hearing in language development. In J. F.Kavanaugh (Ed)., Otitis Media and Child Development, York Press:Maryland, pp. 3-11.Meck, W., Smith, R., & Williams, C. (1988). Pre- and postnatal cholinesupplementation produces long-term facilitation of spatial memory.Developmental Psychobiology, 21(4), 339-353.Menyuk, P. (1986). Predicting speech and language problems with persistentotitis media. In J. F. Kavanagh (Ed)., Otitis Media and Child Development,York Press: Maryland, .pp.83-96.Moore, B. C. J. (1982). The nature of sound and the structure of the auditorysystem. In B. C. J. Moore, An Introduction to the Psychology of Hearing.(2nd ed.) Academic Press. New York, N.Y. pp. 1-40.Moore, D., Hutchings, M., King, A. & Kowalchuk, N. (1989). Auditory brain stemof the ferret: some effects of rearing with a unilateral ear plug on thecochlea, cochlear nucleus, and projections to the inferior colliculus.The Journal of Neuroscience, 2.(4), 1213-1222.Morrongiello, B. A. (1989). Infants' monaural localization of sounds: effects ofunilateral ear infection. LAsaua,joc_kre,n Ir aw, 597-602.Morris, R. G. M. (1981). Spatial localization does not require the presence oflocal cues. Learning and Motivation, 12, 239-260.Nozza, R. (1988). Auditory deficit in infants with otitis media with effusion:more than a "mild" hearing loss. In D. Lim, C. Bluestone, J. Klein, and J.Nelson (Eds.), Recent Advances in Otitis Media, B.C. Decker Inc.: Toronto,pp. 267-280.69Ohashi, Y., Nakai, Y., Ikeoka, H., Koshimo, H., Esaki, Y. & Kato, S. (1988). Ciliaryactivity during experimental otitis media with effusion induced bylipopolysaccharide. In D. Lim, C. Bluestone, J. Klein, and J. Nelson (Eds.),Recent Advances in Otitis Media, B.C. Decker Inc.: Toronto, pp. 299-303.Oldfield, S. R., & Parker, S. P. A. (1984). Acuity of sound localization: atopography of auditory space. I. normal hearing. Perception il, 581-600.Paradise, J. L. (1981). Otitis media during early life: how hazardous todevelopment? a critical review of the evidence. Pediatrics, a, 868-873.Poucet, B., Chapuis, N., Durup, M., & Thinus-blanc, C. (1986). A study ofexploratory behavior as an index of spatial knowledge in hamsters.Animal Learning and Behavior, 14(1), 93-100.Proud, G. & Odoi, H. (1970). Effects of eustachian tube ligation. Annals ofOtology. Rhinology. and Laryngology, 7.2., 30-32.Roberts, J., & Schuele, C. (1990). Otitis media and later academic performance:The linkage and implications for intervention. Topics in LanguageDisorders, 11(1), 43-62.Roberts, J., Burchinal, M., Koch, M., Footo, M., & Henderson, F. (1988). Otitismedia in early childhood and its relationship to later phonologicaldevelopment. Journal of Speech and Hearing Disorders, 53_, 424-432.Roberts, J., Sanyal, M., Burchinal, M., Collier, A., Ramey, C., & Henderson, F.(1986). Otitis media in early childhood and its relationship to laterverbal and academic performance. Pediatrics, 21(3), 423-431.Roland, P., & Brown, 0. (1990). Tympanostomy tubes: A rational clinicaltreatment for middle ear disease. Topics in Language Disorders, 11(1),23-28.Rourke, B., & Finlayson, M. (1978). Neuropsychological significance ofvariations in patterns of academic performance:^Verbal and visual-spatial abilities. Journal of Abnormal Child Psychology, b_(1), 121-133.Rudy, J. & Hyson, R. (1983b). Ontogenesis of learning: III. Variation in therat's differential reflexive and learned responses to sound frequences.Developmental Psychobiology, 1(3), 285-300.Ryan, A. (1976). Hearing sensitivity of the mongolian gerbil, Merionesunguiculatis. J. Acoust. Soc. Am., 52(5), 1222-1225.Shurin, P., Johnson, C., & Wegman, D. (1986). Medical aspects of diagnosis andprevention of otitis media. In J. F. Kavanagh (Ed)., Otitis Media and ChildDevelopment, York Press: Maryland, .pp.60-69.70Steig-Pearce, P., Saunders, M., Creighton, D., & Sauve, R. (1988). Hearing andverbal-cognitive abilities in high-risk preterm infants prone to otitismedia with effusion. Developmental and Behavioral Pediatrics, Q(6),346-351.Stenfors, L., Hellstrom, S. & Salen, B. (1982). The role of the attic space inexperimental otitis media with effusion. Acta Otolaryngol, 386, 146-148.Strange, W. (1986). Speech input and the development of speech perception.In J. F. Kavanagh (Ed)., Otitis Media and Child Development, York Press:Maryland, .pp.13-25.Sutherland, R., & Dyck, R. (1984). Place navigation by rats in a swimming pool.Canadian Journal of Psychology, 2E2), 322-347.Sutherland, R. J., Whishaw, I. Q., & Kolb, B. (1980). Abnormalities in EEG andspatial performance following intrahippocampal injections ofneurotoxins, Society for Neuroscience Abstracts, 6, 565.Sutherland, R. J., Whishaw, I. Q. , & Kolb, B. (1982). A behavioral analysis ofspatial localization following electrolytic, kainate-induced damage tothe hippocampal formation in the rat, Behavioural Brain Research, 55,189-208.Tabachnik, B. G. & Fidell, L. S. (1983). Multivariate analysis of variance andcovariance. In B. G. Tabachnik, and L. S. Fidell, Using MultivariateStaistics. Harper and Row, New York, N. Y. pp.222-291.Teele, D. W. , Klein, J. 0., Rosner, B. A., & The Greater Boston Otitis Media Group.(1984). Otitis media with effusion during the first three years of life anddevelopment of speech and language. Pediatrics, 2_4(/), 282-287.Tees, R. C. (1967a). The effects of early auditory restriction in the rat on adultduration discrimination. The Journal of Auditory Research, j, 195-207.Tees, R. C. (1967b). Effects of early auditory restriction in the rat on adultpattern discrimination. J. Comp. Physiol. Psychol., 63, 389-393.Tees, R. C. (1990). Experience, perceptual competences, and rat cortex. In B.Kolb, and R. C. Tees (eds.), Cerebral Cortex of the Rat. Cambridge, Mass.pp. 506-536.Tees, R. C., Burhmann, K., & Hanley, J. (1989). The effect of early experienceon water amze spatial learning and memory in rats. Developmental Psychobiology, 23(5), 427-439.71Tees, R.C., Midgley, G., & Nesbit, J. (1981). The effect of early visual experienceon spatial maze learning in rats. Developmental Psychobiology, 14(5),425-438.Thinus-Blanc, C., Bouzouba, L., Chaix, K., Chapuis, N., Durup, M. & Poucet, B.(1987). A study of spatial parameters encoded during exploration inhamsters. Journal of Experimental Psychology, 1.314), 418-427.Todd, N. W. (1986). High risk populations for otitis media. In J. F. Kavanagh(Ed)., Otitis Media and Child Development, York Press: Maryland, .pp.52-59.Ventry, I. (1980). Effects of conductive hearing loss: Fact or fiction. Journal of Speech and Hearing Disorders, 45(2), 143-156.Wallace, I., Gravel, J., McCarton, C., & Ruben, R. (1988). Otitis media andlanguage development at 1 year of age. isnmald332esslaadHearLglDisorders, a, 245-251.Watkins, R. (1990). Processing problems and language impairment inchildren. Topics in Language Disorders, 11(1), 63-72.Webster, A. (1984). Auditory neuronal sizes after a unilateral conductivehearing loss. Exp. Neurol„ 7_9, 130-140.Webster, A., Bamford, J., Thyer, N., & Ayles, R. (1989). The psychological,educational, and auditory sequelae of early, persistent secretory otitismedia. J. Child Psychol. Psychiat., 3.1)_(4), 529 -546.Webster, D. & Webster, M. (1977). Neonatal sound deprivation affects brainstem auditory nuclei. Arch Otolaryngol, 103, 392-396.Widemar, L., Hellstrom, S., Nordling, A. & Johansson, U. (1988). Structuralchanges of the middle ear in experimentally evoked serous andpurulent otitis media--a study of two barrier regions. In D. Lim, C.Bluestone, J. Klein, and J. Nelson (Eds.), Recent Advances in Otitis Media,B.C. Decker Inc.: Toronto, pp. 464-466.Winer, B. J. Brown, D. R., & Michels, K. M. (1991). Statistical Principles inExperimentalataign. (3rd ed.) McGraw-Hill Inc., New York, N. Y.Yost, J. P., & Nielsen, R. N. (1985). The Fundamentals of Hearing. Oxford Press.,New York, N. Y.Zinkus, P. (1986). Perceptual and academic deficits related to early chronicotitis media. In J. F. Kavanagh (Ed)., Otitis Media and Child Development,York Press: Maryland, .pp.107-128.727 3APPENDICESAppendix 1. Descriptive statistics for audiogram data in control (SC), mildly effused(MBE), and severely effused (SBE) animals. This table includes the means andstandard deviations of threshold scores during an effusion episode, betweenepisodes, and after recovery.74MEANSDURING EFFUSION^BETWEEN EPISODES^AFTER RECOVERY1 kHz^8 kHz 16 kHz 1 kHz^8 kHz 16 kHz 1 kHz^8 kHz 16 kHzGROUPSC 58.80 35.40 27.80 57.00 39.60 28.80 58.60 41.60 27.60MBE 54.2 45.40 45.60 61.00 52.40 36.00 51.20 48.80 30.80SBE 52.2 67.80 57.20 52.00 67.80 56.20 54.80 51.40 34.60SD'SDURING EFFUSION^BETWEEN EPISODES^AFTER RECOVERY1 kHz^8 kHz 16 kHz 1 kHz^8 kHz 16 kHz 1 kHz^8 kHz 16 kHzGROUPSC 2.28 2.41 3.11 2.24 5.18 2.86 4.77 6.11 3.43MBE 4.21 4.45 6.50 4.00 2.61 4.74 4.82 4.09 0.84SBE 3.90 1.48 2.59 4.58 3.03 3.49 4.66 2.41 3.13Appendix 2. The ANOVA summary table for thresholds estimated for control (SC),mildly effused (MBE), and severely effused (SBE) rats.76SOURCEBetween^subjdf SS MS F T12Gp 2 3939.6 1969.8 60.61^* 0.910S's w. Gp's 1 2 390.0 32.5Within^subjEff 2 901.91 450.96 38.82* 0.764Gp by Eff 4 223.99 223.99 19.28* 0.764Eff by S's w. 2 4 286.93 11.96Gp'sFreq 2 7024.18 3512.09 293.76* 0.961Gp by Freq 4 3806.89 951.72 79.61* 0.930Freq by S'sw. Gp's2 4 286.93 11.96Freq by Eff 4 669.51 167.38 12.31* 0.506Gp by Freq 8 1080.49 135.06 9.93* 0.623Freq by Effby S's w. Gp's 4 8 652.67 13.6077Appendix 3. The ANOVA summary table of simple interactions of effusion state byfrequency. Included are findings from controls (SC), mildly effused (MBE),and severely effused (SBE) animals.78SOURCE df SS MS F 12CONTROLSEff by Freq 4 85.69 21.42 1.58 0.116MILDEff by Freq 4 538.09 134.42 9.88* 0.452SEVEREEff by Freq 4 1126.22 281.56 20.70* 0.63379Appendix 4. Summary table of the simple simple main effects of effusion state ateach frequency for mildly (MBE) and severely (SBE) effused rats.80SOURCE df SS MS F 12MILDEff at 1kHz 2 252.13 126.07 9.27* 0.278Eff at 8kHz 2 66.53 33.27 2.45 0.093Eff at 16kHz 2 563.73 281.87 20.73* 0.463SEVEREEff at 1kHz 2 24.40 12.20 0.90 0.036Eff at 8kHz 2 896.53 448.27 32.96* 0.579Eff at 16kHz 2 1630.53 815.27 59.95* 0.71481Appendix 5. Descriptive statistics for performance in the Open Field Paradigm. Datafrom rears, defecations, time spent in each quadrant, and number of contactsmade with each object are included. Animals included are SC (saline controls),MME (mildly monaurally effused), MBE (mildly binaurally effused), SME(severely monaurally effused), and SBE (severely binaurally effused) rats.82GROUPREARSPHASE 1^PHASE 2DE1±CATIONSPHASE 1^PHASE 2SC MEAN 28.50 22.83 2.75 0SD 8.38 10.33 2.49 0MME MEAN 19.92 15.83 0.92 1.33SD 13.14 11.96 1.32 1.37MBE MEAN 19.92 14.58 1.42 1.25SD 13.98 15.28 1.17 1.66SME MEAN 20.83 22.00 2.33 2.00SD 9.39 14.91 2.31 2.41SBE MEAN 25.50 25.50 2.58 1.58SD 13.01 13.87 1.98 2.39GROUPTIME SPENT IN QUAD 1PHASE 1^PHASE 2TIME SPENT IN QUAD 2PHASE 1^PHASE 2SC MEAN 171.50 111.92 88.00 131.25SD 92.30 61.56 36.56 70.11MME MEAN 215.17 200.92 74.67 103.75123.09 139.69 52.10 79.32MBE MEAN 137.00 168.92 81.58 61.50SD 131.15 116.98 104.20 59.51SME MEAN 161.08 139.33 94.42 93.17SD 117.05 118.29 45.54 73.62SBE MEAN 170.25 149.92 78.75 109.67SD 87.65 100.27 47.82 81.77GROUPTIME SPENT IN QUAD 3PHASE 1^PHASE 2TIME SPENT IN QUAD 4PHASE 1^PHASE 2MEAN 120.92 116.58 106.67 120.25SD 71.39 63.42 55.27 70.89MME MEAN 102.25 59.75 87.92 111.83SD 95.92 63.91 52.62 123.67MBE MEAN 118.00 120.50 144.25 126.25SD 80.57 99.63 117.71 107.48SME MEAN 123.42 133.58 101.08 111.92SD 82.49 79.87 58.09 92.32SBE MEAN 111.17 95.67 119.83 124.75SD 55.69 51.68 73.50 107.93Quad. 1 - denotes the quadrant containing the object that was changed in orientation and location within aquadrant in Phase 2.Quad. 2 - denotes the quadrant containing the object that was replaced with a novel one in Phase^2.Quad. 3 - denotes the quadrant containing the object that was left alone for Phase 2.Quad. 4 - denotes the quadrant with no object in it.GROUPMEANSDOBJECT 1 CONTACTSPHASE 1^PHASE 2^9.17^5.754.63 2.30OBJECT 2 CONTACTSPHASE 1^PHASE 26.92^7.333.68 2.64MME MEAN 5.91 5.36 3.27 3.18SD 3.42 4.06 2.10 2.56MBE MEAN 4.30 4.60 3.10 2.30SD 3.95 3.81 3.41 3.27SME MEAN 5.36 5.55 5.36 4.73SD 3.33 3.62 4.72 3.38SBE MEAN 5.63 6.75 5.12 5.50SD 2.26 3.96 2.99 4.72GROUPOBJECT 3 CONTACiSPHASE 1^PHASE 2SC MEAN 8.00 7.92SD 3.84 4.46MME MEAN 3.91 2.91SD 4.04 3.51MBE MEAN 5.00 3 .80SD 5.35 4.34SME MEAN 6.91 5.64SD 5.70 4.13SBE MEAN 6.38 5.38SD 3.89 3.78Object 1 - denotes the object that was changed in orientation and locationwithin a quadrant in Phase 2.Object 2 - denotes the object that was replaced with a novel one in Phase 2.Object 3 - denotes the object that was left alone for Phase 2.Appendix 6. Descriptive statistics for performance on the visual spatial water mazetask. Data from all five groups (SC, MME, MBE, SME, and SBE) are presented.88CONTROLSDAY 1 DAY 2 DAY 3 DAY 4 DAY 5 DAY 6 DAY 7 DAY 8 DAY 9 DAY 10MEAN 25.79 24.08 19.76 26.39 23.59 21.53 18.80 16.05 14.23 16.21SD 17.39 18.09 14.08 19.65 17.66 13.43 10.00 7.78 6.38 9.04MILD MONAURALSDAY 1 DAY 2 DAY 3 DAY 4 DAY 5 DAY 6 DAY 7 DAY 8 DAY 9 DAY 10MEAN 37.70 38.90 31.07 31.81 27.81 33.22 23.16 28.22 25.40 23.71SD 18.42 25.59 17.15 17.60 21.68 24.63 11.80 16.37 15.08 13.50MILD BINAURALSDAY 1 DAY 2 DAY 3 DAY 4 DAY 5 DAY 6 DAY 7 DAY 8 DAY 9 DAY 10MEAN 33.99 22.08 22.99 22.94 18.40 17.55 18.59 15.55 17.25 19.96SD 28.66 15.01 17.59 15.87 13.38 10.89 13.08 8.02 12.04 12.85SEVERE MONAURALSDAY 1 DAY 2 DAY 3 DAY 4 DAY 5 DAY 6 DAY 7 DAY 8 DAY 9 DAY 10MEAN 38.44 26.71 26.65 28.79 21.37 12.09 14.85 12.86 11.62 10.82SD^26.13 17.48 18.45 26.20 17.34 5.22^5.51^3.21^4.20^3.28SEVERE BINAURALSDAY 1 DAY 2 DAY 3 DAY 4 DAY 5 DAY 6 DAY 7 DAY 8 DAY 9 DAY 10MEAN 27.46 22.11 17.39 15.92 14.76 12.99 13.32 11.72 12.14 10.50SD^21.78 20.67 10.86 10.34 9.57^8.10^6.75^4.11^5.91^3.73Appendix 7. The ANOVA summary table for the visual spatial water maze task.Results include findings from control (SC), mild monaural (MME), mildbinaural (MBE), severe monaural (SME), and severe binaural (SBE) animals.91SOURCE d f SS MS Wilks A F 12Between subjGp 4 10831.47 2707.87 1.75 0.135S's w. Gp's 45 69598.35 1546.63Within subjDay 9 13326.12 1480.68 0.479 4.47* 0.521Gp by Day 3 6 4021.53 111.71 0.388 1.12+ 0.211Day by S's w. 405 49082.88 121.9Gp's*1.<0.010.0+ - approximate F valueAppendix 8. Descriptive statistics for performance in the auditory spatial water mazetask. Results from all five groups are included.93CONTROLSDAY^MEAN SD1 35.31 15.562 32.05 18.313 32.25 22.474 28.42 19.465 25.99 15.746 26.01 17.497 21.69 15.458 20.00 9.669 25.07 20.991 0 19.54 10.171 1 17.01 7.911 2 19.55 12.881 3 13.01 5.861 4 14.31 8.411 5 14.62 8.331 6 13.30 6.581 7 10.30 3.531 8 8.99 2.441 9 9.74 3.2720 9.60 3.8221 9.99 3.9322 9.46 4.1723 9.49 3.6124 10.54 3.9225 8.50 2.5194MILD MONAURAL RATSDAY MEAN Si)1 38.29 14.402 38.56 11.563 32.79 12.654 33.00 10.825 26.31 10.026 23.36 8.577 21.61 6.968 20.15 7.939 24.19 10.0810 14.96 1.631 1 16.84 4.2812 15.55 4.111 3 18.87 8.3414 19.11 5.931 5 17.67 7.251 6 19.36 5.551 7 17.92 4.851 8 14.57 3.551 9 17.52 8.0020 16.65 7.8421 16.59 5.6522 15.29 4.5423 17.14 5.192 4 16.65 6.2225 13.94 3.5095MILD BINAURAL RATSDAY^MEAN SD1 44.77 22.322 42.84 18.993 35.10 17.074 36.36 19.665 37.29 19.236 30.31 14.287 29.71 11.628 25.50 10.439 26.15 10.9110 25.76 10.241 1 22.60 10.401 2 22.61 13.501 3 21.25 12.011 4 16.55 5.911 5 20.81 12.451 6 14.92 4.261 7 16.16 3.841 8 14.62 4.721 9 13.92 7.0620 15.41 5.4021 13.95 4.6122 13.44 3.5023 13.19 5.4124 15.16 5.6625 15.65 8.4896SEVERE MONAURAL RATSDAY MEAN SD1 35.14 23.052 29.72 18.383 27.00 13.904 28.79 18.695 25.26 17.716 28.67 16.367 28.06 17.738 29.12 18.759 18.15 9.721 0 20.24 13.721 1 20.74 10.031 2 20.27 13.4013 19.25 8.001 4 17.87 8.151 5 23.51 6.661 6 24.91 12.9517 18.90 6.031 8 16.52 8.431 9 17.24 5.2520 16.30 10.262 1 16.90 5.4122 15.87 5.0423 17.21 13.3524 14.79 3.782 5 16.75 8.8097SEVERE BINAURAL RATSDAY MEAN SD1 34.91 19.542 35.99 20.923 28.99 15.434 28.10 18.615 28.65 15.376 31.27 20.237 24.41 16.098 26.75 19.009 19.15 14.2910 25.57 17.891 1 22.21 11.7912 20.27 13.4013 15.25 6.161 4 13.41 4.1015 17.72 7.271 6 13.91 4.2317 15.94 7.611 8 12.75 4.751 9 12.14 3.7420 14.81 5.3221 13.27 3.9622 12.12 4.192 3 11.07 4.6124 12.29 3.8225 10.96 2.449.8 •Appendix 9. The ANOVA summary table for the auditory spatial water maze task.Results include data collected from all five groups.99SOURCEBetween subjd f SS MS Wilks A F 712Gp 4 4333.52 1083.38 0.86 0.071S's w. Gp's 45 56772.49 1261.61Within^subjDay 24 13326.12 1480.68 0.171 4.44* 0.829Gp by Day 96 8331.25 86.78 0.117 0.67+ 0.415Day by S's w. 1080 90448.06 83.75Gp's* R < 0.01+ - approximate F value