COGNITION AND EMOTION, 1998, 12 (3), 307 330

Affective Style and Affective Disorders: Perspectives from Affective Neuroscience

Richard J. Davidso n
U niversity of W isconsin M adiso n, U SA

Individual diffe rences in emotional reactivity or affe ctive style can be fruitfully decomposed into more elementary c onstituents. Several separable features of affe ctive style are identi ed such as the threshol d for reactivity, the peak ampl itude of response, the rise time to peak and the rec overy time . The latter two characteristics constitute components of affec tive chronome try. The circuitry that underlies tw o fundamental forms of motivation and emotion approach and w ithdraw al-relate d processes is described. Data on individual differences in functional activity in certain components of these circuits are next review ed, w ith an emphasis on the nomological netw ork of associations surrounding individual diffe rences in asymme tric prefrontal activati on. The relevance of such differences for understanding the nature of the affe ctive dysfunction in affe ctive disorders is then considered. The article ends by considering w hat the prefrontal cortex ``does in certain components of affe ctive style and highlights some of the important questions for future research.

I. INTRODUCTION
Among the most striking features of human emotion is the variability that is appare nt across individuals in the quality and intensity of disposition al mood and emotional reactions to similar incentive s and challe nge s. The broad range s of differences in these varie d affective phenomena has been

Requests for reprints should be sent to Ric hard J. Davidson, Laboratory for Affec tive Neuroscienc e, Department of Psychology, University of Wisconsin Madison, 1202 West Johnson Street, M adison, WI 53706, USA; e-mail: davidson@ macc.wisc.edu. This research was supported by NIMH grants MH43454, MH40747, Research Scientist Award K05-MH00875, and P50-MH52354 to the Wisconsin Center for Affective Science (R.J. Davidson, Director), by a NARSAD Establis hed Investigator Aw ard, and by a grant from the John D. and Catherine T. M acArthur Foundation. I thank the many indiv iduals in my laboratories who have contributed importantly to this research ove r the years, inc luding Andy Tomarken, Steve Sutton, Wil Irwin, Heather Abercrombie, Jeff Henriques, Chris Larson, Stacey Schaefer, Terry Ward, Darren Dottl, Isa Dolski, as well as the many collaborators outside my lab too numerous to name.

1998 Psychology Press Ltd

308

DAVIDSON

referred to as ``affective style (Davidson, 1992 ). Differences among people in affectiv e style appe ar to be associate d w ith temperament (Kagan, Reznick & Snidm an, 1988), personality (Gross, Sutton, & Ketelaar, in press) and vulne rability to psyc hopathology (Meehl, 1975). Moreov er, such diffe rences are not a unique human attribute , but appe ar to be present in a number of different species (e.g. Davidson, Kalin, & Shelton, 19 93; Kalin, 1993). In the next section of this article, conceptual distinc tions among the various compone nts of affective style w ill be introduc ed and methodological challe nge s to their study w ill be hig hlighte d. The third section will present a brief ove rview of the anatom y of tw o basic motivational/ emotional systems the approac h and w ithdraw al systems. Section four will cons ider individual diffe rences in these basic systems and indic ate how such differences m ight be studie d. The fth section will addre ss the relation betw een such individual differences and psychopathology. It is our intuition that some of the individual differences in basic processes of affective style are central to determining either resilie nce or vulne rability. Such diff erences can be conceptualise d as diathe ses w hich affect an individual s response to a stressful life event. Finally, the last section w ill conside r some of the implic ations of this perspective for assessment, treatment and plastic ity.

II. THE CONSTITUENTS OF AFFECTIVE STYLE

Many phenom ena are subsum ed unde r the rubric of affective style. A conc ept featured in many disc ussions of affective de velopme nt, affectiv e disorde rs and personality is ``emotion regulation (Thompson, 1994). Emotion regulation refers to a broad constellation of processes that se rve to either amplify , attenuate , or maintain the strength of emotional reactions. Include d among these processes are certain features of attention w hich regulate the extent to w hich an org anis m can be distrac ted from a potentially ave rsive stimulus (Derryberry & Reed, 1996 ) and the capac ity for self-generated image ry to replac e emotions that are unw ante d, w ith more desirable image ry scripts. Emotion regulation can be both automatic and controlle d. Automatic emotion regulation may result from the prog ressive automisation of processes that initially w ere voluntary and controlle d and have evolve d to become more autom atic w ith practice. We hold the view that regulatory processes are an intrinsic part of emotional behav iour and rarely does an emotion get gene rated in the absence of recruiting associated regulatory processes. For this reason, it is often conceptually dif cult to distinguis h sharply betw een w here an emotion ends and regulation begins. Even more proble matic is the methodologic al challe nge of operationalis ing these diffe rent compone nts in the stream of affectiv e behaviour.

AFFECTIVE STYLE AND AFFECTIVE DISORDERS

309

Whe n consi dering the question of indiv idual differences in affective behaviour, one must specify the partic ular respons e systems in w hich the individual diffe rences are being explore d. It is not necessarily the case that the same pattern of individual differences w ould be found across response systems. Thus, for example , an individual may have a low threshold for the elicitation of the subje ctive expe rience (as re e cted in self-reports) of a partic ular emotion but a relativ ely high threshold for the elicitation of a partic ular phy siologic al change . It is important not to assume that indiv idual differences in any parame ter of affective responding will necessarily generalise ac ross response systems, w ithin the same emotion. Equally important is the question of whether indiv idual differences associated w ith the generation of a partic ular speci c emotion w ill nec essarily generalise to other em otions. For example , are those individuals who are behaviourally expressiv e in respons e to a fear challe nge als o like ly to show comparably high levels of expressiv ity in response to positiv e incentive s? Although systematic research on this question is still require d, initial evide nce suggests that at least certain aspects of affective style may be emotion-s peci c, or at least valence-speci c (e .g. Wheeler, Davidson, & Tomarke n, 199 3). In addition to emotion regulation, there are probably intrinsic diff erences in certain compone nts of emotional responding. There are like ly individual diffe rences in the threshold for eliciting compone nts of a partic ular emotion, give n a stimulus of a certain intensity. Thus, some individuals are like ly to produc e facial signs of disgust on presentation of a partic ular intensity of nox ious stimulus, whe reas other individuals may require a more intense stim ulus for the elicitation of the same respons e at a comparable intensity . This sugg estion implie s that dose-response functions may reliably diffe r across individuals . Unfortunate ly, systematic studie s of this kind hav e not been performed, in part because of the dif culty of creating stimuli that are grade d in intensity and designe d to elicit the same em otion. In the olfactory and gustatory modalitie s, there are possibili ties of creating stimuli that differ systematically in the c oncentration of a disgust-produc ing compone nt and then obtaining psychophys ical threshold func tions that w ould reveal such indiv idual differences. How ever, the produc tion of such inte nsi ty-g raded stim uli in other modalitie s w ill like ly be more complicate d, althoug h w ith the dev elopment of large , normative ly rated complex stimulus sets, this may be poss ible . An example is the Interna tional Affective Picture Sys tem (Lang, Bradle y, & Cuthbe rt, 19 95) develope d by Peter Lang and his colle ague s. This set include s a large number of visual stimuli that have been rated on vale nce and arous al dimensions and that com prise locations throug hout this two-dim ensional space. The density of stimulus exemplars at all levels within this space allow for the pos sibility of selecting stimuli that are grade d in intensi ty for the sort of dose-response studie s described above .

310

DAVIDSON

There are also like ly to be individual diffe rences in the peak or amplitude of the response . On presentation of a series of graded stimuli that differ in intensity, the maximum amplitude in a certain system (e.g. intensity of a facial contrac tion, change in heart rate, etc.) is like ly to differ systematically across subje cts. Some individuals will respond w ith a large r amplitude peak compared w ith othe rs. Again, such individual differences may w ell be quite speci c to partic ular systems and w ill not necessarily generalise across systems, even within the same emotion. Thus, the indiv idual w ho is in the tail of the distribution in his/he r heart rate response to a fearful stimulus w ill not necessarily be in the tail of the distribution in his/he r facial response . Anothe r parame ter that is like ly to differ systematically across individuals is the rise time to peak. Some individuals will rise quic kly in a certain response system, whereas others w ill rise more slowly. There may be an association betw een the peak of the response and the rise tim e to the peak w ithin certain systems for partic ular emotions. Thus, it may be the case that for ange r-related emotion, thos e individuals w ith hig her peak voc al response s als o show a faster rise time, but to the best of my know ledge , there are no systematic data related to such differences. Finally, anothe r compone nt of intrinsic diffe rences across individuals is the re covery time . Follow ing perturbation in a partic ular system, some individuals recove r quic kly and others recove r slow ly. For example, follow ing a fear-provoking encounte r, some individuals show a persisting heart rate elevation that might last for minute s, w hereas other individuals show a comparable peak and rise time, but recov er much more quickly . Of course, as with other param eters, there are like ly to be dif ferences in recove ry time across different respons e systems. Some individuals may recove r rapidly in their expressive behaviour, w hile recove ring slow ly in certain autonomic channe ls. The potential sig ni cance of such dis sociations has not been systematically exam ined. The speci c param eters of indiv idual diffe rences that are deline ated abov e describe affective chro nometry the temporal dy namic s of affectiv e responding. V ery little is know n about the factors that gov ern these individual diffe rences and the extent to w hich such differences are speci c to partic ular emotion response systems or generalis e across emotions (e.g. is the heart rate recove ry follow ing fear similar to that follow ing disgus t? ). Moreove r, the general issue of the extent to which these diffe rent parame ters that have been identi ed are orthogonal or correlate d features of emotional responding is an empiric al que stion that has yet to be answ ered. For reasons that I hope to make clear later, affec tive chronome try is a partic ularly important feature of affectiv e style and is like ly to play a key role in determining vulne rability to psychopatholog y. It is also a feature of affective style that is methodo-

AFFECTIVE STYLE AND AFFECTIVE DISORDERS

311

logic ally tractable and can yield to expe rimental study of its neural subs trates. We also hold that affectiv e style is critical in unde rstanding the continuity betw een norm al and abnorm al functioning and in the predic tion of psychopathology and the deline ation of vulne rability. On the opposite side of the spectrum, such individual difference s in affective style will also feature centrally in any comprehensive theory of resilie nce. The fact that some individuals reside ``off the diag onal and appe ar to maintain very high levels of psyc holog ical w ell-be ing despite their exposure to obje ctive life adve rsity is like ly related to their affective style (R yff & Singe r, in press). Some of these im plic ations w ill be discussed at the end of this article. We rst cons ider some of the neural substrate s of two fundam ental emotion systems. This provide s the foundation for a conside ration of individual differences in these systems and the neural circuitry responsible for such differences.

III. THE ANATOMY OF APPROACH AND WITHDRAWAL

Although the focus of my empiric al research has been on measures of prefrontal brain activity, it must be em phasise d at the outset that the circuit instantiating emotion in the human brain is comple x and involv es a numbe r of interrelated structures. Preciously few empiric al studie s using modern neuroimaging procedures that afford a high degree of spatial resolution have yet been perform ed (see George et al., 1995; Paradiso et al., 1997, for example s). Therefore, hypothes es about the set of structures that partic ipate in the produc tion of emotion must necessarily be speculativ e and base d to a larg e extent on the inform ation av ailable from the anim al literature (e.g. LeDoux, 198 7), and from theoretical acc ounts of the processes involve d in human emotion. Based upon the av ailable strands of theory and evide nce, numerous scientists have propose d tw o bas ic circuits each mediating different forms of motivation and emotion (see e.g. Davidson, 1995; Gray, 199 4; Lang, Bradle y, & Cuthbe rt, 199 0). The approac h system facilitate s appe titive behaviour and generates certain type s of positive affect that are approac h-related (e.g. enthusiasm , pride , etc., see Depue & Collins, in press for a review ). This form of pos itive affect is usually generated in the contex t of moving toward a desired goal (see Lazarus, 1991 and Stein & Trabasso, 199 2, for theoretical accounts of emotion that plac e a premium on goal states). The representation of a goal state in w orking memory is hypothes ised to be im plemented in dorsolate ral pre frontal cortex. The medial prefrontal cortex seem s to play an important role in maintaining

312

DAVIDSON

representations of behavioural-re inforc ement continge ncies in w orking memory (Thorpe , Rolls, & Maddis on, 198 3). In addition, output from the medial prefrontal cortex to nucleus accumbens (NA) neurone s m odulate s the transfe r of motivationally relevant inform ation through the NA (Kalivas , Churchill, & Klitenick, 19 93). The basal ganglia are hypothes ised to be involve d in the expression of the abstract goal in action plans and in the antic ipation of rew ard (Schultz , Apicella, R omo, & Sc arnati, 199 5a; Schultz et al., 19 95b). The NA, partic ularly the caudom edial shell region of the NA, is a major conve rgence zone for motivationall y relevant inform ation from a myriad of limbic structures. Cells in this region of the NA increase their ring rate during reward expe ctation (see Schultz et al., 1995a). There are probably other structures inv olve d in this c ircuit w hich depend on a number of factors including the nature of the stimuli signalli ng appe titiv e inform ation, the extent to w hic h the behaviouralreinforc ement continge ncy is nove l or ov erlearned, and the nature of the antic ipate d behavioural respons e. It should be noted that the activation of this approac h system is hy pothe sised to be associated w ith one partic ular form of positiv e affect and not all forms of such emotion. It is speci cally predicted to be associated w ith pre-goal attainme nt posi tive affec t, that form of positiv e affect that is elicited as an org anis m move s clos er tow ard an appe titiv e goal. Post-goal attainme nt positiv e affect represents another form of positive em otion that is not expe cted to be associated with activ ation of this circuit (see Davids on, 19 94 for a more extended dis cussion of this distinc tion). This latter type of positive affect may be phenomenolog ically expe rienced as contentment and is expe cted to occur w hen the prefrontal cortex goe s off-line after a desired goal has been achieved. Cells in the NA have also been show n to decrease their ring rate during post-goal consummatory behaviour (e.g. Henrikse n & Giacchino, 1993). Law ful individual diffe rences can enter into many different stag es of the approac h system. Such individual differences and their role in modulating vulnerabil ity to psychopathology w ill be cons idered in detail later. For the moment, it is important to unde rscore tw o issues. One is that there are individual differences in the tonic level of activation of the approac h syste m w hich alters an individual s prope nsity to expe rience approac hrelated positiv e affect. Second, there are like ly to be individual diff erences in the capac ity to shift betw een pre- and post-g oal attainme nt positiv e affect and in the ratio between these two forms of positiv e affect. On reaching a desired goal, some individuals w ill immediate ly replac e the just-ac hieved goal w ith a new desired goal, and so w ill hav e little opportunity to e xpe rience pos t-goal attainme nt positive affect, or contentment. There m ay be an optimal balance betw een these two forms of positiv e affect, althoug h this issue has not receive d systematic study.

AFFECTIVE STYLE AND AFFECTIVE DISORDERS

313

There appe ars to be a second system concerned with the neural imple mentation of w ithdraw al. This system facilitates the withdraw al of an individual from sources of ave rsive stimulation and generates certain forms of negative affect that are w ithdraw al-related. Both fear and disgus t are associated w ith increasing the distanc e betw een the organism and a source of av ersiv e stimulation. From invasive animal studie s and human neuroimaging studie s, it appe ars that the amy gdala is critically involve d in this system (e.g . LeDoux, 19 87). Using functional magnetic resonanc e imaging (fMRI) we have recently demonstrate d for the rst time activation in the human amyg dala in response to av ersive pictures compared w ith neutral control pictures (Irw in et al., 199 6). In addition, the temporal polar region also appe ars to be activate d during w ithdraw al-related emotion (e.g. Reiman, Fusselman, Fox, & R aichle, 1989; but see Drevets, V ideen, MacLeod, Halle r, & Raichle, 1992 a). These effects, at least in humans, appe ar to be more pronounc ed on the right side of the brain (se e Davids on, 1992 , 1993 for review s). In human electrophysiologi cal studie s, the right frontal region is also activate d during w ithdraw al-related negativ e affective states (e.g. Davidson, Ekman, Saron, Senulis, & Friesen, 1990 a). At present, it is not entirely clear w hether this EEG change re ects activ ation at a frontal site or w hether the activity recorded from the frontal scalp region is volume -conduc ted from other cortical loci. The resolution of this uncertainty must aw ait additional studie s using positron emission tomography (PET) or fMRI, w hich hav e suf c ient spatial resolution to diff erentiate among diffe rent anterior cortical regions. In addition to the temporal polar region, the amygdala and poss ibly the prefrontal cortex, it is als o like ly that the bas al ganglia and hypothalamus are involv ed in the motor and autonom ic compone nts, respectiv ely, of w ithdraw al-related negativ e affect (see Smith, DeVito, & Astley, 199 0). The nature of the relation between these tw o hy pothe sised affect systems also rem ains to be deline ated. The emotion literature is replete w ith different proposals regarding the interrelations among different forms of positive and negative affect. Some theorists have propose d a single bivalent dimension that rang es from unple asant to pleasant affect, w ith a second dimension that re ects arous al (e.g. R ussell, 1980). Other theorists have sug gested that affect spac e is best described by tw o orthogonal positive and negative dimensions (e.g. Watson & Tellegen, 1985). Still other w orkers have sug gested that the degree of orthogonality betw een positive and neg ative affect depends upon the temporal frame of analysis (Diener & Emmons, 198 4). This formulation holds that whe n assessed in the m oment, positive and negative affect are reciproc ally related, but w hen examine d ove r a longe r time-frame (e.g. dis positional affect), they are orthogonal. It must be emphasise d that these analyse s of the relation between positive and negativ e affect are all base d exclusi vely on measures

314

DAVIDSON

of self-re port and therefore their generalisabili ty to other measures of affect are uncertain. However, based on new data to be described later, w e believ e that a grow ing corpus of data does inde ed indic ate that one func tion of positive affect is to inhibit concurrent negativ e affect.

IV. INDIVIDUAL DIFFERENCES IN ASYMMETRIC PREFRONTAL ACTIVATION: WHAT DO THEY REFLECT?

This section w ill present a brief ove rview of recent w ork from my laboratory designed to exam ine individual differences in m easures of prefrontal activ ation and their relation to different aspe cts of emotion, affective style, and related biologica l cons tructs. These ndings w ill be used to addre ss the question of w hat unde rlying constitue nts of affectiv e style such individual differences in prefrontal activ ation actually re ect. In both infants (Davidson & Fox, 1989 ) and adults (Davids on & Tomarke n, 19 89) we notic ed that there w ere large individual diff erences in baseline electrophysiolog ical m easures of prefrontal activation and that such individual variation w as associated with differences in aspe cts of affective reactiv ity. In infants , Davidson and Fox (19 89) reported that 10 -month-old infants who cried in respons e to maternal separation w ere more like ly to have less left and greate r right-side d prefrontal activation during a preceding resting baseline compared w ith those infants w ho did not cry in respons e to this challe nge . In adults, w e rst noted that the phasi c in uence of pos itive and negative emotion elic itors (e.g. lm clips) on measures of prefrontal activ ation asymmetry appe ared to be superimpose d on more tonic individual dif ferences in the direction and absolute magnitude of asymm etry (Davidson & Tomarken, 1989 ). During our initial explorations of this phenomenon, w e needed to determine if base line electrophysiolog ical measures of prefrontal asymmetry w ere reliable and stable ove r time and thus could be used as a trait-like measure. Tomarken, Davids on, Wheele r, and Doss (19 92) recorded base line brain electrical activity from 90 normal subje cts on tw o occasions separate ly by approximate ly three w eeks. At each testing session, brain activ ity w as recorded during eight 1-minute trials, four eyes open and four eyes close d, presented in counte rbalanc ed order. The data were scored visually to remove artefac t, and then Fourie r-transforme d. Our focus was on pow er in the apha band (8 13Hz), although w e extrac ted pow er in all frequency bands (see Davids on, C hapm an, Chapman, & Henrique s, 1990 for a disc ussion of pow er in diffe rent frequency bands and their relation to activ ation). We computed coef cient alpha as a measure of internal consistency reliability from the data for each session. The coef cient alphas w ere quite high, with all value s exceeding .85, indic ating that the electro-

AFFECTIVE STYLE AND AFFECTIVE DISORDERS

315

phys iolog ical measures of asy mmetric activation inde ed show ed excellent internal consiste ncy reliability. The test-retest reliability was adequate w ith intrac lass correlations rang ing from .65 to .75 , depending on the speci c sites and methods of analys is. The major nding of import from this study was the demons tration that measures of activation asym metry based on pow er in the alpha band from ante rior scalp electrodes show ed both high internal cons istency reliability and acceptable test-retest reliability to be conside red a trait-like inde x. The large sample size in the reliability study disc ussed above e nabled us to select a small group of extreme left and extreme rig ht-frontally activate d subje cts for mag netic resonanc e imag ing (MRI) scans to determine if there existed any gros s morphom etric differences in anatom ical structure between these subgroups. None of our measures of regional volume tric asym metry revealed any diffe rence betw een the groups (unpublishe d observations) . These ndings sugge st that w hate ver differences exist between subje cts w ith extreme left versus right prefrontal activation, those differences are like ly func tional and not structural. On the bas is of our prior data and theory, we reasone d that extreme left and extreme rig ht frontally activate d subje cts w ould show systematic differences in disposi tional positiv e and negativ e affect. We adm inis tered the trait version of the Positive and Negative Affect Scales (PANAS; Watson, C lark, & Tellegen, 1988) to examine this question and found that the left-frontally activate d subje cts reported more positive and less negativ e affect than their right-frontally activate d counterparts (Tomarken et al., 1992; see Fig. 1, ov erleaf). More recently w ith Sutton (Sutton & Davidson, 19 97) w e show ed that sc ores on a self-report measure designe d to operationalise Gray s concepts of B ehavioral Inhibition and B ehav ioral Activation (the BIS/BAS scales; Carver & White , 199 4) w ere even more strongly predicted by electrophysiolog ical measures of prefrontal asym metry than were scores on the PANAS scale s (see Fig. 2, situated on p. 337 of the Colour Plate Section). Subje cts w ith greater le ft-side d prefrontal activation reported more relative B AS to B IS activity compared w ith subje cts exhibiting more right-side d prefrontal activation. We also hypothe sised that our measures of prefrontal asym metry would predict reactivity to expe rime ntal elicitors of emotion. The model that w e hav e develope d ov er the past several years (see Davidson, 19 92, 1994 , 19 95 for background) features individual diffe rences in prefrontal activation asymmetry as a re ection of a diathe sis w hich modulate s reactivity to emotionally signi cant events. According to this model, individuals w ho differ in prefrontal asy m metry should respond differently to an elicitor of positive or negativ e em otion, even w hen bas eline mood is partiale d out. We (Wheeler et al., 19 93) performed an experiment to exam ine this question. We presente d short lm clips desig ned to

316

DAVIDSON

FIG. 1.

Dispositional positive and negative affect (from scores on the PANAS-General Positive and Negative Affect Sc ale) in subjects who were classi ed as extreme and stable left-frontally active ( N = 14) and extreme and stable right-frontally active ( N = 13) on the basis of electrophysiologic al measures of base line activation asymmetrie s on two occasions separated by three wee ks. From Tomarken et al. (1992).

elicit positive or negative emotion. Brain electrical activity was recorded prior to the presentation of the lm clips. Just after the clips were presented, subjec ts w ere asked to rate their emotional expe rience during the preceding lm clip. In addition, subje cts comple ted scales that w ere desig ned to re ect their mood at bas eline . We found that individual differences in prefrontal asy mmetry predicted the emotional response to the lms even after the varianc e contribute d by baseline mood w as statistic ally remove d. Thos e individuals with more left-sided prefrontal activation at bas eline reported more positiv e affect to the pos itive lm clips and those w ith more rightsided prefrontal activ ation reported more negative affect to the negative lm clips. These ndings support the idea that individual differences in electrophysiolog ical measures of prefrontal activation asy mmetry mark some aspe ct of vulnerabil ity to positive and negative emotion elicitors. The fact that such relations w ere obtaine d follow ing the statistic al removal of base line mood indic ates that any difference betw een left and right frontally activ ated in bas eline m ood cannot account for the prediction of lmelicite d emotion effects that were observed. In a very recent study, w e (Davidson, Dolski, Laron, & Sutton, in prep.) exam ined relations betw een indiv idual diffe rences in prefrontal activation asy mmetry and the emotion-m odulated startle. In this study, w e presented

AFFECTIVE STYLE AND AFFECTIVE DISORDERS

317

pictures from the Internation al Affective Picture Syste m (Lang et al., 1995) while acous tic startle probe s w ere presented and the EMG-measured blink response from the orbic ularis oculi muscle region w as recorded (see Sutton, Davids on, Donze lla, Irw in, & Dottl, 1997 for basic methods). Startle probes w ere presented both during the 6-second slide exposure as we ll as 500 milliseconds follow ing the offset of the pic tures, on separate 1 trials. We interpreted startle magnitude during picture exposure as prov iding an inde x related to the peak of emotional respons e, w hereas startle magnitude follow ing the offse t of the pictures w as take n to re ect the recove ry from e motional challe nge . Used in this way , startle probe methods can potentially provide new inform ation on the time course of emotional responding. We expe cted that individual diffe rences during actual picture presentation w ould be less pronounc ed than individual differences follow ing picture pre sentation as an acute emotional stimulus is like ly to pull for a normative response across subje cts, yet individuals are likely to differ dram atically in the time to recove r. Similarly , w e predicted that individual differences in prefrontal asym metry w ould account for more varianc e in predicting magnitude of recove ry (i.e. startle magnitude poststimulus) than in predic ting startle mag nitude during the stimulus. Our ndings were consiste nt w ith our predic tions and indic ated that subje cts with greater right-side d prefrontal activation show a large r blink magnitude follow ing the offset of the negativ e stimuli, after the varianc e in blink magnitude du rin g the negative stimulus w as partiale d out. Measures of prefrontal asym metry did not reliably predict startle magnitude during picture presentation. The ndings from this study are cons istent w ith our hypothes is and indic ate that individual differences in prefrontal asym metry are associate d w ith the time course of affective responding, partic ularly the recove ry follow ing emotional challe nge . In addition to the studies just described using se lf-report and psyc hophys iolog ical measures of emotion, w e have also exam ined relations

In this initial study on the recovery func tion assessed with startle probe measures, we had only a single post-stimulus probe at 500ms follow ing the offset of the picture. Readers may be surprised that the interval be tween the offset of the picture and the presentation of the probe was so short. However, it should be noted that these emotional pictures are not particularly intense and so the lingering effects of emotion follow ing the presentation of such pic tures is likely not to last very long in most individuals. Future studies will probe furthe r out follow ing the offset of the pic ture. Because, at most, only a sing le probe can be presented for each picture so that habituation effec ts are minimised, each new probe position requires a substantial incre ase in the overall numbe r of pictures presented. There is a nite limit to the number of pictures containe d in the IAPS. Even more importantly, we have found that it is critical to keep the picture vie wing period to well under one hour to minimise fatigue and bore dom.

318

DAVIDSON

betw een individual diffe rences in electrophysiolog ical measures of prefrontal asym metry and othe r biolog ical indic es w hich in turn have been related to diff erential reactivity to stressful events. Tw o recent exam ples from our laboratory include measures of immune function and cortisol. In the case of the former, w e examine d diffe rences betw een left- and rig htprefrontally activ ated subje cts in natural kille r cell activity, becaus e decline s in NK activity have been reported in response to stressful, negative events (Kiecolt-Glase r & Glaser, 1991 ). We predicted that subjects w ith right prefrontal activation w ould exhibit low er NK ac tivity compared w ith their left-activate d c ounte rparts because the former type of subje ct has been found to report more dispositional negative affect, to show hig her relative B IS activity and to respond more intensely to negativ e emotional stimuli. We found that rig ht-frontally activate d subje cts indee d had low er levels of NK activity compared to their left-frontally activate d counte rparts (Kang et al., 1991). Recently, in collaboration w ith Kalin, our laboratory has been study ing sim ilar individual differences in scalp-re corded measures of prefrontal activ ation asym metry in rhesus monke ys (Dav idson, Kalin, & Shelton, 19 92, 1993). R ecently, w e (Kalin, Larson, Shelton, & Davidson, in press) acquire d measures of brain electric al activity from a large sample of rhesus monke ys (N = 50). EEG measures w ere obtaine d during periods of manual restraint. A subsample of 15 of these monke ys w ere tested on tw o occasions four m onths apart. We found that the test-re test correlation for measures of prefrontal asym metry w as .62, sugg esting similar stability of this metric in monke y and man. In the group of 50 anim als, w e also obtaine d measures of plasm a cortisol during the early morning. We hy pothe sised that if individual differences in prefrontal asymmetry w ere associated w ith dispositional affec tive style, such differences should be correlate d with cortisol, because indiv idual differences in base line cortisol have been related to various aspe cts of trait-related stressful behaviour and psychopatholog y (see e.g. Gold, Goodw in, & Chrousos, 1988). We found that animals w ith rig ht-side d prefrontal activation had highe r levels of bas eline cortisol than their left-frontally activate d counte rparts (see Fig. 3). Moreove r, when blood sample s w ere colle cted two years follow ing our initial testing, animals clas si ed as show ing e xtreme right-side d prefrontal activ ation at age 1 year had signi cantly highe r baseline cortisol levels w hen they w ere 3 years of age compared w ith animals w ho were classi ed at age 1 year as displaying extreme left-sided prefrontal activation. These ndings indic ate that indiv idual diffe rences in prefrontal asy mmetry are present in nonhuman primates and that such differences predict biologic al measures that are related to affective style.

AFFECTIVE STYLE AND AFFECTIVE DISORDERS

319

FIG. 3.

Basal morning plasma cortisol from 1-year-old rhesus monkeys classi ed as left ( N = 12), middle ( N = 16), or right ( N = 11) frontally activated based upon electrophysiolo gical measurements. From Kalin et al. (in press).

V. AFFECTIVE STYLE AND PSYCHOPATHOLOGY

Virtually all forms of psychopatholog y involve some abnormality in emotional processes, although the nature of these abnorm alitie s is like ly to differ among different disorde rs. The study of precisely w hat is abnorm al in the emotion-proc essing systems of individuals w ith different forms of psychopathology is very much at the earlie st stage s of inve stigation. We have used our nding s in normal subje cts as a foundation to probe the unde rlying neural substrate s of affective and anxie ty dis orders w ith a major goal of unde rstanding more prec isely the nature of the abnorm ality in emotional-proc essing in affectiv e disorde rs. One of the important sources of data on relations betw een brain function and emotion has come from studie s of the affective styles of patie nts w ith localise d brain lesions (se e Robinson & Dow nhill, 19 95 for a review ). Robinson and his colle ague s have reported that damage to the left frontal region is more like ly to be associated w ith depression than damage to any other cortical region. Moreove r, among patie nts w ith left hemisphe re damag e, more severe depressive sym ptomatology is present in those patie nts w hos e damage is closer to the frontal pole (Robinson, Kubos, Starr, R ao, & Price , 19 84). Studie s of regional brain function with neuroimaging of patie nts with psyc hiatric depressions have fairly consi stently reveale d a pattern of decrease d blood ow or m etabolism in left prefrontal

320

DAVIDSON

regions at rest (Bax ter et al., 1989 ; Bench et al., 19 92; 199 3; Martinot et al., 19 90; se e George, Ketter, & Pos t, 19 94 for a review; se e also Drevets et al., 19 92b for a more complex pattern associated w ith pure familial depression). We have conduc ted several studies exam ining regional brain electrical activ ity in depression. We hy pothe sised that most depre ssion is fundam entally associated with a de cit in the approac h/appe titive motivation system and should therefore be speci cally accompanie d by decreased activation in the left prefrontal region as measured by scalp ele ctrophysiol ogy. Henrique s and Davidson (1991 ) obtaine d support for this hypothe sis. Moreove r, in anothe r study , the se authors demons trated that the decrease in left prefrontal activ ation found among depressives w as also present in recove red depressives w ho w ere currently euthymic , c ompared with ne ver depressed controls who were sc reened for lifetime history of psychopatholog y in both themselves as w ell as their rst degree relativ es (Henrique s & Davidson, 1990 ). Tog ether the ndings from patie nts w ith localise d unilateral brain damag e, as w ell neuroim aging and electrophysiology studie s in psyc hiatric patie nts w ithout frank lesions, conve rge on the notion that depression is associated w ith a de cit in at least the prefrontal compone nt of the approach system. We view this pattern of left prefrontal hypoac tivation as a neural re ection of the decreased capac ity for pleasure , loss of interest, and generalis ed decline in goal-re lated motivation and behaviour. Consi stent w ith this notion are the data from anothe r recent behavioural study from my laboratory where w e demonstrated using signal detection methods that depressed subje cts w ere speci cally hypore active to rew ard incentive s (Henriques , Glow acki, & Davidson, 1994). In this study, w e adminis tered a verbal memory task unde r rew ard, punis hment, and neutral incentive conditions. The rew ards and punis hments w ere monetary. Sig nal detection measure s of sensitivity and response bias were computed. Nondepressed control subje cts exhibite d a more libe ral response bias unde r both reward and punishm ent incentive s. In other w ords, they w ere more like ly to conside r a stimulus as a signal if they w ere rew arded for correct hits or punishe d for misses. Depressed subjec ts show ed a patte rn quite sim ilar to the controls in response to the punishm ent conting ency. How ever, they faile d to modify their response bias during the rew ard condition. In other words, the depressed subje cts we re less respons iv e to rewards compared with controls , how ever, the groups show ed no signi cant differences in response to punishme nt. B ased on the evide nce review ed earlie r, w e hypothe sised that in contrast to depression, anxie ty disorde rs w ould be associated with an increase in right-side d rather than a decrease in left-sided pre frontal activation, particularly during an acute episode of anxie ty. To test this hypothe sis, w e (Davidson, Marshall, Tomarke n, & Henriques , 1997 ) expose d social phobic s who w ere partic ularly fearful of making public speeches to the

AFFECTIVE STYLE AND AFFECTIVE DISORDERS

321

threat of having to m ake a public speech. We recorded brain electrical activity during an antic ipation phas e w here subje cts w ere presented w ith an audiotape d countdow n that noted how much more time there w as until they we re to make their speech. The tape d-re corded message was presented every 30 seconds for a total of 3 minute s. We found that the phobic s show ed a large and highly signi cant increase ov er base line in right-si ded prefrontal and right-side d parie tal activ ation. During the same antic ipation period, the controls show ed a very different patte rn of regional chang es. The only change to reach signi cance w as in the left poste rior temporal region. We interpret this latter change as like ly a consequence of verbal rehearsal in antic ipation of making the public speech. No region in the right hemisphe re exceeded an even libe ral statis tical threshold for increased activation relative to a base line condition. The change in prefrontal activation among the phobic s is consis tent with our hypothe sis of increased rig htsided activation associate d w ith an increase in anxie ty. The increase in rig ht parie tal activation is c onsi stent w ith Heller and Nitschke s (this Issue) hypothes is of increased right-s ided activation associated w ith the arousal compone nt of anxie ty. Indeed, simultaneous measures of heart rate in this study indic ated that the phobic s had highe r heart rate compared w ith the controls, partic ularly during the antic ipation phase . Research using self-report measure s of positiv e and negative affect as we ll as expe rienced increase s in autonomic arousal indic ate that decreased positive affect is unique ly assoc iated w ith depression, w hereas increased autonom ic arousal is unique ly associated with anxie ty. How ever, reported negativ e affect is something that has been found to be common to both anxie ty and depression (Watson et al., 1995). We have hypothe sised that the decrease in left prefrontal activation may be speci c to depression, whereas the increase in right-side d prefrontal activation (as w ell as right parie tal activation) may be spec i c to certain compone nts of anxie ty. Conside rably more research is require d to understand the contribution being made by the activate d right prefrontal region to negative affect. Other w ork (see Posner & Petersen, 1990 for a review ) indic ates that portions of the right prefrontal region are activate d during certain type s of vigilanc e and attention (e.g. Knight, 19 91; see Pos ner & Petersen, 1990 for a review). Anxie ty-related negative affe ct is accompanie d by heightened vigilanc e (e.g . McNally , this Issue) which may be re ected in the rig ht prefrontal increase. One comm on region w e believe to be assoc iated w ith both anxie ty and depression is the amy gdala. Even though there is a burge oning literature on the anatom y and func tion of the amy gdala (see Aggleton, 1993 for a review), relative ly little research has been conduc ted in intac t humans, ow ing in large measure to the dif culty in imaging function in a structure that is relativ ely small (the adult human amygdala is approximate ly 1cm in

322

DAVIDSON

volume ). However, from w hat is know n from both the anim al and human studie s, it appe ars that the amygdala plays an important role in assigning affective signi cance, partic ularly of negative vale nce, to both sensory as w ell as cognitive input (see LeDoux , 1992 for a review ). Using positron emission tomography (PET) to measure regional blood ow , several groups have reported increased blood ow in the amygdala in response to both behavioural (e.g. Schneider et al., 1995 ) and pharm ac ologic al (e.g. Ketter et al., 1996) elicitors of negative affect. We have recently reported activation in the human amyg dala using functional mag netic resonance imag ing (fMR I) in respons e to ave rsive pictures (Irwin et al., 1996). These studie s sugge st that activation in the human amyg dala occurs in respons e to a broad rang e of elicitors of negative affect. 15 B oth fMR I and O PET are ill-sui ted, for diff erent reasons, for exam ining individual diffe rences in resting or baseline levels of activation in the amygdala. As it is currently used, fMRI requires that at least tw o conditions be compare d. What is measured is a relativ e dif ference in MR sig nal intensity between tw o or more conditions. Currently, fMRI is not cali15 brated in real physiolog ical units. Although O PET can be calibrate d in real units, it re ects activity ov er a very short period of time (approximately one minute ) and thus , for psyc hometric reasons, is poorly suited to capture trait-like diffe rences. (It w ould be the equivale nt of developing a sing le-item self-report ins trument for assessing individual diffe rences.) PET used w ith uorode ox ygluc ose (FDG) as a tracer, on the other hand, is w ell-suite d to capture trait-like effects because the period of activ e uptake of tracer in the brain is approxim ately 30 minute s. Thus, it is inhe rently more reliable as the data re ect activity aggre gate d ov er this 30 -minute period. We have used resting FDG-PET to exam ine individual differences in glucose metabolic rate in the amygdala and its relation to disposi tional negative affect in depressed subje cts (Abercrombie et al., submitte d). We acquire d a resting FDG-PET scan as w ell as a structural MR scan for each subje ct. The structural MR scans are used for anatomic al localis ation by coregistering the tw o image sets. Thus, for each subjec t, w e used an autom ated algorithm to t the MR scan to the PET image . Regions of interest (ROIs) w ere then draw n on each subje ct s MR scan to outline the amyg dala in e ach he misphe re. These ROIs w ere draw n on coronal sections of subje cts MR image s and the R OIs w ere then autom atically transfe rred to the core gis tered PET image s. Glucose metabolis m in the left and rig ht amyg dala ROIs were then extrac ted. The inter-rater reliability for the e xtrac ted gluc ose m etabolic rate is highly signi cant w ith intrac lass correlations between tw o inde pendent raters .97. Figure 4 (situate d on p. 338 of the C olour Plate Section) illustrate s R OIs draw n around the amyg dala on MR scans of three subje cts and the associated coregistered PET imag es from the same subje cts. We found that subje cts with greater gluc ose

AFFECTIVE STYLE AND AFFECTIVE DISORDERS

323

metabolism in both the right and left amy gdala report greater disposition al negativ e affect on the PANAS scale (see Fig. 5). These ndings indic ate that individual diff erences in resting gluc ose metabolism in the amyg dala are present and that they predict dis positional negative affect among depressed subje cts. Most nonde pressed controls score so low on the PANAS trait negativ e scale that it is not possible to e xamine the same relation in this group because of the severe truncation of range for the PANAS scores. The ndings reviewed in this section indic ate that the framew ork adopte d for the study of individual differences in fundame ntal approac h and w ithdraw al-related processes can be usefully applie d in the study of psychopathology. A de cit in the approac h system is view ed as a unique attribute of depressive dis orders that is re ected in decreased left prefrontal activation. The acute symptom s of anxie ty, as w as described in our study with social phobic s, w as associated w ith a pronounc ed increase in both rig ht-side d prefrontal and parie tal activation. From research conduc ted in our laboratory as w ell as recent nding s in the literature, it appe ars that

FIG. 5.

Scatter plot of correlations between dispositional negative affect assessed with the PANAS-General Negative Affect Scale and PET M RI-c oregistration-deriv ed re gional glucose metabolism in the left and right amygdalae for all subjects ( N = 17). Subjects we re tested on two different PET cameras. Those teste d in a Sie mens CTI 93 3/04 PET came ra are displayed by close d squares and those tested in the GE Advance are de picted by the open squares. From Aberc rombie et al. (submitted).

324

DAVIDSON

amygdala activation may be a generic compone nt of negative affect that is present in both anx iety and depression. Thus, differences between these disorde rs may be more pronounc ed for cortical syste ms that are critically involve d in affect regulation and affect-cognition interac tion, whereas subcortical contribution (in partic ular, the amy gdala) may be common to both type s of disorde rs and may in part be responsible for the substantial comorbidity betw een these disorde rs (Watson et al., 19 95).

VI. IMPLICATIONS AND CONCLUSIONS

Earlie r in this article, the cons tituents of affective style w ere described. We cons idered individual differenc es in threshold, peak amplitude , rise time to peak, and recove ry time . Toge ther, these constitute parame ters of affectiv e chronome try and dictate important features of the time course of affectiv e responding. Follow ing a description of the functional neuroanatom y of the approac h and w ithdraw al systems, individual differences in prefrontal activ ation asy mmetry w ere discussed and their relation to affective style and psychopatholog y desc ribed. In the light of this information, w e now return to the question pose d in the title of Section IV : What do individual differences in prefrontal asymmetry re ect? On the basis of ndings from several new studie s in my laboratory, I sugge st that at least one important compone nt of w hat prefrontal cortex ``does in affective responding is modulate the time course of emotional responding, partic ularly recove ry time. There are several facts critical to making this claim. First, there are extensive reciproc al conne ctions betw een amygdala and the prefrontal cortex (PFC), particularly the medial and orbital zones of the PFC (Amaral, Price, Pitkane n, & C armichael, 19 92). The glutamate rgic e fferents from the PFC like ly synapse on GABA (gamma-aminobutyric acid) neurone s (Amaral et al., 1992 ) and thus provide an important inhibitory input to the amygdala. Second, LeDoux and his colle ague s (Morgan, R omanski, & LeDoux, 19 93; but see Gew irtz, Falls, & Davis, 1997 ) dem onstrate d in rats that lesions of medial prefrontal cortex dramatic ally prolong the mainte nance of a conditione d ave rsive response . In other words, animals w ith medial prefrontal lesions retain ave rsiv e associations for a much longe r duration of time than normal anim als. These nding s imply that the medial PFC normally inhibits the amygdala as an active compone nt of extinc tion. In the abs ence of this normal inhibitory input, the amyg dala remains unchecked and continue s to maintain the learned ave rsive respons e. Third, are the data from my laboratory cited in Section IV , indic ating that individual differences in prefrontal activation asy mmetry signi cantly predict the magnitude of the post-stim ulus startle follow ing removal of the varianc e attributable to startle magnitude during the presentation of the emotional

AFFECTIVE STYLE AND AFFECTIVE DISORDERS

325

picture. In partic ular, left prefrontal activation appe ars to facilitate tw o processes simultane ously: (1) it maintains represe ntations of behav iouralreinforc ement continge ncies in working mem ory (Thorpe et al., 1983 ); (2) it inhibits the amygdala. In this w ay, the time course of negative affect is shortened whe reas the time course of pos itive affect is accentuate d. Finally, fourth, new nding s using PET from my laboratory indic ate that in norm al subje cts, gluc ose metabolism in left medial and late ral prefrontal cortex is strongly associated reciprocally w ith gluc ose metabolic rate in the amygdala (Abercrombie et al., 1996). Thus , subje cts w ith greater left-sided prefrontal metabolis m hav e low er metabolic activity in their amygdala. These ndings are consiste nt with the lesion study of LeDoux and colle ague s and imply that prefrontal cortex plays an important role in modulating activity in the amygdala. At the same time, the left prefrontal cortex is also like ly to play a role in the mainte nanc e of reinforc ement-relate d behavioural approach. Perhaps the damping of negative affect and shortening of its time course facilitate s the mainte nanc e of approac h-related positive affect. The questions that are featured in this article are more tractable now than ever before. With the adv ent of echoplanar methods for rapid functional magnetic resonanc e imaging (MRI), suf cient data can be collec ted within individuals to examine func tional conne ctions among regions hypothes ised to constitute important elements of the approac h and w ithdraw al circuits discusse d earlie r. Individual differences in different aspe cts of these systems can then be studie d with greater precision. Functional magne tic resonanc e imaging (fMR I) methods als o lend themselves to addre ss questions related to affectiv e chronom etry. In partic ular, w e can calculate that the slope of MRI signal inte nsity decline s follow ing the offse t of an av ersive stimulus to provide an inde x of the rapidity of recove ry from activation in select brain regions . Positron emission tomography (PET) methods using new radioligands that permit quanti cation of receptor density for speci c neurotrans mitters in different brain regions is yielding ne w insights directly relevant to questions about affective style (see e.g . Farde, Gustav sson, & Jo nsson, 19 97). Trait-like differences in affective style are like ly re ected in relative ly stable differences in characteristics of the underlying neurochemical systems. Using PET to exam ine such individual differences promise s to prov ide important synthe ses between neurochemical and neuroanatomic al approac hes to unde rstanding the biologic al bas es of affective style. Affective neuroscience seeks to unde rstand the unde rlying prox imal neural subs trates of elementary cons tituents of emotional processing. In this article, I have provide d a model of the functional neuroanatom y of approac h and withdraw al m otivational/e motional systems and illustrate d the many varie ties of individual diffe rences that might occur in these

326

DAVIDSON

syste ms. R esearch on prefrontal asym metries associate d w ith affectiv e style and psyc hopathology was used to illustrate the pote ntial promise of some initial approac hes to the study of these questions. Modern neuroim aging methods used in conjunc tion with theoretically sophis ticated models of emotion and psy chopathology offer great promise in advanc ing our unde rstanding of the basic mechanism s giving rise to affective style and affective and anxie ty disorde rs.
M anusc ript received 17 June 1997

REFERENCES
Abercrombie, H.C., Larson, C.L., Ward, R.T., Sc haefer, S.M., Holden, J.E., Perlman, S.B., Turski, P.A., Krahn, D.D., & Davidson, R.J. (submitted). Metabolic rate in the amygdala predic ts ne gative affect and depression severity in depressed patie nts: A FDG-PET Study. Abercrombie, H.C., Schae fer, S.M., Larson, C.L., Ward, R.T., Holden, J.F., Turski, P.A., Perlman, S.B . & Davids on, R.J. (19 96). Me dial prefrontal and amygdalar glucose metabolism in depressed and control subjec ts: An FDG-PET study. Psychophysiology, 33 , S17. Aggleton, J.P. (1993). The contribution of the amygdala to normal and abnormal emotional states. Trends in Neuroscience . 16 , 328 333. Amaral, D.G., Price, J.L., Pitkanen, A., & Carmichael, S.T. (1992). Anatomical organization of the primate amygdaloid complex. In J.P. Aggleton (Ed.), The amygd ala : Neu robiolog ical aspects of emotion, me mory and mental dysfunction (pp. 1 66). New York: Wiley -Liss. Baxter, L.R., Schw artz, J.M., Phelps, M.E., Mazziota, J.C., Guze, B.H., Selin, C.E., Gerner, R.H., & Sum ida, R.M. (198 9). Re duction of prefrontal corte x glucose metabolis m common to three types of depression. Archives of General Psychiatry, 46 , 243 252. Bench, C.J., Friston, K.J., Brown, R.G., Frackowiak, R.S.J., & Dolan, R.J. (1993). Regional cerebral blood ow in de pression measured by pos itron emission tomography : the relationship with clinical dimensions. Psychological Medicine , 23 , 579 590. B e nc h, C.J., Friston, K.J., B row n, R.G., S c ott, L.C ., Frac kow iak, S .J., & Dolan, R.J. (1992). The anatomy of melancholi a-focal abnormalitie s of cerebral blood ow in major depre ssion. Psychologic al Medicine , 22 , 607 615. Carver, C.S., & White, T.L. (1994). Behavioral inhibition, behavioral activation and affective responses to impending reward and punishment: The BIS/BAS scales. Journal of Personality and Social Psycholo gy, 67 , 319 333. Davidson, R.J. (1992). Emotion and affective style: Hemispheric substrates. Psychological Science , 3 , 39 43. Davidson, R.J. (1993). Cerebral asymmetry and emotion: Conceptual and methodological conundrums. Cognition and Emotion , 7 , 115 138. Davidson, R.J. (19 94). Asymmetric brain function, affective style and psychopathology : The role of early experience and plasticity. De velopment and Psychopathology, 6 , 741 758. Davidson, R.J. (1995). Cerebral asymmetry, emotion and affective style. In R.J. Davidson & K. Hugdahl (Eds.), Brain asymmetry (pp. 361 387). Cambridge, MA: MIT Press. Davidson, R.J., Chapman, J.P., Chapm an, L.P., & Henrique s, J.B. (1990). Asymmetrical brain electrical activity disc riminate s betw een psychome tric ally-matche d verbal and spatial cognitive tasks. Psychophysiology, 27 , 528 543.

AFFECTIVE STYLE AND AFFECTIVE DISORDERS

327

Davidson, R.J., Dolski, I., Laron, C., & Sutton, S.K. (in prep.). Electrophy siolog ic al measures of prefrontal asymmetry predict recovery of emotion-modulate d startle . Davidson, R.J., Ekman, P., Saron, C., Senulis, J., & Friesen, W.V . (1990a). Approach/ withdraw al and cerebral asymmetry: Emotional expression and brain physiolog y: I. Journal of Personality and Social Psychology, 58 , 330 341. Davidson, R.J., & Fox, N.A. (1989). Frontal brain asymmetry predicts infants response to maternal separation. Journal of Abnormal Psychology, 98 , 127 131. Davidson,R .J.,K alin,N .H.,& Shelton ,S.E. (1992). Lateralizedeffects of diazepamon frontal brain electrical asymmetries in rhesus monkeys. Biological Psychiatry, 32 , 438 451. Davidson,R .J.,K alin,N.H .,& Shelton,S .E. (1993). Lateralizedresponseto diazepampredicts temperamentalstyle in rhesus monkeys. Behavioral N euroscience, 107 , 1106 1110. Davidson, R.J., M arshall, J.R., Tomarken, A.J., & Henriques, J.B. (1997). While a phobic waits: Regional brain elec trical and autonomic activity predict anxiety in social phobics during anticipation of public speaking. Submitted for publication. Davidson, R.J., & Tomarken, A.J. (19 89). Laterality and emotion: An electrophysiological approach. In F. Bolle r & J. Grafman (Eds.), Handbook of neuropsycholo gy. Amsterdam: Elsevier. Depue , R.A., & Collins, P.F. (in press). Neurobiology of the structure of pe rsonality: Dopamine , facilitation of inc entive-reward motiv ation, and extraversion. Behavioral and Brain Sciences . Derrybe rry, D., & Re ed, M.A. (1 996). Re gulatory proce sses and the de velopme nt of cognitiv e representations. Development and Psychopathology, 8 , 215 234. Diener, V.E., & Emmons, R.A . (1984). The independence of positive and negativ e affec t. Journal of Personality and Social Psychology, 47 , 1105 1117. Drevets, W.C., Videen, T.O., M acLeod, A.K., Haller, J.W., & Raichle, M.E. (1992a). PET images of blood changes during anxie ty: Correction. Science , 256 , 1696. Drevets, W.C., Vide en, T.O., Price, J.L., Preskorn, S.H., Carmichael, S.T., & Raic hle, M.E. (1992b). A func tional anatom ic al study of unipolar depression. Journal of Neuroscience , 12 , 3628 3641. Farde, L., Gustavsson, J.P., & Jo nsson, E. (199 7). D2 dopamine receptors and personality. Nature , 385 , 590. Ge org e , M.S ., Ke tte r, T.A., Pare kh, P.I., Horw i tz , B ., He rs ov i tc h, P., & Pos t, R.M. (1995). Brain activity during transient sadness and happiness in he althy women. American Journal of Psychiatry, 152 , 341 351. George, M.S., Ketter, K.A., & Post, R.M. (1994). Prefrontal cortex dysfunction in clinical depression. Depression , 2 , 59 72. Gew irtz , J.C., Falls, W .A., & Davis, M. (19 97). Normal conditione d inhibitio n and extinc tion of freezing and fear-potentiated startle follow ing ele ctrolytic lesions of medial prefrontal cortex in rats. Behavioral Neuroscience , 111 , 712 726. Gold, P.W., Goodw in, F.K., & Chrousos, G.P. (1988). Clinic al and biochemic al manife stations of depression: Relation to the ne urobiology of stress. Ne w England Journal of Medicine , 314 , 348 353. Gray, J.A. (1994). Three fundamental emotion systems. In P. Ekman & R.J. Davidson (Eds.), The nature of emo tion : F undame ntal questio ns (pp. 243 247). New York: Oxford Unive rsity Press. Gross, J.J., Sutton, S.K., & Ketelaar, T.V. (in press). Relations betw een affect and personality: Support for the affect-le vel and affective -reactivity vie ws. Personality and Social Psychology Bulle tin . Henriksen, S.J., & Giacchino, J. (1993). Func tional characteristic s of nucle us accumbens neurons: Evidence obtained from in vivo electrophysiologic al recordings. In P.W. Kalivas

328

DAVIDSON

& Barnes, C.D. (Eds.), Lim bic motor circuits and neuropsychiatry (pp. 101 124). Boca Raton, FL: CRC Press. Henriques, J.B., & Davidson, R.J. (1990). Regional brain electrical asymmetries discriminate betwe en previously depressed and healthy control subjects. Journal of Abnormal Psychology, 99 , 22 31. Henriques, J.B., & Davidson, R.J. (1991). Left frontal hypoactivation in depression. Journal of Abnormal Psychology , 100 , 535 545. Henriques, J.B., Glow acki, J.M ., & Davidson, R.J. (1994). Rew ard fails to alter response bias in depre ssion. Journal of Abnormal Psychology , 103 , 460 466. Irw i n, W ., Dav i ds on, R.J., Low e , M.J. , Moc k, B .J., S ore ns on, J. A., & Turs ki , P.A. (1996 ). Human amygdala activation dete cted with echo-planar functional magnetic resonance imaging. NeuroReport, 7 , 1765 1769. Kagan, J., Reznick, J.S., & Snidman, N. (1988). Biologic al bases of childhood shyness. Science , 240 , 167 171. Kalin, N.H. (1993). The neurobiology of fe ar. Scienti c American , 268 , 94 101. Kalin, N.H., Larson, C., Shelton, S.E., & Davidson, R.J. (in press). A symmetric frontal brain activity, cortisol, and behaviour associated with fearful tempe rament in Rhe sus monkeys. Behavioral Neuroscience . Kaliv as, P.W., Churc hill, L., & Klitenick, M.A. (1993). The circuitry mediating the translation of motiv ational stimuli into adaptive motor responses. In P.W. Kalivas & Barnes, C.D. (Eds.), Limbic motor circuits and neuropsychiatry (pp. 237 287). Boca Raton, FL: CRC Press. Kang, D.H., Davids on, R.J., Coe , C.L., Whe ele r, R.W ., Tomarken, A.J., & Ershl er, W .B . ( 1991 ). Frontal brai n as ymme try and i mmune func ti on. Beh av io r al Ne uroscience , 105 , 860 869. Ketter, T.A., Andre ason, P.J., George, M.S., Lee, C., Gill, D.S., Parekh, P.I., Willis, M.W., Herscovitch, P., & Post, R.M. (1996). Anterior paralimbic mediation of proc aineinduced emotional and psychose nsory experie nces. Archives of General Psychia try, 53 , 59 69. Kiecolt-Glaser, J.K., & Glaser, R. (1991). Stress and immune function in humans. In R. A der, D.L. Felten, & N. Cohe n (Eds.), Psychoneuroimmunology (2nd ed.) (pp. 849 868). San Diego, CA: Academic Press. Knight, R.T. (1991). Evoke d potential studies of attention capacity in human frontal lobe le sions. In H.S. Levin, H.M. Eisenberg, & A.L. Benton (Eds.), Frontal lobe function and dysfunction (pp. 139 153). New York: Oxford University Press. Lang, P.J., Bradle y, M.M ., & Cuthbert, B.N. (1990). Emotion, attention and the startle re ex. Psychological Revie w , 97 , 377 398. Lang, P.J., Bradley, M., & Cuthbert, B. (1995). Internatio nal Affe ctiv e Picture System: Technical manual and affectiv e ratings . Gainesville FL: Center for Research in Psychophysiology, Univ ersity of Florida. Lazarus, R.S. (1 991). Emotio n and adaptation . Oxford, UK: Oxford University Press. LeDoux, J.E. (1987). Emotion. In V.B. Mountcastle (Ed.), Handbook of physiology, Section 1: The nervous system: Vol. V. Higher functions of the brain (pp. 419 459). Bethesda, MD: American Physiologic al Society. LeDoux, J.E. (1992). Emotion and the amygdala. In J.P. Aggleton (Ed.), The amygdala: Ne ur obio lo gi cal asp ects of e mo ti on, me mo ry and me n ta l dysfu nc ti on (pp. 339 351). New York: Wile y-Liss. Martinot, J.L., Hardy P., Feline, A., Huret, J.D., Maz oye r, B., Attar-Levy, D., Pappata, S., & Syrota, A. (1990). Left frontal gluc ose hypometabolism in the de pressed state: A con rmation. American Journal of Psychiatry, 147 , 1313 1317.

AFFECTIVE STYLE AND AFFECTIVE DISORDERS

329

Me ehl, P.E. (1975). Hedonic capacity: Some conjecture s. Bulletin of the Menninger Clinic , 39 , 295 307. Morgan, M.A., Romanski, L., & LeDoux, J.E. (1993). Extinction of emotional learning: Contribution of medial prefrontal corte x. Ne uroscience Letters , 163 , 109 113. Paradiso, S., Robinson, R.G., Andreason, N.C., Downhill, J.E., Davidson, R.J., Kirchner, P.T., Watkins, G.L., Boles, L.L., & Hichw a, R.D. (1997). Emotional activation of limbic circuitry in elderly and normal subje cts in a PET study. Am erican Journal of Psychiatry, 154 , 382 389. Posner, M.I., & Petersen, S.E. (1990). The attention system of the human brain. Annual Review of Neuroscience , 13 , 25 42. Reiman, E.M., Fusselman, M.J.L., Fox, B.J., & Raichle , M .E. (1989). Neuroanatomical correlates of anticipatory anxiety. Science , 243 , 1071 1074. Robinson, R.G., & Downhill, J.E. (1995). Lateralization of psychopathology in response to focal brain injury. In R.J. Davidson & K. Hugdahl (Eds.), Brain asymmetry (pp. 693 711). Cambridge, MA: MIT Press. Robinson, R.G., Kubos, K.L., Starr, L.B., Rao, K., & Price, T.R. (198 4). M ood disorders in stroke patients: Importance of location of lesion. Brain , 107 , 81 93. Russell, J.A. (19 80). A circumplex mode l of emotion. Journal of Personality and Social Psychology, 39 , 1161 1178. Ryff, C.D., & Singer, B. (in press). The contours of positive human he alth. Psychological Inquiry. Schneider, F., Gur, R.E., M ozley, L.H., Smith, R.J., Mozley, P.D., Censitis, D.M., Alavi, A., & Gur, R.C. (1995). Mood effects on limbic blood ow correlate with emotional selfrating: A PET study with oxygen-15 labeled water. Psychiatric Research: Neuroimaging , 61 , 265 283. Schultz , W., Apic ella, P., Romo, R., & Scarnati, E. (1995a). Context-depende nt activity in primate striatum re ecting past and future behavioral events. In J.C. Houk, J.L. Davis, & Beiser, D.G. (Eds.), Models of informatio n processing in the basal ganglia (pp. 11 28). Cambridge, MA: MIT Press. Schultz , W., Romo, R., Ljungberg, T., Mire now icz , J., Holle rman, J.R., & Dickinson, A. (1995b). Rew ard-related signals carried by dopamine ne urons. In J.C. Houk, J.L. Davis, & Beiser, D.G. (Eds.), Models of informatio n processing in the basal ganglia (pp. 233 248). Cambridge, MA: MIT Press. Smith, O.A., DeVito, J.L., & Astley , C.A. (1 990). Neurons controlling cardiovascular responses to emotion are located in lateral hypothalamus-peri fornic al region. American Journal of Physiology , 259 , R943 R954. Stein, N.L., & Trabasso, T. (1992). The organization of emotional experience: Creating links among emotion, thinking, language and intentional action. Cognition and Emotion , 6 , 225 244. Sutton, S.K., & Davidson, R.J. (1997). Prefrontal brain asymmetry: A biological substrate of the behavioral approach and inhibition systems. Psychological Science , 8 , 204 210. Sutton, S.K., Davidson, R.J., Donzella, B., Irwin, W., & Dottl, D.A. (1997). M anipulating affective state using extended picture presentation. Psychophysiology, 34 , 217 226. Thompson, R.A . (1994). Emotion regulation: A theme in search of de nition. In N.A. Fox (Ed.), The develop men t of emo tio n regula tion : Biolo gic al and behavio ral aspects. Monographs of the Society for Research in Child Development, 59 , (Se rial No. 240), 25 52. Thorpe, S., Rolls, E., & M addison, S. (1983). The orbitofrontal cortex: Neuronal activity in the behaving monkey. Experimental Brain Research , 49 , 93 113. Tomarken, A.J., Davidson, R.J., Wheeler, R.E., & Doss, R.C. (1992). Individual diffe rences

330

DAVIDSON

in anterior brain asymmetry and fundamental dimensions of emotion. Journal of Personality and Social Psychology , 62 , 676 687. Wats on, D., Clark, L.A., We ber, K., Assenheimer, J.S., Strauss, M.E., & McCormick, R.A. (1995). Testing a tripartite model: I. Evaluating the converge nt and discriminant validity of anxiety and de pression symptom scales. Journal of Abnormal Psychology, 104 , 3 14. Watson, D., Clark, L.A., & Tellegen, A. (1988). Developmental and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54 , 1063 1070. Watson, D., & Tellegen, A. (1985). Toward a consensual structure of mood. Psychological Bulle tin , 98 , 219 235. Whee le r, R.E., Davidson, R.J., & Tomarken, A.J. (1993). Frontal brain asymmetry and emotional reactivity: A biological substrate of affe ctiv e style. Psychophysiology , 30 , 82 89. Woods, R.P., Mazziotta, J.C., & Cherry, S.R. (1993). M RI-PET registration with automate d algorithm. Journal of Computer Assisted Tomography, 17 , 536 546.