Abstract

Brain structure changes in size with normal aging, but the rate at which different structures change is controversial. We used magnetic resonance imaging (MRI) performed twice, 4 years apart, to compare rates of age-related size change of the corpus callosum, which has been inconsistently observed to thin with age, with change in the lateral ventricles, which are well established to enlarge. Subjects were 215 community dwelling, elderly men (70–82 years old at initial MRI), who were participants in a longitudinal study of cardiovascular risk factors. Percent change in size was significant for both the callosal and ventricular measures, but annual rate of ventricular expansion (2.9%) was significantly greater than annual rate of callosal thinning (–0.9%). Callosal regions showed statistically equivalent rates of shrinkage; ventricular dilatation was symmetrical. Neither callosal and ventricular rates of change correlated with each other (r = 0.01), nor did genu and splenium rates of change correlate with each other (r = 0.05). Tests of speeded processing were administered contemporaneously with both MRIs to examine functional ramifications of observed brain changes. Decline in the Mini-Mental State Examination was related to thinning of the splenium, and decline in Stroop test word reading was selectively related to thinning of the callosal body. These longitudinal data support the contentions that differential rates of change occur in different brain regions in normal aging, age-related callosal thinning contributes to functional declines, and rate of change in one region can be independent of rate of change in another region, even within a brain structure.

Introduction

The corpus callosum is the major neural system of white matter tracts in the brain and functions as the principal conduit of interhemispheric communication. Post-mortem study of the corpus callosum indicates significant age-related structural in-crease during development to adulthood (Yakovlev and Lecours, 1967) and deterioration, including breakdown of myelin and microtubules and deletion of small diameter fibers, with senescence (Pandya and Seltzer, 1986; Aboitiz et al., 1996). Although in vivo study using quantitative magnetic resonance imaging (MRI) provides convincing longitudinal evidence for substantial increase in callosal size throughout childhood (age 5–18 years) (Giedd et al., 1999), the fate of callosal size in aging adults is controversial. Many (Cowell et al., 1992; Biegon et al., 1994; Johnson et al., 1994; Pozzilli et al., 1994; Pfefferbaum et al., 1996; Hampel et al., 1998; Sullivan et al., 2001b; Thompson et al., 2002) but not all (Doraiswamy et al., 1991; Weis et al., 1993; Janowsky et al., 1996) published in vivo studies have concluded that if age-related thinning occurs it is modest, especially in relation to concurrent changes in other brain regions [e.g. (Pfefferbaum et al., 2000)]. Studies likely to observe age effects have focused on elderly adults [e.g. (Janowsky et al., 1996)]. A meta-analysis examined the influence of age as well as other factors on corpus callosum morphology (Driesen and Raz, 1995). An assessment of 21 studies yielded only modest and non-significant estimates of correlations between total midsagittal callosal area and age. The authors concluded that although age-related decreases in callosal area may occur, longitudinal studies that use the same analytical techniques across sessions are essential for a valid test of this possibility.

A number of factors contribute to the controversy regarding aging of callosal macrostructure. Firstly, the corpus callosum is located adjacent and superior to the lateral ventricles, which expand significantly in normal aging [e.g. (Pfefferbaum et al., 1994; Gur et al., 1999; Resnick et al., 2000)]. Associated with ventricular expansion is callosal arching and thinning (Peterson et al., 2001), even when its overall area remains relatively unchanged (Pfefferbaum et al., 2000). Thus, callosal dys-morphology resulting from passive stretching by the ventricles must be considered when measuring age-related changes. Further, despite the distinct appearance of the corpus callosum on midsagittal MR images, its regional divisions are not defined by internal or external landmarks [e.g. (Peterson et al., 2001)]. Consequently, a variety of unique geometrically based, and not necessarily comparable, schema have been devised to demarcate callosal regions. Finally, virtually all conclusions about callosal aging have been based on cross-sectional study, span-ning either the normal adult range [e.g. (Pfefferbaum et al., 1996; Peterson et al., 2001; Sullivan et al., 2001b)] or restricted older samples [e.g. 50–90 years (Laissy et al., 1993; Janowsky et al., 1996; Hampel et al., 1998)]. The dearth of longitudinal study in adults undoubtedly contributes to the inconsistencies in characterization of aging of callosal structure as well as its function.

The purpose of this study was to address, for the first time, callosal aging in a large sample of community dwelling elderly adults who were studied twice with the same MR imaging protocol over 4 years. The lateral ventricles were also measured for two reasons. First, change in ventricular size served as a control region, which would be assumed to change significantly given that a number of longitudinal MRI studies have already established substantial ventricular dilatation in normal aging. Accordingly, we tested the hypothesis that little if any shrinkage would occur in callosal area, especially in comparison with marked ventricular expansion expected over the same interval in the same subjects. Secondly, we tested the hypothesis that expansion of the lateral ventricles, which are adjacent to the corpus callosum, would contribute substantially to callosal change in contour without altering its overall size. Finally, we examined the functional ramifications of potential callosal thinning by correlating percent change in callosal size with percent change in two cognitive tests with known sensitivity to age-related performance declines [for a review see (Salthouse, 2000)]: Trail Making, which assesses visual search, working memory, and psychomotor speed, and the Stroop Test, which assesses response inhibition and reading speed.

Materials and Methods

Subjects

Data for this study were collected as part of the National Heart, Lung, and Blood Institute (NHLBI) Twin Study, an ongoing longitudinal genetic study of cardiovascular disease risk factors and their effect on brain structure and function in old age [e.g. (Carmelli et al., 1998; Swan et al., 1999)]. All participants were World War II veterans, born in 1917–1927 and 42–56 years old when first examined in 1969–1972 (Feinleib et al., 1977). These men underwent brain scanning with MRI on two occasions: 1995–1997 and 1999–2000. Data used in this analysis were collected at three locations: Stanford University School of Medicine, Indiana University, and West Suburban Imaging Center in Massachusetts. All data from both scanning sessions were collected on 1.5 T GE Signa systems and were analyzed as a single data set at SRI International, using in-house software for image alignment and region of interest quantification written by A.P. The average (mean ± SD) interscan interval was 4.0 ± 0.4 years. Procedures for this project were approved at each participating institution by its review board for use of human subjects in research.

The group with repeated scans comprised 215 men who were 71.9 ± 2.7 years old at the first MRI and 75.9 ± 2.7 years old at the second MRI. The group included 34 monozygotic twin pairs, 37 dizygotic twin pairs, and 73 singletons; genetic analysis of the longitudinal twin-pair data will be performed in a separate study. On average, these men had 13.7 ± 2.9 years of education, and 201 of the 215 met criteria for right-handedness (Oldfield, 1971). Average Mini-Mental Status Examination (MMSE) score was 27.5 ± 2.2 at the first MRI and 27.2 ± 2.4 at the second MRI. Although 15 men at MRI 1 and 26 at MRI 2 scored in the impaired range (<25) on the MMSE, these men were included in analysis because all were independently living.

MRI Acquisition and Quantification

Corpus Callosum

The corpus callosum was measured on data from multislice, sagittal, single-echo, spin-echo series (TR = 300–500 ms, TE = 8–14 ms, thickness = 4 mm, skip = 1 mm, encompassing the midline) on both the native images and on an extracted midline slice from the native data after interpolation, alignment, and reslicing. The native image analysis was performed on the slice which contained the fullest view of the cross-section of the corpus callosum. For extraction of the aligned midline slice, the anterior com-missure (AC) was identified on the native sagittal slice on which it was best visualized. The sagittal slices were then stacked and interpolated to high (1 mm) isotropic resolution. The left/right midline was determined from a coronal reconstruction passing through the AC. The AC and posterior commissure (PC) were identified on a resliced midline sagittal image. Head tilt angle was defined on a high resolution coronal image and head rotation was defined on a high resolution axial image using an interactive rotating cursor. The resultant landmarks and angles were used to align the volume in a uniform space and orientation anchored on the AC. The midsagittal image was then extracted for semi-automated edge identification of the corpus callosum (Fig. 1). Inter-rater reliability (for E.V.S. and A.P.) was determined with intraclass correlations (ICC) and was high (n = 50, total area r = 0.99).

We tested the potential improvement of the extraction and realign-ment procedure over native image analysis by computing the effect size (the mean MRI 1 minus MRI 2 difference divided by the standard deviation of the difference scores) for both methods.

In addition to the total cross-sectional area of the corpus callosum, regional areas and shape-related variables (height and length) were quantified. Accordingly, the corpus callosum silhouette was rotated to a plane parallel to the inferior extremes of the rostrum anteriorly and splenium posteriorly. The midpoint along this plane between the anterior extreme of the genu and posterior extreme of the splenium was used as the center of a circle, and radii were projected at +30° and +150° angles from the plane, thus dividing the corpus callosum into genu + rostrum, body, and splenium. From this rotated image, the height and length of the callosal silhouette were also determined.

Lateral Ventricles

The lateral ventricles were measured on three slices taken from a multislice, coronal, single-echo proton density-weighted or dual-echo, fast spin-echo scan (T1 weighted or dual echo data manipulated to give T1-weighted contrast; thickness = 5 mm, skip = 0 mm, encompassing the entire brain). Similar to callosal quantification, ventricle quantification on the coronal images started with identification of the AC on sagittal reconstructed images. Head tilt angle was defined on a coronal image and head rotation was defined on a high resolution axial. Three coronal slices (at the level of the AC, AC + 10 mm, AC – 10 mm) perpendicular to the AC–PC plane were extracted for semi-automated edge identification, the sum of the resultant three areas was the volume estimate for the left and right ventricles separately (Fig. 2). Inter-rater ICC (for E.V.S. and A.P.) for ventricular measurements was high (n = 30, left r = 0.99, right r = 0.99).

In order to validate the adequacy of the three-slice estimate of lateral ventricular volume in representing total ventricular volume, we used a manual identification procedure applied to semi-automated segmentation images to measure the entire volume of the lateral ventricles across all the slices on which it appeared (~15 slices) on 30 subjects randomly selected from the two scanning sessions. The ICC between the sampled and fully volumed measurements was 0.94 (P = 0.0001).

Intracranial Volume (ICV)

An estimate of supratentorial intracranial volume (ICV) was derived from measurements on both the midsagittal image and the coronal image at the level of the AC. Interactive software allowed the operator to determine the left, right and superior parietal lobe and the inferior temporal lobe dural margins on the coronal slice; the anterior frontal lobe and posterior occipital lobe dural margins were determined from the sagittal image. These dimensions were then used to model intracranial volume as an elliptical solid; inter-rater intraclass correlation for 30 subjects was r = 0.83.

Neuropsychological Tests

Three pencil-and-paper based tests of psychomotor speed, which have been reliably shown to be sensitive to age-related performance declines, were included in this analysis. All tests were timed, required subjects to work quickly and accurately, and were administered at both test sessions. The Trail Making Test (Lezak, 1995) has two parts: for Trails A, subjects connected in ascending numerical order 25 numbered circles, placed irregularly on a page; for Trails B, subjects connected 25 circles, which contain either a letter or a number, in ascending alphabetical and numerical order. This test yielded two scores: time (s) to complete Trails A, a measure of visual search; and time to complete Trails B, a measure of set shifting and visual search. The Stroop Test (Stroop, 1935) comprised three conditions timed individually: color, requiring subjects to name the color of the ink in which a row of x's was printed; word, requiring subjects to read words printed in black ink that named colors; and color word, the interference condition, requiring subjects to say the color of ink in which a named color was printed but the ink and the word were not congruent; the score was the number of correct responses given in 45 s for each condition.

Statistical Analysis

Percent change per year for each MRI and cognitive measure was calculated as the difference between the scores at time 1 and 2, divided by the score at time 2 and the number of years between examinations. For each measure, the 95% confidence intervals (CI 95%) for the mean difference and the effect size were calculated. Within-subject analysis of variance (ANOVA) and paired t-tests examined brain regional and cognitive rates of change scores; for these analyses only, ventricular and callosal length change (expressed as enlargement or stretching) and Trail Making Test score differences (expressed in number of seconds to completion) were transformed so that more negative values were in the direction of greater change for the worse. Pearson correlation was used to test the relationships between rates of change of brain and cognitive measures; to adjust for multiple comparisons, family-wise Bonferroni correction for four tests (α = 0.05) were applied and required P ≤ 0.025. These correlations were based on rates of change values, not trans-formed, where negative callosal and Stroop values and positive ventricular and Trails values were in the direction of declining condition.

This analysis was based on all available MRI data, including those from twin pairs and singletons. Because a potential bias can exist in making statistical estimates from a sample including non-independent observa-tions (i.e. twin pairs), we used bootstrap methods (Efron and Tibshirani, 1986) (i.e. resampling the data with replacement) to derive empirical estimates of the standard errors of our statistics. To accomplish this, we created 1000 bootstrap data sets using the twins as genetically unrelated individuals.

Results

Rates of Change in Callosal and Ventricular Measures

The realignment procedure produced larger effect sizes for estimates of change in size of total corpus callosum (effect size native = 0.434, aligned = 0.588) and its three regions (genu native = 0.336, aligned = 0.470; body native = 0.209, aligned = 0.370; splenium native = 0.475, aligned = 0.492). Therefore, all sub-sequent analyses were based on the aligned measurements.

Table 1 presents 95% CIs for MRI1 – MRI2 mean differences and effect sizes of each brain measure. The effect sizes for the callosal measures were considerably smaller than those of the ventricular measures. Within the corpus callosum, regional variation was not significant.

Figure 3 displays average rates of change for each regional MRI measurement. Within-subject ANOVA comparing percent rates of change in the three sections of the corpus callosum and left and right ventricles yielded a significant repeated measures effect [F(4,856) = 70.934, P = 0.0001]. Follow-up tests indicated that although percent change in size was significant for all callosal and ventricular measures (for all paired t-tests, P = 0.0001), rate of annual ventricular expansion (left + right ventricles = 2.9%) was significantly greater than the annual rate of callosal shrinkage (total area = –0.9%) [t(214) = 11.93, P = 0.0001]. Regional analysis showed statistically equivalent rates of callosal thinning [genu = –0.9%; body = –0.7%; splenium = –1.0% per annum; F(2,428) = 1.133, n.s.] and of lateralized ventricular dilatation (left = 2.8%; right = 2.9%). Symmetrical ventricular change was detected longitudinally despite reliable lateralized differences detected cross-sectionally, in that the right ventricle was significantly larger than the left at MRI 1 [t(214) = 13.140, P = 0.0001] and at MRI 2 [t(214) = 15.251, P = 0.0001]. In addition, one-group t-tests indicated that height [t(214) = 6.137, P = 0.0001] but not length [t(214) = –0.398, n.s.] of the corpus callosum changed significantly over the retest interval; in particular, callosal height expanded. As was expected, intra-cranial volume showed a small and statistically non-significant annual change (–0.06%), reflecting primarily measurement error [cf. (Pfefferbaum et al., 1995; Shear et al., 1995)]. Bootstrap resampling tests performed for each measure yielded the same pattern of results described above. None of the regional brain changes was related to handedness.

Correlations between Changes in Brain Measures

All correlations and P-values are presented in Table 2. Within the corpus callosum, rate of change in body area correlated significantly with rate of change in genu and splenium areas. However, genu and splenium rates of change were not correlated (r = 0.05). Although change in the left and right ventricles were highly correlated with each other (r = 0.79), ventricular change showed no significant correlation with change in any callosal region (r = –0.06 to 0.10, n.s.).

Expansion in callosal height correlated with thinning of callosal body (r = 0.27) but not of the genu (r = 0.11) or splenium (r = 0.00), whereas increases in callosal length correlated with thinning of the genu (r = 0.25) and splenium (r = 0.37) but not of the body (r = 0.06). In addition, callosal height but not length was related to enlargement of left, right, and total ventricular size (r = 0.31 to 0.35). Finally, increases in callosal height correlated only modestly with callosal lengthening (r = 0.19).

Rates of Change in Neuropsychological Test Scores

One sample t-tests, assuming that the population mean = 0, yielded highly significant decline in performance on Trails A [t(208) = 6.065, P = 0.0001], Trails B [t(206) = 6.314, P = 0.0001], and the three Stroop Test measures [words t(199) = –3.722, P = 0.0003; colors t(197) = –7.628, P = 0.0001; color words t(193) = –3.628, P = 0.0004] (Fig. 3). One subject was excluded from the Stroop word condition analysis as an outlier; his difference score in the word reading condition was 6.74 SD below the group mean. The decline on the MMSE was not significant [t(212) = –1.36, P = 0.1752]. A within-subjects ANOVA, which tested differences among the two Trails and three Stroop measures, revealed, as expected, a significant repeated measures effect [F(4,954) = 10.270, P = 0.0001]. Follow-up tests revealed that the two Trails scores showed greater decline than the three Stroop scores. Although differences between measures within Trails were not significant [t(205) = –0.809, p = 0.4196], word reading performance on the Stroop test tended to show less decline than did color naming or color word reading [F(193,581) = 1.219, P = 0.0528]. None of the changes in test performance was significantly related to handedness; however, a trend was noted for decline in MMSE, which was greater in left than right-handed men [t(211) = 1.894, P = 0.0596].

Relationships between Changes in Brain Measures and Test Scores

Pairwise correlations between percent change in brain meas-ures and changes in cognitive measures yielded modestly significant relationships, meeting the corrected P-value for multiple comparisons, between splenium shrinkage decline in MMSE (r = 0.22, P < 0.002) and between thinning of the body and decline in Stroop word reading (r = 0.17, P < 0.02). Multiple regression analysis examined the independent contributions of change in the three callosal regions to change in word reading scores. Only the contribution from change in the body (β = 0.241, P = 0.041) to the word reading task was significant compared with that from the change in the genu (β = –0.038, P = 0.73) and splenium (β = 0.058, P = 0.587). When MMSE change was entered as an additional factor, the significant predictor of change in body remained essentially unchanged (MMSE β = –0.006, P = 0.948).

Discussion

This study focused on measuring change in size of the corpus callosum and lateral ventricles in elderly community dwelling men, who had received the same MRI sequences twice, with a 4-year interval. Longitudinal quantitative analysis revealed subtle but statistically significant thinning of the midsagittal area of the corpus callosum and of its three major regions, genu, body, and splenium. Unlike callosal development during childhood and adolescence, which undergoes relatively greater growth of posterior than anterior regions (Giedd et al., 1999), rate of callosal thinning in old age was similar in its three divisions. Although percent change in size was significant for both the callosal and ventricular measures, the annual rate of ventricular expansion (nearly 3%) was about three-fold greater than the annual rate of callosal thinning (~–1%).

The observation that the callosal and ventricular measures of size and shape exhibit different rates of change within the same individual over the same intervals comports with a theme in development of the human brain, namely, that its structures grow at different rates, a feature that contributes to variability in size of a given structure (Yakovlev and Lecours, 1967; Lange et al., 1997). Despite commensurate rates of change in genu and splenium size, the striking lack of correlation between these regional rates of change suggests that independent factors underlie these changes. Supporting this possibility are studies of age-related changes that report dissociable regional and structural brain tissue shrinkage, for example, in sulcal and ventricular expansion (Blatter et al., 1995; Jernigan et al., 1991; Pfefferbaum et al., 1990, 1994, 1998; Raz et al., 1997). Noted difference in heritability of selective brain structures is further indication of the relative independence of morphological change in certain brain structures (Carmelli et al., 2001). Quantitative analysis based on the full twin model report high heritability of regional and total callosal and ventricular size with less environmental influence on callosal size than on ventricular size (Pfefferbaum et al., 2000). This pattern of genetic and environ-mental influences held for measures of callosal microstructure, detected with MR diffusion tensor imaging, as well as its overall size and macrostructure (Pfefferbaum et al., 2001). Considerably less contribution from genetic sources was detected in different cortical regions (Carmelli et al., 2001), hippocampus and temporal horns (Sullivan et al., 2001a), even within the same individuals. Taken together, these results support the relative independence of the susceptibility of these structures to influ-ences from genes, the environment, and aging. Identification of genetic factors presenting significant contributions to variability in regional brain morphology may serve to explain hetero-scedasticity in certain measures in old age [e.g. (Raz et al., 1997; Pfefferbaum et al., 2000)].

The importance of considering the potential interactions and dissociations of contextual structural changes is evident from the morphological interaction of ventricular and callosal change in size and contour. Although genu and splenium area losses were not correlated, changes in both these callosal regions were related to increase in height but not length of the corpus callosum. By contrast, contemporaneous thinning of the body was related to increase in length but not height. Only changes in callosal height occurred in conjunction with expansion of the lateral ventricles. This observation is consistent with our earlier conclusions (Pfefferbaum et al., 2000) and also with relation-ships reported by Peterson et al. (Peterson et al., 2001), who used an empirically driven approach to identify latent features of callosal morphology affected by aging studied cross-sectionally. Of eight factors that explained >90% of the observed variance, ventricular enlargement was related to all three factors involved in upward arching (i.e. height) of the corpus callosum. Further, in our genetic analysis (Pfefferbaum et al., 2000), the strong phenotypic relationship detected between callosal height and ventricular size was explained by significant genetic as well as significant environmental influences common to these measures. Thus, while based only on correlational analysis, these converging results suggest a significant causal interaction between structural features of the corpus callosum and lateral ventricles that is mediated by aging.

Age-related expansion of the lateral ventricles has been docu-mented in previous longitudinal studies. The rate of ventricular dilatation observed in the present study, which sampled rather than volumed the entire ventricular space, is largely consistent with other studies in elderly groups. For example, the longi-tudinal study of Resnick et al. (Resnick et al., 2000) reported ~3.8% per year increase in ventricular size in men and the cross-sectional study of Coffey et al. (Coffey et al., 1992) suggested a 3.2% per year increase. Our previous longitudinal study conducted in healthy adults who were younger than the men in the present study revealed a 1.5% ventricular volume increase (Pfefferbaum et al., 1995). Absence of evidence for asymmetrical aging, despite cross-sectional asymmetry (right > left), has also been noted previously (Resnick et al., 2000). Cross-sectional studies of the midsagittal area of the corpus callosum report a range of age-related change from essentially none [reviewed above and including our own (Sullivan et al., 2001b)] to modest correlations with age but without description of the actual rate of change (Janowsky et al., 1996; Hampel et al., 1998; Teipel et al., 1998). Figure 2 of Allen et al. (Allen et al., 1991) suggests 0.2% change per year. Image realignment undoubtedly enhanced our ability to detect ventricular asymmetry as well as relatively small callosal thinning. The meta-analysis of Driesen and Raz indicated modest and non-significant age-related changes in callosal area (Driesen and Raz, 1995). Although the changes we observed in this longitudinal study were indeed small, the overall effect size of total callosal shrinkage, 0.59 ± CI 95% = 0.46–0.73, was greater than the effect size of the correlations between callosal area and age reported in the meta-analysis of 21 cross-sectional studies, 0.24 ± CI 95% = 0.12–0.36 or 0.33 ± CI 95% = 0.22–0.45 (depending on the estimation method). Thus, the upper bound of the CI (0.45) was below the lower bound of the CI 95% of the present study, suggesting underestimation of callosal shrinkage in cross-sectional study. It is clear, how-ever, that none of these studies reported callosal changes of a magnitude matching or even approaching ventricular changes. Thus, despite partial measurement of the lateral ventricles, our sampling was validated against volumetric measurement and was shown adequate to provide a comparison brain measurement to callosal change. Further, the changes may be restricted to older age (i.e. >70 year) and are certainly smaller than the normal between-subject variance, which limits the power of cross-sectional studies to detect such change.

The observed brain structure–function relationships provides support that regional shrinkage of the corpus callosum, although quite small even in the elderly, may have functional relevance by contributing to diminishing efficiency of interhemispheric trans-fer of information [cf. (Woodruff et al., 1997)]. The relationship between thinning of the callosal body and reading speed is consistent with the known reliance of word reading on temporal cortical regions and white matter systems, including the arcuate fasiculus [cf. (Cabeza and Nyberg, 2000; Liotti et al., 2000)]. Such relationships also lend credence to the geometric defi-nitions of callosal subdivisions applied in this study [cf. (Peterson et al., 2001)] and to the multiple component processes assessed by the Stroop Test [e.g. (Pujol et al., 2001)].

In conclusion, these longitudinal data support the contention that differential rates of change occur in different brain regions in normal aging and that rate of change in one region may be independent of rate of change in other regions, even within a brain structure. Further, within the context of longitudinal callosal measurement, the lateral ventricles provide a critical measure of the local environment that can influence change in callosal size and contour. In addition, significant, albeit modest, regional callosal thinning in old age may contribute to age-related performance declines on selective tests of speeded processing, especially those involving efficiency of information sharing across the hemispheres. Whether the same extent and pattern of callosal shrinkage observed in these elderly com-munity dwelling men occur in younger adult men and in women remains to be examined.

Notes

This work was supported by grants from the National Heart, Lung, and Blood Institute (HL51429), National Institute on Aging (AG17919). National Institute on Alcohol Abuse and Alcoholism (AA05965, AA10723), and National Institute of Mental Health (MH58007). We also acknowledge our national collaborators and research technicians of the NHLBI, in particular Charles DeCarli, MD, Terry Reed, PhD, Philip A. Wolf, MD, and Bruce L. Miller, MD, for their rigor in overseeing data collection at their research sites.

Table 1

Effects sizes and confidence intervals (CI 95%) for change in brain structure size

 Mean change (mm2CI 95% Effect size CI 95% 
Corpus callosum 
    Genu 5.55 3.97–7.13 0.47 0.34–0.61 
    Body 7.84 4.99–10.68 0.37 0.24–0.50 
    Splenium 6.10 4.45–7.74 0.50 0.36–0.63 
    Total 19.37 14.96–23.78 0.59 0.47–0.73 
Lateral ventricle 
    Left 24.16 21.35–26.97 1.16 1.02–1.29 
    Right 26.40 23.41–29.38 1.19 1.05–1.32 
 Mean change (mm2CI 95% Effect size CI 95% 
Corpus callosum 
    Genu 5.55 3.97–7.13 0.47 0.34–0.61 
    Body 7.84 4.99–10.68 0.37 0.24–0.50 
    Splenium 6.10 4.45–7.74 0.50 0.36–0.63 
    Total 19.37 14.96–23.78 0.59 0.47–0.73 
Lateral ventricle 
    Left 24.16 21.35–26.97 1.16 1.02–1.29 
    Right 26.40 23.41–29.38 1.19 1.05–1.32 
Table 2

Pearson correlations (r) and P-values between brain measures (% change/year)

  Corpus callosum Lateral ventricles 
  Genu Body Splenium Total Length Height Left Right 
Corpus callosum 
    Body r  0.25 – – – – – – – 
 P  0.0002  
    Splenium r  0.05 0.33 – – – – – – 
 P n.s. 0.0001       
    Total r  0.54 0.86  0.62 – – – – – 
 P  0.0001 0.0001  0.0001      
    Length r  0.25 0.06  0.37 0.20 – – – – 
 P  0.0003 n.s.  0.0001 0.003     
    Height r 0.11 0.27  0.00 0.21 0.19 – – – 
 P n.s. 0.0001 n.s. 0.002 0.0060    
Lateral ventricle 
    Left r –0.06 0.05 –0.01 0.00 0.10 0.31 – – 
 P n.s. n.s. n.s. n.s. n.s. 0.0001   
    Right r –0.08 0.06  0.01 0.01 0.08 0.35 0.79 – 
 P n.s. n.s. n.s. n.s. n.s. 0.0001 0.0001  
    Total r –0.07 0.06  0.00 0.01 0.10 0.35 0.94 0.95 
 P n.s. n.s. n.s. n.s. n.s. 0.0001 0.0001 0.0001 
  Corpus callosum Lateral ventricles 
  Genu Body Splenium Total Length Height Left Right 
Corpus callosum 
    Body r  0.25 – – – – – – – 
 P  0.0002  
    Splenium r  0.05 0.33 – – – – – – 
 P n.s. 0.0001       
    Total r  0.54 0.86  0.62 – – – – – 
 P  0.0001 0.0001  0.0001      
    Length r  0.25 0.06  0.37 0.20 – – – – 
 P  0.0003 n.s.  0.0001 0.003     
    Height r 0.11 0.27  0.00 0.21 0.19 – – – 
 P n.s. 0.0001 n.s. 0.002 0.0060    
Lateral ventricle 
    Left r –0.06 0.05 –0.01 0.00 0.10 0.31 – – 
 P n.s. n.s. n.s. n.s. n.s. 0.0001   
    Right r –0.08 0.06  0.01 0.01 0.08 0.35 0.79 – 
 P n.s. n.s. n.s. n.s. n.s. 0.0001 0.0001  
    Total r –0.07 0.06  0.00 0.01 0.10 0.35 0.94 0.95 
 P n.s. n.s. n.s. n.s. n.s. 0.0001 0.0001 0.0001 
Figure 1.

Example of the outlining of the corpus callosum on the midsagittal slice at the first MRI (top), the second MRI (middle), and an overlay of the two (bottom). In the overlay figure, the earlier study is in white and the later one is in black. Note the increase in callosal height in the later compared with the earlier study.

Figure 1.

Example of the outlining of the corpus callosum on the midsagittal slice at the first MRI (top), the second MRI (middle), and an overlay of the two (bottom). In the overlay figure, the earlier study is in white and the later one is in black. Note the increase in callosal height in the later compared with the earlier study.

Figure 2.

Example of the outlining of the lateral ventricles on each of the coronal slices on which they were measured. The top triplet of figures displays the first MRI, the middle triplet displays the second MRI, and the bottom triplet displays overlays of the two studies, where the earlier study is in gray and the later one is in white.

Figure 2.

Example of the outlining of the lateral ventricles on each of the coronal slices on which they were measured. The top triplet of figures displays the first MRI, the middle triplet displays the second MRI, and the bottom triplet displays overlays of the two studies, where the earlier study is in gray and the later one is in white.

Figure 3.

Mean (± SEM) percent change per year for each brain measure (left) and cognitive measure (right). With aging, values of the ventricles, callosal height and length, and time to complete the Trail Making subtests increased, and values of callosal size, MMSE, and Stroop Test output decreased.

Figure 3.

Mean (± SEM) percent change per year for each brain measure (left) and cognitive measure (right). With aging, values of the ventricles, callosal height and length, and time to complete the Trail Making subtests increased, and values of callosal size, MMSE, and Stroop Test output decreased.

References

Aboitiz F, Rodriguez E, Olivares R, Zaidel E (
1996
) Age-related changes in fibre composition of the human corpus callosum: sex differences.
Neuroreport
 
7
:
1761
–1764.
Allen LS, Richey MF, Chai YM, Gorski RA (
1991
) Sex differences in the corpus callosum of the living human being.
J Neurosci
 
11
:
933
–942.
Banich MT, Milham MP, Atchley R, Cohen NJ, Webb A, Wszalek T, Kramer AF, Liang ZP, Wright A (
2000
) fMRI studies of Stroop tasks reveal unique roles of anterior and posterior brain systems in attentional selection.
J Cogn Neurosci
 
12
:
988
–1000.
Biegon A, Eberling JL, Richardson BC, Roos MS, Wong STS, Reed BR, Jagust WJ (
1994
) Human corpus callosum in aging and Alzheimer's disease: a magnetic resonance imaging study.
Neurobiol Aging
 
15
:
393
–397.
Blatter DD, Bigler ED, Gale SD, Johnson SC, Anderson C, Burnett BM, Parker N, Kurth S, Horn S (
1995
) Quantitative volumetric analysis of brain MRI: normative database spanning five decades of life.
Am J Neuroradiol
 
16
:
241
–245.
Cabeza R, Nyberg L (
2000
) maging cognition II: an empirical review of 275 PET and fMRI studies.
J Cogn Neurosci
 
12
:
1
–47.
Carmelli D, DeCarli C, Swan GE, Jack LA, Reed T, Wolf PA, Miller BL (
1998
) Evidence for genetic variance in white matter hyperintensity volume in normal elderly male twins.
Stroke
 
29
:
1177
–1181.
Carmelli D, DeCarli C, Swan GE, Wolf PA, Read T (
2001
) Differential heritability of regional cortical brain volumes in older male twins.
Neurology
 
56
:
A262
.
Coffey CE, Wilkinson WE, Parashos IA, Soady SAR, Sullivan RJ, Patterson LJ, Figiel GS, Webb MC, Spritzer CE, Djang WT (
1992
) Quantitative cerebral anatomy of the aging human brain — a cross-sectional study using magnetic resonance imaging.
Neurology
 
42
:
527
–536.
Cowell C, Allen L, Zalatimo N, Denenberg V (
1992
) A developmental study of sex by age interactions in human corpus callosum.
Dev Brain Res
 
66
:
187
–192.
Doraiswamy PM, Figiel GS, Husain MM, McDonald WM, Shah SA, Boyko OG, Ellinswood EH, Krishnan KRK (
1991
) Aging of the human corpus callosum: magnetic resonance imaging in normal volunteers.
J Neuropsychiatry Clin Neurosci
 
3
:
392
–397.
Driesen NR, Raz N (
1995
) The influence of sex, age, and handedness on corpus callosum morphology: a meta-analysis.
Psychobiology
 
23
:
240
–247.
Efron B, Tibshirani R (
1986
) Bootstrap methods for standard errors, confidence intervals and other measures of statistical accuracy.
Stat Sci
 
1
:
54
–77.
Feinleib M, Garrison RJ, Fabsitz RR, Christian JC, Hrubec Z, Borhani NO, Kannel WB, Rosenman RR, Schwartz JT, Wagner JO (
1977
) The NHLBI twin study of cardiovascular risk factors: methodology and summary of results.
Am J Epidemiol
 
106
:
284
–295.
Giedd JN, Blumenthal J, Jeffries NO, Castellanos FX, Liu H, Zijdenbos A, Paus T, Evans AC, Rapoport JL (
1999
) Brain development during childhood and adolescence: a longitudinal MRI study [letter].
Nat Neurosci
 
2
:
861
–863.
Gur R, Turetsky B, Matsui M, Yan M, Bilker W, Hughett P, Gur R (
1999
) Sex differences in brain gray and white matter in healthy young adults: correlations with cognitive performance.
J Neurosci
 
19
:
4065
–4072.
Hampel H, Teipel SJ, Alexander GE, Horwitz B, Teichberg D, Schapiro MB, Rapoport SI (
1998
) Corpus callosum atrophy is a possible indicator of region-and cell type-specific neuronal degeneration in Alzheimer Disease.
Arch Neurol
 
55
:
193
–198.
Janowsky JS, Kaye JA, Carper RA (
1996
) Atrophy of the corpus callosum in Alzheimer's disease versus healthy aging.
J Am Geriatr Soc
 
44
:
798
–803.
Jernigan TL, Archibald SL, Berhow MT, Sowell ER, Foster DS, Hesselink JR (
1991
) Cerebral structure on MRI. 1. Localization of age-related changes.
Biol Psychiatry
 
29
:
55
–67.
Johnson SC, Farnworth T, Pinkston JB, Bigler ED, Blatter DD (
1994
) Corpus callosum surface area across the human adult life span: effect of age and gender.
Brain Res Bull
 
35
:
373
–377.
Laissy JP, Patrux B, Duchateau C, Hannequin D, Hugonet P, Ait-Yahia H, Thiebot J (
1993
) Midsagittal MR measurements of the corpus callosum in healthy subjects and diseased patients: a prospective survey.
Am J Neuroradiol
 
14
:
145
–154.
Lange N, Giedd JN, Castellanos FX, Vaituzis AC, Rapoport JL (
1997
) Variability of human brain structure size: ages 4–20 years.
Psychiatry Res Neuroimaging
 
74
:
1
–12.
Lezak MD (1995) Neuropsychological assessment, 3rd edn. New York: Oxford University Press.
Liotti M, Woldorff MG, Perez R, Mayberg HS (
2000
) An ERP study of the temporal course of the Stroop color–word interference effect.
Neuropsychologia
 
38
:
701
–711.
Oldfield RC (
1971
) The assessment and analysis of handedness: the Edinburgh inventory.
Neuropsychologia
 
9
:
97
–113.
Pandya DN, Seltzer B (1986) The topography of commissural fibers. In: Two hemispheres — one brain: functions of the corpus callosum (Lepore F, Ptito M, Jasper HH, eds), pp. 47–74. New York: Alan R. Liss.
Peterson BS, Feineigle PA, Staib LH, Gore JC (
2001
) Automated measurement of latent morphological features in the human corpus callosum.
Hum Brain Mapp
 
12
:
232
–245.
Pfefferbaum A, Sullivan EV, Jernigan TL, Zipursky RB, Rosenbloom MJ, Yesavage JA, Tinklenberg JR (
1990
) A quantitative analysis of CT and cognitive measures in normal aging and Alzheimer's disease.
Psychiatry Res Neuroimaging
 
35
:
115
–136.
Pfefferbaum A, Mathalon DH, Sullivan EV, Rawles JM, Zipursky RB, Lim KO (
1994
) A quantitative magnetic resonance imaging study of changes in brain morphology from infancy to late adulthood.
Arch Neurol
 
51
:
874
–887.
Pfefferbaum A, Sullivan EV, Mathalon DH, Shear PK, Rosenbloom MJ, Lim KO (
1995
) Longitudinal changes in magnetic resonance imaging brain volumes in abstinent and relapsed alcoholics.
Alcohol Clin Exp Res
 
19
:
1177
–1191.
Pfefferbaum A, Lim KO, Desmond J, Sullivan EV (
1996
) Thinning of the corpus callosum in older alcoholic men: a magnetic resonance imaging study.
Alcohol Clin Exp Res
 
20
:
752
–757.
Pfefferbaum A, Sullivan EV, Rosenbloom MJ, Mathalon DH, Lim KO (
1998
) A controlled study of cortical gray matter and ventricular changes in alcoholic men over a five year interval.
Arch Gen Psychiatry
 
55
:
905
–912.
Pfefferbaum A, Sullivan EV, Swan GE, Carmelli D (
2000
) Brain structure in men remains highly heritable in the seventh and eighth decades of life.
Neurobiol Aging
 
21
:
63
–74.
Pfefferbaum A, Sullivan EV, Carmelli D (
2001
) Genetic regulation of regional microstructure of the corpus callosum in late life.
Neuroreport
 
12
:
1677
–1681.
Pozzilli C, Bastianello S, Bozzao A, Pierallini A, Giubilei F, Arentino C, Bozzao L (
1994
) No differences in corpus callosum size by sex and aging. A quantitative study using magnetic resonance imaging.
J Neuroimaging
 
4
:
218
–221.
Pujol J, Vendrell P, Deus J, Junque C, Bello J, Mari-Vilalta JL, Capdevila A (
2001
) The effect of medial frontal and posterior parietal demyelin-ating lesions on Stroop interference.
NeuroImage
 
13
:
68
–75.
Raz N, Gunning FM, Head D, Dupuis JH, McQuain J, Briggs SD, Loken WJ, Thornton AE, Acker JD (
1997
) Selective aging of the human cerebral cortex observed in vivo: differential vulnerability of the prefrontal gray matter.
Cereb Cortex
 
7
:
268
–282.
Resnick SM, Goldszal AF, Davatzikos C, Golski S, Kraut MA, Metter EJ, Bryan N, Zonderman AB (
2000
) One-year changes in MRI brain volumes in older adults.
Cereb Cortex
 
10
:
464
–472.
Salthouse TA (
2000
) Aging and measures of processing speed.
Biol Psychol
 
54
:
35
–54.
Shear PK, Sullivan EV, Mathalon DH, Lim KO, Davis LF, Yesavage JA, Tinklenberg JR, Pfefferbaum A (
1995
) Longitudinal volumetric computed tomographic analysis of regional brain changes in normal aging and Alzheimer's disease.
Arch Neurol
 
52
:
392
–404.
Stroop JR (
1935
) Studies of interference in serial verbal reactions.
J Exp Psychol
 
18
:
643
–662.
Sullivan EV, Pfefferbaum A, Swan G, Carmelli D (
2001
) Heritability of hippocampal size in elderly twin men: equivalent influence from genes and environment.
Hippocampus
 
11
:
754
–762.
Sullivan EV, Rosenbloom M, Desmond J, Pfefferbaum A (
2001
) Sex differences in corpus callosum size: relationship to age and intracranial size.
Neurobiol Aging
 
22
:
603
–611.
Swan GE, Reed T, Jack LM, Miller BL, Markee T, Wolf PA, DeCarli C, Carmelli D (
1999
) Differential genetic influence for components of memory in aging adult twins.
Arch Neurol
 
56
:
1127
–1132.
Teipel J, Hampel H, Alexander GE, Schapiro MB, Horwitz B, Teichberg D, Daley E, Hippius H, Moller H-J, Rapoport SI (
1998
) Dissociation between corpus callosum atrophy and white matter pathology in Alzheimer's disease.
Neurology
 
51
:
1381
–1385.
Thompson P, Narr KL, Blanton RE, Toga AW (2002) Mapping structural alterations of the corpus callosum during brain development and degeneration. Proceedings of the NATO ASI on the Corpus Callosum (Zaidel E, Iacoboni M, eds). New York: Kluwer Academic, in press.
Weis S, Kimbacher M, Wenger E, Neuhold A (
1993
) Morphometric analysis of the corpus callosum using MR: correlation of measurement with aging in healthy individuals.
Am J Neuroradiol
 
14
:
637
–645.
Woodruff PWR, Phillips ML, Rushe T, Wright IC, Murray RM, David AS (
1997
) Corpus callosum size and inter-hemispheric function in schizophrenia.
Schizophr Res
 
23
:
189
–196.
Yakovlev PI, Lecours A-R (1967) The myelogenetic cycles of regional maturation of the brain. In: Regional development of the brain in early life (Minkowski A, ed.), pp. 3–70. Oxford: Blackwell Scientific.