Modern Psychological Studies

Volume 4 Number 1 Article 3

1996

Age-related dendritic changes in human occipital and prefrontal cortices: a quantitative Golgi study

Kelly A. Courns Colorado College

Bob Jacobs Colorado College

Follow this and additional works at: https://scholar.utc.edu/mps

Part of the Psychology Commons

Recommended Citation Courns, Kelly A. and Jacobs, Bob (1996) "Age-related dendritic changes in human occipital and prefrontal cortices: a quantitative Golgi study," Modern Psychological Studies: Vol. 4 : No. 1 , Article 3. Available at: https://scholar.utc.edu/mps/vol4/iss1/3

This articles is brought to you for free and open access by the Journals, Magazines, and Newsletters at UTC Scholar. It has been accepted for inclusion in Modern Psychological Studies by an authorized editor of UTC Scholar. For more information, please contact [email protected]. Modem Psychological Studies Copyright 1996 by 1996, Vol. 4, No. 1, pp. 10-20 Modem Psychological Studies

Age-related Dendritic Changes in Human Occipital and Prefrontal Cortices: A Quantitative Golgi Study

Kelly A. Courns and Bob Jacobs Laboratory of Quantitative Neuromorphology Department of Psychology Colorado College

Qualitative (Scheibe!, 1992) and quantitative (Jacobs & Scheibel, 1993) research indicates a general decline in dendritic neuropil with increasing age. The present study extends previous human dendritic research by examining quantitatively age-related changes in 2 cortical areas: prefrontal cortex (area 10) and occipital cortex (area 18). Tissue blocks were obtained from the left hemisphere of 10 neurologically normal subjects, ranging in age from 23 to 81 years. Blocks were stained with a modified rapid Golgi technique. Supragranular pyramidal cells were quantified on a Neurolucida computer/microscope interface system (Microbrightfield, Inc.). Dendritic system complexity was determined by several dependent measures: total dendritic length, mean dendritic length , dendritic segment count (DSC), dendritic spine number, and dendritic spine density. All dependent measures, except DSC, decreased with age, with a substantial (approximately 50%) decrease in dendritic spines. Although area 10 exhibited greater dendritic aborizations than area 18, dendritic declines were slightly more pronounced in area 10 than in area 18. The present results quantitatively document the ongoing, dynamic refinement of dendritic systems across the human life span, and suggest that higher order cortical areas (e.g., area 10) may be more susceptible to age-related changes.

Those cognitive changes in processing speed and Other age-associated histological changes have style that develop, benignly during the process of been observed across several cortical areas. Declines in normal aging, and more catastrophically in dementias such as Alzheimer's disease, can be related in large brain weight, cortical thickness, and in the number of measure to the loss of dendritic surface, with resultant large have been observed in midfrontal, disappearance of synaptic connections and superior temporal, and inferior parietal areas of aged progressive restriction in the computational power of human brains (Terry , DeTeresa, & Hansen, 1987). the brain. (Scheibel, 1992, p. 159) More specifically, a 13-20% reduction in area, which appears to be related to a reduction in dendritic Several age-related neural changes have been systems, has been revealed in individuals between the documented in both nonhuman and human animals ages of 70 and 85 years in the superior temporal gyms (Flood & Coleman, 1988). In particular, qualitative (Anderson, Hubbard, Coghill, & Slidders, 1983). (Scheibel, 1992) and quantitative (Jacobs & Scheibel, Similar decreases in density have been observed 1993) analyses reveal age-related decreases in dendritic in human somatosensory and visual cortices (Flood & complexity. Scheibel noted that aged human brains exhibited irregular swelling of the soma, loss of dendritic spines, and decreases in both basilar and apical I would like to thank the Associated Colleges of the dendritic systems. Age-related declines in basilar Midwest for developing the Minority Scholar Program, which gave me the invaluable opportunity to do the dendritic length have also been quantitatively observed present research. I am grateful to Dr. Bowerman, the El in Wernicke's area (Jacobs & Scheibel). More recently, Paso County Coroner, and Dr. Sherwin from Penrose Baca et al. (1995) documented age-related decreases in Hospital for providing brain tissue. I would also like to basilar dendritic length and dendritic spine number in thank Lori Larsen, Serapio Baca, and Jonathan Driscoll for preparing me for the present study. Lastly, several human cortical areas and suggested that age- I thank my family for their continued support. Partial related changes may not be uniform across all cortical support for this work was provided by the National regions. Science Foundation Division of Undergraduate Education through grant DUE - #9550790.

10 MODERN PSYCHOLOGICAL STUDIES AGE-RELATED DENDRITIC CHANGES IN HUMAN CORTEX

Coleman, 1988). changes in two cortical areas: prefrontal cortex (area 10) Several indirect measures further suggest decreases and occipital cortex (area 18). These two regions were in dendritic systems with age. Chugani, Phelps, and chosen because preliminary results (Baca et al., 1995) Mazziotta (1987) found lower local cerebral metabolic indicate that cortical areas higher in the information rates for glucose (CMRG1c) in adults than in children. processing hierarchy—as determined by Benson's These changes were specifically associated with age- (1994) functional schema—may be preferentially related alterations in dendritic systems (Jacobs et al., susceptible to age-related changes. Age-related decreases 1995; Jacobs & Scheibel, 1993). Age-related decreases in dendritic complexity and especially in dendritic spine in mean CMRGIc (Yoshii et al., 1988) and decreases in number/density are expected in both cortical areas. cerebral blood flow (Kety, 1956; Leenders et al., 1990; However, it is predicted that unimodal association Melamed, Lavy, Bentin, Cooper, & Rinot, 1980) have cortex (area 18), which is involved in basic visual been documented along with decreases in blood volume. processing, will be less affected by aging than Such metabolic changes parallel declines in synaptic supramodal association cortex (area 10), which is density with increasing age (Leenders et al.). More involved in executive (cognitive) control of mental directly, age-related decreases in the synaptophysin functions. The magnitude of age-related decreases is content have been observed in older rat brains (Saito et expected to be greater in supramodal cortex because, al., 1994), a decrease that correlates closely with loss of with increasing age, higher order cortical areas may not synapses (Masliah, Terry, Alford, DeTeresa, & Hansen, be as fully utilized as cortices involved in more basic 1991). Bondareff and Geinisman (1976) and Feldman (e.g., visual) processing. (1976) found a decrease in deafferented dendritic spines, indicating a correlation between loss of synapses and Method decreases in dendritic spine number. Apart from Huttenlocher's (1979) intensive Subjects developmental work, the present study is the first to examine quantitatively age-related changes in dendritic Tissue was removed from the left hemisphere of 10 spines across the human life span. The current project neurologically normal subjects, ranging in age from 23 extends previous human dendritic research (Jacobs & to 81 years (M = 50±19.5; see Table 1). The average Scheibel, 1993; Larsen, Swanson, Wainwright, & autolysis time was 10.4±4.6 hours. Brains were divided Jacobs, 1994) by specifically examining age-related into a younger group (n = 5; 23-48 years of age; M = 33.8±9.0) and an older group (n = 5; 51-81 years of Table 1 age; M = 66.2±10.8). Brains were obtained from the County Coroner's office and a local hospital. The Subject summary research protocol was approved by The Colorado College Human Subjects Institutional Review Board Age Gender Race Autolysis Cause of death (#H94-004). (years) time (hours) Apparatus

23 M C 12.0 Motor vehicle accident (chest trauma) All cells were quantified on a Neurolucida computer/microscope interface system 32 F C 20.0 Bulimia (Microbrightfield, Inc.) with an Olympus BH-2 32 M C 9 .0 Suicide (asphyxia) microscope under a 40X (.70) dry objective. 34 F C 14.5 Heroine overdose Histological Techniques 48 F C 4.0 Myxoid leiomyosarcoma 51 F C 6.0 Respiratory distress/ Tissue was immersion fixed in 10% neutral emphysema buffered formalin for 2-4 weeks before staining. One 3- 64 F C 11.0 Myocardial infarction 5 mm tissue block was removed from the frontal pole 66 M C 9.0 Pneumonia/cardiac (area 10) and the lateral surface of the occipital lobe arrest (area 18). Coded tissue blocks were stained using a 69 M C 7.0 Prostate cancer modified rapid Golgi technique (Scheibel & Scheibel, 1978). Adjacent cortical blocks were stained with a 81 M C 11.0 Gastro-intestinal bleeding cresyl echt violet technique (Gridley, 1960) to determine laminar depth and cortical thickness. Golgi-

MODERN PSYCHOLOGICAL STUDIES 11 Kelly A. Courns and Bob Jacobs stained tissue was serially sectioned at 120 tm on a systems, Pearson product correlations across TDL, vibratome such that sections were perpendicular to the soma size, and DSN averaged 0.98, indicating high long axis of the gyrus. Tissue sections were then agreement between raters. An ANOVA (a = 0.5) serially mounted on glass slides with permount and indicated no significant difference between raters on coverslipped. these measures.

Cell Selection Criteria and Quantification Design Techniques Cortical area (10 & 18) and age group (younger vs. Supragranular pyramidal cells from each tissue older) constituted the independent measures for the block were chosen based on accepted criteria (Jacobs & present study. Age-related changes in dendritic systems Scheibel, 1993): (a) The soma-apical across each individual were also noted. orientation is perpendicular to the pial surface. (b) The Dendritic system complexity was determined by soma is located centrally within the 120 gm section several dependent measures: total dendritic length depth. (c) The apical shaft is at least 100 gm in length. (TDL), mean dendritic length (MDL), dendritic segment (d) At least three primary basilar dendritic shafts are count (DSC), dendritic spine number (DSN), and present, each with at least two secondary branches and dendritic spine density (DSD). Dendritic systems were their consequent branch systems. (e) Neurons show no further broken down by proximal (lst-3rd) and distal obvious evidence of incomplete impregnation.- (0 Cells (4th order and higher) segments and by segment order to are relatively unobscured by adjacent neuronal determine where on the dendritic tree age-related changes structures. (g) Higher order branches should have were most pronounced. Dependent values are provided natural terminations, either characterized by naturally as an average per cell per group (±SEM). tapered ends, growth conebearing tips, or by terminal Because of the small sample size, only descriptive clusters of dendritic spines. The last criterion was not statistics were used for the present study. Pearson absolute because some had been sectioned product-moment correlations between age and the during processing. Cells with incomplete endings were dependent variables were also performed. included in the present study because exclusion of these neurons would have biased the sample towards smaller Results neurons (Buell & Coleman, 1981). Ten cells were randomly selected from each block As expected, all dependent measures, except DSC, and traced if they met the above criteria. First, the soma decreased with age, with the greatest decline observed in was traced at its widest extent. Then, at least 100 .tm spine measurements (see Table 2). Area 10 exhibited of the apical shaft was traced for orientation purposes. greater dendritic systems than area 18 (see Figure 1). In Finally, the basilar dendrites, including dendritic comparing the younger and older age groups, area 10 were traced. Unlike Horner (1993), no distinction was exhibited a greater decrease in dendritic measures (MDL: made between thin, stubby, or mushroom-shaped 17.7%, DSN: 51.8%, DSD: 53.3%) than did area 18 spines because of time-constraints. (MDL: 11.7%, DSN: 49.1%, DSD: 48.1%). Dendritic values tended to be 12.8% higher for males than for Intra- and Inter-rater Reliability females.

Cells were traced by six different raters: KC = 60 Summary of Neuronal Population cells; Serapio Baca = 10 cells; Bob Jacobs = 20 cells; Lori Larsen = 50 cells; Rebecca Swanson = 20 cells; Measurements from Nissl stains revealed that Marcy Wainwright = 40 cells. Intrarater reliability was sampled cells originated in cortical layers II and EEL determined by having each rater trace the same dendritic There was little difference in cortical and laminar system (including somata and dendritic spines) ten thickness between age groups (see Table 3), or between times. The average coefficient of variation across all cortical areas. Area 10 was characterized by greater raters for TDL, soma size, and DSN was 0.053, laminar and cortical thickness than area 18. Soma size indicating little variation in tracings. The first five tended to decrease slightly from the younger (M = tracings were compared with the second five tracings in 230.8±9.92 gm) to the older (M = 200.9±10.10 gm) a split plot design (a = 0.5). There was no significant group (r[200] = -0.21; p < 0.003). A decline in soma difference within raters for any of these measures. All size was associated with a decrease in MDL (4200] = raters were normed before quantification to maximize 0.45; p < 0.01), TDL (r[200] = 0.22; p < .002), DSC interrater reliability. In tracings of ten different dendritic

12 MODERN PSYCHOLOGICAL STUDIES AGE-RELATED DENDRITIC CHANGES IN HUMAN CORTEX

Table 2 Area 10 Area 18 A Breakdown of Dependent Measures across Individuals Younger group Age TDL MDL DSC DSN DSD

100 23 4000.66 61.94 64.55 1724.40 0.38 32 3313.76 71.18 48.17 1311.53 0.29

34 3466.45 63.02 54.35 985.35 0.27 48 3518.27 64.14 55.05 896.90 0.19 Older group 51 3008.72 54.31 55.25 586.00 0.14 100 64 3427.49 57.70 59.75 549.90 0.12 Figure 1. Representative tracing of neurons from 66 3301.50 53.40 61.65 718.35 0.16 prefrontal cortex (area 10) and occipital cortex (area 18) of 69 3648.73 57.24 63.65 543.00 0.12 a 23-year-old male and an 81-year-old male. Area 10 neurons possessed greater dendritic neuropil than area 18 81 3667.01 56.68 64.60 678.70 0.15 neurons. Dendritic complexity, especially with regard to spine number, is greater in the 23-year-old than in the 81- year-old. Grand 3482.60 61.08 57.52 930.57 0.21 mean

23-48 3554.58 66.29 54.06 1245.94 0.28 order segments but remained stable from fourth to sixth mean order, averaging 78.3 gm (see Figure 3). There was a 51-81 3410.61 55.87 60.98 615.19 0.14 general decrease in MDL with age (r[200] = -0.37; p < mean 0.01), but not in TDL.

Note. Dependent variable values represent mean measurements Dendritic Segment Count. Unlike dendritic per neuron for the 20 cells quantified for each individual. length measures, the number of segments tended to Abbreviations: TDL = total dendritic length; MDL = mean dendritic length; DSC = dendritic segment count; DSN = dendritic increase with age (see Figure 4). DSC was 11% lower spine number; DSD = dendritic spine density. in the younger group (54.1±1.41) than in the older group (61.0±1.48) and 12% higher in area 10 (r[200] = 0.32; p < 0.01), and DSN (r[200] = 0.31; p < (60.9±1.55) than in area 18 (54.2±1.35). For the total 0.01). population, DSC was 5.1% higher in the distal part of the dendritic tree than in the proximal part. Dendritic Measures There were 17.9% more distal segments in the younger group over the older group, but only 5.0% Dendritic Length Measures. There was an more proximal segments. An order by order analysis inverse relationship between age and measures of dendritic length. TDL was 4.2% higher in the younger Table 3 group (3,555±108.0 pm) than in the older group (3,411±102.8 gm) and 15% higher in area 10 Depth of Cortical Layers and Sampled Cells (pm) (3,725±115.0 gm) than in area 18 (3,240±89.0 pm; Younger Older Area 10 Area 18 see Figure 2). MDL was 19% greater in the younger group group group (66.3±1.29 pm) than in the older group (55.9±0.91 pm) and 2.5% higher in area 10 (61.8±1.34 Layer 1/11 junction 294 282 305 271 pm) than in area 18 (60.3±1.04 gm). Overall, distal Layer II/III junction 573 571 596 548 segments contributed 61% more to TDL and- 104% Sampled cells 850 842 844 848 more to MDL than did proximal segments. TDL increased from first to fourth order and tended to Layer III/IV junction 1407 1385 1429 1363 Gray/white matter 3092 3353 3387 3064 decrease thereafter.-MDL increased from first to higher junction

MODERN PSYCHOLOGICAL STUDIES 13 Kelly A. Courns and Bob Jacobs

(a) D Area 10 100 E • Area 18

80 4000 n)

I 3500 - in

l ( a) 60 l 3000 - Ce

h/ 2500 t 45 40 — 2000 - cu Leng

ic 1500 - it ro 20 dr MDL n 1000 - 2 De 500 - 0

ta l 1

To 0 2 3 4 5 6 7 Young Old Dendritic Segment Order

Figure 3. Mean dendritic length (MDL in 1.1m) broken O Proximal (b) down by order. MDL tended to increase with each • Distal subsequent segment order. Segments 8 and 9 were not

graphed because there were only one or two segments each. 800 rn) Bars represent SEM. u 700 — in ( ll 600 — indicated an increase in DSC up to the fourth order and /Ce h t 500— then a decrease in subsequent segments. There was a

ng r 400 — significant increase in DSC with age (r[2001 = 0.23; p Le

ic 300 -. 7,1, v. < 0.001). it

dr 200 — Agog, Dendritic Spine Measures. As with length 100 —

ta l Den measures, spine measures decreased as a function of 0 To age. DSN was 50.6% less in the older group Young Old (615±25.3) than in the younger group (1,246±52.2) and 31% higher in area 10 (1,056±58.5) than in area 18 (805±40.4; see Figure 5). DSD was 50% less in the older group (0.14±0) than in the younger group

)

m (0.28±0.01) and 10% greater in area 10 (0.22±0.01) g

in than in area 18 (0.20±0.01). An order by order analysis ( ll revealed greater DSN and DSD for all orders in the Ce

h/ younger group than in the older group (see Figure 6). t DSN and DSD were 89% and 155% greater in distal

Leng segments than in proximal segments, respectively. ic it Figure 7 illustrates a decline in DSN with age, with dr n area 10 consisting of more dendritic spines than area 18

l De across all but one individual. There was a significant ta

To decrease in DSN (r[200] = -0.63; p < 0.01) and DSD (r[200] = -0.77; p < 0.01) with age. 1 1 1 1 1 1 1 1 23 32 34 48 51 64 66 69 81 Discussion Subjects by Age

Figure 2. Total dendritic length (TDL in µm) for the The results of the present study revealed a general younger versus the older group broken down by (a) area decline in dendritic systems with age. This decrease in and (b) distal/proximal segments. TDL is also displayed dendritic systems was slightly more pronounced in area for individuals broken down by area (c). Note the higher 10 over area 18. Interpretation of these findings, dendritic values in area 10 and in distal segments. Bars however, is constrained by methodological limitations. represent SEM.

14 MODERN PSYCHOLOGICAL STUDIES AGE-RELATED DENDRITIC CHANGES IN HUMAN CORTEX

(a) ▪ Area 10 Methodological Limitations ▪ Area 18 70 Limited availability of tissue with short autolysis time and detailed biographic -6 60 — information, especially from young individuals, places 3 50— crucial restraints on quantitative dendritic studies. Such ca, limitations make it difficult to obtain multiple subjects E 40- CY) of the same age, which is required for statistical c' 30 analyses to be meaningful (Flood, 1993). Without such large sample sizes, it is difficult to rule out factors 20 (e.g., educational level and personal history) that also 10., influence dendritic systems (Jacobs, Batal, et al., 1993). 0 Young Old Histological limitations. A critical issue in quantitative studies is the histological and O Proximal

▪ Distal Area 10

Area 18 t n Cou t n 1200 me Seg ic

it 800- dr

Den 400—

Young Old

Young Old

Proximal

t (b) n I Distal 60 300 Cou t en

m 250 —

40 — ber m Seg 200 — ic it Nu

dr —0— Area 10 n 20 ine 150. De —•— Area 18 Sp ic

it 100 0 dr I I I 1 I I I 23 32 34 I 48 51 64 66 69 81 Den Subjects by Age 0 Young Old Figure 4. Dendritic segment count (DSC in gm) for the younger versus the older group broken down by (a) area and Figure 5. Dendritic spine number (DSN) for the younger (b) distal/proximal segments. DSC is also displayed for versus the older group broken down by (a) area and (b) individuals broken down by area (c). The older group has distal/proximal segments. DSN is higher in the younger more dendritic segments than the younger group, and area group and in area 10. Distal segments consist of more 10 exhibits more dendritic segments than area 18. Bars dendritic spines than proximal segments. Bars represent represent SEM. SEM.

MODERN PSYCHOLOGICAL STUDIES 15 Kelly A. Courns and Bob Jacobs

(a) 2000 500 .ta 1600 — --B-- Young E z 400— --•-- Old a) 1200 — 0. E ci) zo, 300— c co 200-- • 400 — a, 100— 0 23 32 34 48 51 64 66 69 81

I T I I Subjects by Age 2 3 4 5 6 7 Dendritic Segment Order Figure 7. Dendritic spine number (DSN) per individual broken down by area. Bars represent SEM. (b) 0.4 Several studies have reported variable results due to different Golgi techniques, especially between the 0.35 Golgi-Cox and the rapid Golgi method. Rapid Golgi 0.3- stains tend to be more sensitive to post-mortem a,c fixation delay than the Golgi-Cox method (Buell, • 0.25 — 1982). This delay tends to have adverse effects, a,c rna- 0.2 — particularly on dendritic spines (deRuiter, 1983). To avoid degenerative changes in brain tissue, deRuiter .c 0.15 — suggested that fixation delay should not exceed four —fi-- Young • 0.1 — hours. Although the average autolysis time in the 0.05 — Old present study was 10.4 hours, there was no correlation between autolysis time and TDL, which is in 0 1 I agreement with previous findings (Jacobs & Scheibel, 2 3 4 5 6 7 1993), nor did dendritic spines appear to decrease as a Dendritic Segment Order result of increased autolysis time. Nevertheless, the Figure 6. An order by order analysis for the younger and number of dendritic segments did appear to decrease older group: (a) Dendritic spine number (DSN), (b) with increased autolysis time (r[2001 = -0.23; p < Dendritic spine density (DSD). DSN and DSD are 0.01). consistently greater in the younger group than in the older group. Segments 8 and 9 were not graphed because there In spite of the potential problems associated with were only one or two segments each. Bars represent SEM. the rapid Golgi technique, it remains particularly useful for formalin-fixed tissue (Buell, 1982). Furthermore, within our laboratory, the rapid Golgi method has quantification techniques. It has been suggested that provided more complete impregnations and more detail Golgi-stained neurons may not represent a random of dendritic structures (Jacobs & Scheibel, 1993) than population (Cupp & Uemura, 1980). However, in the any other silver technique. Because the rapid Golgi present study, pyramidal cells of varying degrees of technique has been consistently applied in our previous complexity were observed. Other Golgi investigations studies, it is essential to continue with the same have observed neurons with rich dendritic arborizations methodology if the present results are to be compared as well as shrunken, regressing neurons within the with past findings. same cortical area (Buell & Coleman, 1981; Coleman & Flood, 1986), indicating the presence of a random Light microscopy. One restriction of light sample. In summary, there is no indication to date that microscopy is the inability to see spines above and a selective population of neurons is preferentially below dendrites. Because thicker branches tend to have stained with the Golgi technique (Flood, 1993). more dendritic spines than smaller branches, spines on

16 MODERN PSYCHOLOGICAL STUDIES AGE-RELATED DENDRITIC CHANGES IN HUMAN CORTEX thicker branches tend to be underestimated (Horner, pyramidal cells. The greatest decline was found in 1993; Horner & Arbuthnott, 1991; but see Belichenko dendritic spines, which was expected because previous, & Dahlstrom, 1995). Although correction equations qualitative work (Scheibel, 1992) had suggested such a (Feldman & Peters, 1979) and three-dimensional decline with age. Continued reorganization of dendritic reconstructions of dendrites (Belichenko & Dahlstrom) systems throughout the aging process might be attempt to compensate for this underestimation, such responsible for the loss of dendritic spines because compensation techniques would not be practical in the spines are resorbed if they lose their synaptic current study because of the extensive nature of the connections (Bondareff & Geinisman, 1976; Feldman, data. Spine measures in the present project thus 1976; Vaughan, 1977). Resorption of proximal probably represent an underestimation. Nonetheless, the dendritic segments (Simonds & Scheibel, 1989) appears present cell selection criteria (Jacobs & Scheibel, 1993) to be one aspect of this on-going reorganizing process. and sampling procedure ensured that the degree of Such a resorption of dendritic segments throughout underestimation was constant across all samples. one's life span might account for the observed decreases An additional restriction of light microscopy is in dendritic length observed in the present study. sectioning, which often truncates endings. Although Unlike dendritic length and spine measurements, 26% of dendritic segments were sectioned in the present the number of dendritic segments increased with age, study, cell selection criteria (Jacobs & Scheibel, 1993) which is consistent with previous studies. Jacobs and minimized the number of incomplete endings. To avoid Scheibel (1993), for example, found an increase in cut segments entirely, sections would need to approach DSC from the younger (< 50 years) to the older group a thickness of approximately 1,000 p.m, making light (r[20] = 0.23) with the same magnitude of increase as microscopy impossible. Given the limitations of light the present results. Such an increase in dendritic microscopy, one solution would be to eliminate segments would tend to increase TDL values. This neurons with cut segments. However, exclusion of would in turn reduce the observable loss in TDL such neurons would result in an unrepresentative (Jacobs & Scheibel) resulting from decreases in neuronal population (Buell & Coleman, 1981). A segment length. The addition of segments appeared to second solution would be to trace across sections. occur primarily in the distal part of the dendritic arbor Because of the extensive nature of the present data and (recall Figure 4b). This is to be expected because because the Golgi method stains several neurons peripheral segments are believed to represent the most simultaneously (unlike injection methods, cf. plastic portion of the dendritic tree (Carughi, Carpenter, Belichenko & Dahlstrom, 1995), serial tracing would & Diamond, 1989; Scheibel, 1988) and appear to be be impractical. Despite problems associated with preferentially involved in higher-level cognitive dendritic truncation, incomplete endings were consistent functions (Jacobs, Batal, et al., 1993). An addition of across age group and area in the present investigation. segments to the dendritic tree with age might therefore provide a compensatory mechanism for dendritic Neuronal Shrinkage resorption (Coleman & Flood, 1986). The overall regression in dendritic systems can be Soma size decreased from the younger to the older seen as a normal part of the maturation process group. The present findings are consistent with (Scheibel, 1992). Some synaptic connections may be previous studies (Anderson et al., 1983; Terry et al., lost, but other connections are enhanced. In essence, 1987; Vaughan, 1977) documenting neuronal there is a refinement of dendritic systems with shrinkage. This decline in soma size is believed to increasing age. This refinement is perhaps one reason result in an increase in the population of smaller why older individuals can still perform high-level neurons (Terry et al.). Although such a shift in neuron mental tasks, but at a slower rate. Giaquinto (1988) type was not examined in the current study, there did performed experiments on younger and older individuals appear to be a direct relationship between soma size and and found that the number of errors made was the amount of dendritic systems, which would result in approximately the same for both groups, but that there an apparent increase in the number of smaller neurons. was a decline in reaction time with age. With increasing age, there appears to be a general slowing of Dendritic Regression behavior (Giaquinto) and a decrease in the ability to recognize faces (Grady et al., 1995). Such changes may The current findings document age-related, basilar result from dendritic degeneration such as those dendritic regression in area 10 and area 18 supragranular observed in the present study.

MODERN PSYCHOLOGICAL STUDIES 17 Kelly A. Courns and Bob Jacobs

Area 18 versus Area 10 Individual Variations

All dependent measures were higher in area 10 than When the personal history of each individual is in area 18, which is consistent with preliminary results unknown, one cannot determine the extent to which (Larsen et al., 1994). This finding may reflect other factors (e.g., education and environment) differences in the computational and processing influence the results. In the present study, the 81-year- demands placed on dendritic systems in area 10 (Baca et old male, who was a physician . by occupation, al., 1995). According to Benson's (1994) functional exhibited greater TDL than all but the youngest subject hierarchical schema, supramodal association cortex and greater spine number than the 69-year-old male, (prefrontal area 10), the phylogenetically newest part of whose occupation was unknown. Perhaps, the 81-year- the cortex, is perhaps the most functionally complex old's educational achievements were responsible for his area of the brain. It has executive control over other relatively higher TDL and spine measures. Such a cortical functions, thus enabling it to select or inhibit possibility was suggested by Jacobs, Schall, and cognitive activity. Moreover, area 10 appears to be Scheibel (1993), who found a relationship between responsible for high-level cognitive functions (e.g., educational level and dendritic values. They believed discrimination, recognition, recall, attention, visual and that a higher educational level coupled with continued, tactile learning; Roland, 1993). Even though area 10 intellectual activity could contribute to a proliferation does monitor some sensory information, most of this in dendritic systems. Over 30 years of nonhuman information (e.g., visual) is analyzed by -unimodal animal research investigating the effects of differential association cortex (occipital area 18), which specifically environments on the brain have yielded similar results analyzes visual patterns (i.e., orientation, color, spatial (Diamond, Krech, & Rosenzweig, 1964). For example, frequencies, motion; Roland). In summary, area 10 is plasticity in response to a challenging environment has responsible for high-level cognition, and area 18 is been shown in 904-day-old male rat brains (Diamond, responsible for processing visual impulses (Benson). Johnson, Protti, Ott, & Kajisa, 1985), which is Because distal segments appear to be preferentially equivalent to an 80-year-old human brain. These rats involved in high-level cognitive functions, it seems were placed in an enriched environment at 766 days of reasonable to expect that area 10 would have more age and, after 138 days, showed an increase in cortical distal segments than area 18, which was the case in the thickness, presumably as a result of increased dendritic present study. Jacobs, Batal, et al. (1993) found that aborizations. Such findings indicate continued plasticity Wernicke's and Broca's areas, which are involved in throughout the life span of the organism and suggest high-level functions, also exhibited a distal over that dendritic system degeneration may be diminished in proximal advantage in dendritic length and suggested older rats, and possibly humans, in the face of a that distal segments might be involved in higher order challenging environment. cortical activities. Such a distal advantage was not observed in the motor strip areas, which represent Conclusion primary cortex in Benson's (1994) schema. Similar findings have been observed in prefrontal cortices of The current study quantitatively documented life rhesus monkeys (Cupp & Uemura, 1980). span changes in dendritic systems and is the first to A greater decline in dendritic systems in area 10 document life span decreases in spine number in human may be due to decreased utilization of the prefrontal brain tissue. The results of the present investigation cortex with increasing age or due to preferential revealed: (a) A decline in total dendritic length; mean sensitivity of these higher cortical areas to the effects of dendritic length, dendritic spine number, and dendritic aging. Involvement of area 18 in basic visual spine density from the younger to the older group, with processing may make it less susceptible to age-related the greatest decline in spine number. (b) Greater changes because the visual system may be more dendritic systems in area 10 than in area 18. (c) A consistently utilized throughout the life span than slightly more pronounced decrease in dendritic systems higher order areas. Additional support for the current with age in area 10 over area 18. Importantly, the findings is provided by studies of local cerebral blood present project reconfirms previous qualitative flow (Tachibana et al., 1984), which found a greater observations (Scheibel, 1992) by providing quantitative decline in cerebral blood flow in prefrontal cortical areas data. Despite the loss in dendritic systems with age, than in visual areas with age. human (Jacobs, Schall, & Scheibel, 1993) and nonhuman (Diamond et al., 1985) animal studies have

18 MODERN PSYCHOLOGICAL STUDIES AGE-RELATED DENDRITIC CHANGES IN HUMAN CORTEX indicated continued plasticity throughout the aging histology of the rat cerebral cortex. The Journal of process and suggest that challenge can at least partially Comparative , 123, 111-120. Feldman, M. L. (1976). Aging changes in the attenuate the effects of aging. The overall findings morphology of cortical dendrites. In R. D. Terry & S. stress the importance of maintaining a mentally active Gershon (Eds.), Neurobiology of aging (pp. 211- life into old age. 227). New York: Raven Press. Feldman, M. L., & Peters, A. (1979). A technique for References estimating total spine numbers on Golgi-impregnated dendrites. The Journal of Comparative Neurology, Anderson, J. M., Hubbard, B. M., Coghill, G. R., & 118, 527-542. Slidders, W. (1983). The effect of advanced old age on Flood, D. G. (1993). Critical issues in the analysis of the neurone content of the cerebral cortex. Journal of dendritic extent in aging humans, primates, and the Neurological Sciences, 58, 233-244. rodents. Neurobiology of Aging, 14, 649-654. Baca, S., Larsen, L., Fisher, B., Kernan, R., Schall, M., & Flood, D. G., & Coleman, P. D. (1988). Neuron numbers Jacobs, B. (1995). Dendritic and spine analyses and sizes in aging brain: Comparisons of human, across hierarchically arranged areas of human monkey, and rodent data. Neurobiology of Aging, 9, : A quantitative Golgi study. Society for 453-463. Abstracts, 182-9. Giaquinto, S. (1988). Aging and the . New Belichenko, P. V., & Dahlstrom, A. (1995). Studies on the York: John Wiley & Sons. 3-dimensional architecture of dendritic spines and Grady, C. L., McIntosh, A. R., Horwitz, B., Maisog, J. varicosities in human cortex by confocal laser M., Ungerleider, L. G., Mentis, M. J., Pietrini, P., scanning microscopy and Lucifer Yellow Schapiro, M. B., & Haxby, J. V. (1995). Age-related microinjections. Journal of Neuroscience Methods, reductions in human recognition memory due to 57, 55-61. impaired encoding. Science, 269, 218-220. Benson, F. (1994). The neurology of thinking. New York: Gridley, M. (1960). Manual of histologic and special Oxford. staining techniques (2nd ed.). New York: McGraw- Bondareff, W., & Geinisman, Y. (1976). Loss of synapses Hill. in the dentate gyms of the senescent rat. American Homer, C. H. (1993). Plasticity of the dendritic spine. Journal of , 145, 129-136. Progress in Neurobiology, 4, 281-321. Buell, S. J. (1982). Golgi-Cox and rapid Golgi methods as Homer, C. H., & Arbuthnott, E. (1991). Methods of applied to autopsied human brain tissue: Widely estimation of spine density-Are spines evenly disparate results. Journal of and distributed throughout the dendritic field? Journal of Experimental Neurology, 41, 500-507. Anatomy, 177, 179-184. Buell, S. J., & Coleman, P. D. (1981). Quantitative Huttenlocher, P. R. (1979). Synaptic density in human evidence for selective dendritic growth in normal frontal cortex-Developmental changes and effects of human aging but not in senile dementia. Brain aging. Brain Research, 163, 195-205. Research, 214, 23-41. Jacobs, B., Batal, H. A., Lynch, B., Ojemann, G., Carughi, A., Carpenter, K. J., & Diamond, M. C. (1989). Ojemann, L. M., & Scheibel, A. B. (1993). Effect of environmental enrichment during nutritional Quantitative dendritic and spine analyses of speech rehabilitation on body growth, blood parameters and cortices: A case study. Brain and Language, 44, 239- cerebral cortical development of rats. Journal of 253. Nutrition, 119, 2005-2016. Jacobs, B., Chugani, H. T., Allada, V., Chen, S., Phelps, Chugani, H. T., Phelps, M. E., & Mazziotta, J. C. (1987). M. E., Pollack, D. B., & Raleigh, M. J. (1995). Positron emission tomography study of human brain Developmental changes in brain metabolism in functional development. Annals of Neurology, 22, sedated rhesus macaques and vervet monkeys revealed 487-497. by positron emission tomography. Cerebral Cortex, Coleman, P. D., & Flood, D. G. (1986). Dendritic 3, 222-233. proliferation in the aging brain as a compensatory Jacobs, B., Schall, M., & Scheibel, A. B. (1993). A repair mechanism. Progress in Brain Research, 70, quantitative dendritic analysis of Wernicke's area in 227-237. humans. II. Gender, hemispheric, and environmental Cupp, C. J., & Uemura, E. (1980). Age-related changes in factors. The Journal of Comparative Neurology, 327, prefrontal cortex of Macaca mulatto: Quantitative 97-111. analysis of dendritic branching patterns. Jacobs, B., & Scheibel, A. B. (1993). A quantitative Experimental Neurology, 69, 143-163. dendritic analysis of Wernicke's area in humans. I. deRuiter, J. P. (1983). The influence of post-mortem Life span changes. The Journal of Comparative fixation delay on the reliability of the Golgi silver Neurology, 327, 83-96. impregnation. Brain Research, 266, 143-147. Kety, S. S. (1956). Human cerebral blood flow and oxygen Diamond, M. C., Johnson, R. E., Protti, A. M., Ott, C., & consumption as related to aging. Journal of Chronic Kajisa, L. (1985). Plasticity in the 904-day-old male Disorders, 3, 478-486. cerebral cortex. Experimental Neurology, 87, 309- Larsen, L., Swanson, R. L., Wainwright, M. L., & Jacobs, 317. B. (1994). Quantitative dendritic and spine analyses Diamond, M. C., Krech, D., & Rosenzweig, M. R. (1964). of human prefrontal and occipital cortices. Society for The effects of an enriched environment on the Neuroscience Abstracts, 584-13.

MODERN PSYCHOLOGICAL STUDIES 19 Kelly A. Courns and Bob Jacobs

Leenders, K. L., Perani, D., Lammertsma, A. A., Heather, Scheibel, A. B. (1992). Structural changes in the aging J. D., Buckingham, P., Healy, M. J. R., Gibbs, J. M., brain. Handbook of mental health and aging (2nd ed., Wise, R. J. S., Hatazawa, J., Herold, S., Beaney, R. pp. 147-173). New York: Academic Press. P., Brooks, D. J., Spinks, T., Rhodes, C., Scheibel, M. E., & Scheibel, A. B. (1978). The methods of Frackowiak, R. S. J., & Jones, T. (1990). Cerebral Golgi. In R. T. Robertson (Ed.), Neuroanatomical blood flow, blood volume, and oxygen utilization. research techniques (pp. 89-114). New York: Brain, 113, 27-47. Academic Press. Masliah, E., Terry, R. D., Alford, M., DeTeresa, R., & Simonds, S. J. & Scheibel, A. B. (1989). The postnatal Hansen, L. A. (1991). Cortical and subcortical development of the motor speech area: A preliminary patterns of synaptophysin-like immunoreactivity in study. Brain Language, 37, 42-58. Alzheimer's disease. American Journal of Pathology, Tachibana, H., Meyer, J. S., Kitagawa, Y., Rogers, R. L., 138, 235-246. Okayasa, H., & Mortel, K. F. (1984). Effects of aging Melamed, E., Lavy, S., Bentin, S., Cooper, G., & Rinot, on cerebral blood flow in dementia. American Y. (1980). Reduction in regional cerebral blood flow Geriatrics Society Journal, 32, 114-120. during normal aging in man. Stroke, 11, 31-35. Terry, R. D., DeTeresa, R., & Hansen, L. A. (1987). Roland, P. E. (1993). Brain activation. New York: Wiley- Neocortical cell counts in normal human adult aging. Liss. Annals of Neurology, 21, 530-539. Saito, S., Kobayashi, S., Ohashi, Y., Igarashi, M., Vaughan, D. W. (1977). Age-related deterioration of Komiya, Y., & Ando, S. (1994). Decreased synaptic basal dendrites in rat auditory cortex. density in aged brains and its prevention by rearing The Journal of Comparative Neurology, 171, 501- under enriched environment as revealed b y 516. synaptophysin contents. Journal of Neuroscience Yoshii, F., Barker, W. W., Chang, J. Y., Loewenstein, D., • Research, 39, 57-62. Apicella, A., Smith, D., Boothe, T., Ginsbery, M. D., Scheibel, A. B. (1988). Dendritic correlates of human Pascal, S., & Duara, R. (1988). Sensitivity of cerebral cortical function. Archives Italiennes de Biologie, glucose metabolism to age, gender, brain volume, 126, 347-357. brain atrophy, and cerebrovascular risk factors. Journal of Cerebral Blood Flow and Metabolism, 8, 654-661.

20 MODERN PSYCHOLOGICAL STUDIES