Postmortem heart weight: relation to body size and effects of cardiovascular disease and cancer

Cardiovascular Pathology 23 (2014) 5–11

Contents lists available at ScienceDirect

Cardiovascular Pathology

Original Article

Postmortem heart weight: relation to body size and effects of cardiovascular disease and cancer Neena Theresa Kumar a,⁎, Knut Liestøl b, Else Marit Løberg a, Henrik Mikael Reims c, Jan Mæhlen a a b c

Department of Pathology, Oslo University Hospital-Ullevål, Kirkeveien 166, 0407 Oslo, Norway Department of Informatics, The Faculty of Mathematics and Natural Sciences, University of Oslo, Oslo, Norway epartment of Pathology, Oslo University Hospital-Rikshospitalet, Oslo, Norway

a r t i c l e

i n f o

Article history: Received 8 March 2013 Received in revised form 29 August 2013 Accepted 3 September 2013 Keywords: Heart weight Body weight Autopsy Cancer Cardiovascular diseases

a b s t r a c t Background: Gender, body weight, and cardiovascular disease (CVD) are all variables known to influence human heart weight. The impact of cancer is less studied, and the influence of age is not unequivocal. We aimed to describe the relationship between body size and heart weight in a large autopsy cohort and to compare heart weight in patients with cancer, CVD, and other diseases. Methods and Results: Registered information, including cause of death, evidence of cancer and/or CVD, heart weight, body weight, and height, was extracted from the autopsy reports of 1410 persons (805 men, mean age 66.5 years and 605 women, mean age 70.6 years). The study population was divided in four groups according to cause of death; cancer (n=349), CVD (n=470), mixed group who died from cancer and CVD and/or lung disease (n=263), and a reference group with patients who did not die from any of these conditions (n=328). In this last group, heart weight correlated only slightly better with body surface area than body weight, and nomograms based on body weight are presented. Compared to the reference group (mean heart weight: 426 g and 351 g in men and women, respectively), heart weight was significantly lower (men: Pb.05, women: Pb.001) in cancer patients (men: 392 g, women: 309 g) and higher (Pb.001) in patients who died from CVD (men: 550 g, women: 430 g). Similar results were obtained in linear regression models adjusted for body weight and age. Among CVD, heart valve disease had the greatest impact on heart weight, followed by old myocardial infarction, coronary atherosclerosis, and hypertension. Absolute heart weight decreased with age, but we demonstrated an increase relative to body weight. Conclusion: The weight of the human heart is influenced by various disease processes, in addition to body weight, gender, and age. While the most prevalent types of CVD are associated with increased heart weight, patients who die from cancer have lower average heart weight than other patient groups. The latter finding, however, is diminished when adjusting for body weight. Summary: The present study demonstrates that the weight of the human heart is influenced by various disease processes like cancer and CVD, in addition to body weight, gender and, possibly, age. © 2014 Elsevier Inc. All rights reserved.

1. Introduction The weight of the human heart may be influenced by various factors, including body size, gender and, possibly, age, in addition to a number of disease processes. Heart weight increases with body weight [1,2], but recent data [3,4] suggest that the normal heart weight/body weight ratio has increased compared to standard tables from the early 20th century. Men have higher heart weight than women [5], also for equal body weight [6,7]. Some studies [1] have found no correlation between heart weight and age, while

⁎ Corresponding author. Department of Pathology, Oslo University Hospital-Ullevål, P.O. Box 4956, Nydalen, 0424 Oslo, Norway. Tel.: +47 22 11 89 10, +47 22 11 89 13; fax: +47 22 11 82 39. E-mail address: [email protected] (N.T. Kumar). 1054-8807/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.carpath.2013.09.001

others [7,8] have suggested that aging is associated with an increase in heart weight. Importantly, cardiovascular diseases (CVDs), including hypertension, valvular heart disease, myocardial infarction (MI), and congestive heart failure, are well known to be associated with myocardial hypertrophy [9–13]. However, coronary atherosclerosis in the absence of MI is not consistently associated with increased heart weight [14–17]. The presence of chronic lung disease may also affect heart weight due to pulmonary hypertension [18,19]. Atrophy of the heart in conjunction with chronic systemic illness was described by Senac as early as 1749 [20], but the effects of cancer on human heart structure and function are not well known. In a previous study [21], our results suggested lower heart weight among patients who died from cancer, than among other patients. In the present study, we aimed to describe the relationship between body size and heart weight in a large autopsy cohort from

6

N.T. Kumar et al. / Cardiovascular Pathology 23 (2014) 5–11

the current era and to compare heart weight in patients with cancer, CVD, and other diseases. Moreover, we wanted to assess the relationship between heart weight and age, the duration of cancer, and atherosclerosis.

2. Methods 2.1. Study population The study was carried out at the Department of Pathology, Oslo University Hospital–Ullevål, Norway, which serves a large area in southeastern Norway. We obtained data collected among 1633 patients subjected to ordinary medical autopsy during the period from August 1996 to December 2000. The data were sampled consecutively from patients aged 20 years and older. Patients without complete autopsy and patients with missing information about heart weight, body weight, or body height were excluded. This left a total of 1410 persons comprising 57% men and 43% women. The mean age was 68.3 years (range: 20–99).

2.2. Registration of data All the autopsies were carried out at the same academic pathology department by the pathologists on duty and according to the same protocol. Standardized information was extracted from the autopsy reports by three experienced pathologists. Information about the patients' disease history was registered based on the medical records of the hospital. Autopsy findings, including the patients' height and body weight, together with heart weight, coronary atherosclerosis, aortic atherosclerosis and its complications, MI, cardiac dilation and/or hypertrophy, and aortic and/or mitral valve disease (grouped as “heart valve disease”), were registered. Findings of cerebrovascular and neurodegenerative disease, cancer with or without metastases, lung disease, alcohol-related disease, and renal disease were also registered. The degree of coronary and aortic atherosclerosis was assessed as described previously [21]. Briefly, coronary atherosclerosis was classified as none, moderate, or severe and atherosclerosis in the aorta as none/mild, moderate, severe, or severe with complications. The grading was based on the visual findings stated in the autopsy report. MI was categorized as none, recent, old, or both recent and old based on microscopic findings. Cerebral findings were categorized as infarction, hemorrhage, and neurodegenerative changes, also based on microscopic findings. We evaluated and registered the underlying, immediate, and contributing causes of death determined at autopsy (based on the Systematized Nomenclature of Medicine classification system) [22]. Cases in which the original conclusion was considered questionable were discussed among the investigators in order to reach consensus. We also discussed other problems or divergences to achieve the most consistent information in our database.

2.3. Assessment of heart weight, body weight, and body height The hearts had been opened according to standard procedure, and blood and blood clots had been washed out, while epicardial fat and approximately 1 cm of the aortic and pulmonary arteries were included. They were weighed during autopsy to the nearest tenth gram (scales from A/S Viig & Vraalsen & A. P. Foss, Oslo). The bodies were weighed naked the day of the autopsy. Body height was measured from head to heel when the patients were placed on the autopsy table.

2.4. Statistical analysis The study population was divided into four groups according to underlying cause of death; CVD, cancer, other diseases (assigned to be the reference group), and a mixed group. The mixed group consisted of patients with both CVD and cancer (one of which the underlying cause of death and one of which a contributing cause of death), of patients with other diseases as cause of death concurrent with CVD or cancer contributing to death, and of patients with primary lung disease. This group was created in order to minimize potentially opposite effects, and primary lung disease was included because of the possibility of cor pulmonale. The reference group consisted of patients who died from other diseases than cancer, CVD, and lung disease but included some patients with various combinations of CVD (hypertension, n=38; MI, n=35; heart valve disease, n=18) and/or cancer (n=20) that did not cause or contribute to death. Since excluding these patients from the reference group did not significantly change the values of the patients' age, heart weight, or body size measures, they were included in all the analyses. The categorical variable MI was analyzed using dummy variables for three categories (no, recent, or old/recent and old MI). Heart valve disease, hypertension, and coronary and aortic atherosclerosis were analyzed as categorical variables with two categories (presence or absence of findings). Body surface area (BSA) was calculated using the formula 0.007184×body weight(kg) 0.425×height(cm) 0.725. The body mass index (BMI) was calculated using body weight(kg)/height(m) 2. An additional standardized variable (“heart weight deviance”) was calculated as the difference between actual heart weight and expected heart weight divided by the standard deviation (S.D.) of the distribution of the difference. Thus, a value of 0 indicates the expected heart weight given the subject's body weight and gender, while a value of 1 indicates that the heart weight is 1 S.D. higher than its expected value. Here, expected heart weight was calculated for men and women separately from heart weights in the reference group, using regression analysis with correction for body weight. Heart weight was also expressed as percentage of body weight (“relative heart weight”). Student's t test was used to compare two groups, and one-way analysis of variance (ANOVA) with Bonferroni correction was used to compare three or more groups. Pearson correlation was used to assess the relation between the natural logarithm (ln) of heart weight and ln(body weight) and between ln(heart weight) and ln(BSA). Univariate and multiple linear regression analyses were used to further examine the association between heart weight and other variables, and corresponding regression coefficients (β) are presented. Linear regression analyses were used to study the association between heart weight and cancer survival time, while logistic regression analyses were used to relate atherosclerosis and cancer survival time. All analyses were carried out using Statistical Package for the Social Sciences (SPSS) 19 and 20/JMP 9.0 statistical system. 2.5. Ethics The study was approved by the Regional Ethical Committee in Norway (Regional komité for medisinsk og helsefaglig forskningsetikk Sør-Øst). 3. Results 3.1. Subject characteristics Table 1 shows means and S.D. for age, body weight, height, BMI, BSA, heart weight, and heart weight deviance separately for the two genders in all the patients and according to cause of death. In the total study population, all parameters except BMI and heart weight

N.T. Kumar et al. / Cardiovascular Pathology 23 (2014) 5–11

7

Table 1 Characteristics of the total study population and disease groups Men

n Age (years) Body weight (kg) Height (m) BMI (kg/m2) BSA (m2) Heart weight (g) Heart weight deviance

Women

All

Reference group

CVD

Cancer

Mixed

All

Reference group

CVD

Cancer

Mixed

805 66.5 74.8 175 24.4 1.89 475 0.52

168 61.3 75.8 175 24.8 1.90 426 0.01

300 68.4 78.5 175 25.5 1.93 550 1.18

197 65.8 68.3 175 22.2 1.82 392 -0.15

140 69.8 74.6 174 24.5 1.88 490 0.67

605 70.6 62.5 161 23.9 1.65 374 0.38

160 67.0 65.6 161 25.0 1.68 351 0.01

170 73.8 63.8 160 24.7 1.66 430 1.15

152 68.1 59.5 162 22.5 1.62 309 -0.53

123 73.7(11.7)† 60.6 (18.3) 161 (7) 23.2 (6.1) 1.62 (0.23) 409 (121)† 0.94 (1.29)†

(14.3)⁎ (18.2)⁎ (8)⁎ (5.3) (0.23)⁎ (140)⁎ (1.23)

(17.1) (18.8) (8) (5.5) (0.24) (108) (1.00)

(12.5)† (18.4) (8) (5.2) (0.23) (140)† (1.07)†

(14.3)‡ (15.9)† (8) (4.7)† (0.21)‡ (89)‡ (1.01)

(12.3)† (17.6) (7) (5.4) (0.21) (148)† (1.31)†

(13.4) (18.5) (8) (6.5) (0.23) (105) (1.35)

(15.2) (21.7) (8) (7.5) (0.27) (76) (1.00)

(11.5)† (16.7) (7) (6.0) (0.20) (105)† (1.29)†

(13.6) (16.6)‡ (8) (5.8)‡ (0.22) (66)† (1.08)†

Total N=1410. Mean values (S.D.). ⁎ Pb.001 vs. women (t test). † Pb.001 vs. reference group (ANOVA). ‡ Pb.05 vs. reference group (ANOVA).

In the reference group, ln(heart weight) was strongly and significantly correlated to ln(body weight) (r=0.609, Pb.001) and ln(BSA) (r=0.638, Pb.001). Nomograms showing the relationship between heart weight and body weight are presented for men and women separately (Fig. 2). Relative heart weight was 0.58% in the reference group, and there was no significant difference between men and women. It ranged from 0.48% in the highest weight quartile (weight above 83 kg) to 0.69% in the lowest weight quartile (weight below 55 kg). Among men in the reference group, mean heart weight was lower with increasing age across all age quartiles (Fig. 3A). Among women, heart weight decreased in a similar fashion across the upper three quartiles but was lowest in the youngest quartile (Fig. 3A). Body weight varied in a similar manner across age quartiles (Fig. 3B). Thus, when adjusted for gender and body weight in a linear regression model, heart weight increased with increasing age (Pb.001). Similarly, heart weight deviance increased with higher age quartiles (Fig. 3C). This association was slightly stronger among women (P=.001) than among men (P=.012) based on a linear regression model.

group (Table 1). Accordingly, the heart weight deviance was negative in cancer patients and positive in CVD patients. Further, age at death was higher in CVD patients than in cancer patients, both being significantly higher than in the reference group. Among men, body weight, BMI, and BSA were significantly lower in cancer patients and nonsignificantly higher in patients who died from CVD, compared to the reference group. In women, body weight and BMI were significantly lower in cancer patients compared to the reference group. For women who died from CVD, body weight, BMI, and BSA were nonsignificantly lower than in the reference group but higher than in the other disease groups (Table 1). The relationship between heart weight and body weight in the main disease groups (mixed group excluded) is shown in scatter plots for men and women separately (Fig. 4). For any given body weight, heart weight tended to be higher in CVD patients and lower in cancer patients than in the reference group. In a linear regression model with adjustment for body weight, gender, and age, heart weight was significantly lower in patients who died from cancer and higher in patients who died from CVD than in the reference group (Table 2). Results were comparable for heart weight deviance (results not shown). When the genders were analyzed separately, heart weight was also significantly higher in those who died from CVD than in the reference group. Heart weight was numerically lower in the cancer group compared to the reference group for both genders but significantly only among women (Table 2). This analysis was also carried out adding categorical variables for coronary atherosclerosis, heart valve disease, hypertension, and recent or old MI, which did not change the regression coefficients notably (results not shown).

3.3. Differences between disease groups

3.4. Heart weight and its association with CVD and COPD

Mean heart weight was significantly lower in cancer patients and higher in patients who died from CVD, compared to the reference

Within the CVD group, 363 patients (77%) died from heart disease, 45 (10%) from cerebral stroke, and 62 (13%) from other types of CVD

deviance differed significantly between men and women. The distribution of heart weight, body weight, and BSA in all the patients is shown more closely in histograms (Fig. 1). Of all patients, 37.5% were recorded with MI, 19.6% with hypertension, 10.9% with valvular heart disease, 19.6% with chronic obstructive pulmonary disease (COPD), and 38.4% with cancer (regardless of the cause of death). 3.2. Heart weight in relation to body size and age

Fig. 1. Histograms of heart weight (g), body weight (kg), and BSA in the total study population (N=1410).

8

N.T. Kumar et al. / Cardiovascular Pathology 23 (2014) 5–11

Fig. 2. Nomograms relating body weight (kg) to heart weight (g) in women and men. Centiles: solid line, 50%; dotted line, 2.5% and 97.5%, dashed line, 10% and 90%. Scale is logarithmic.

(mainly ruptured aortic aneurisms). Patients who died from heart disease had significantly heavier hearts than those who died from cerebral stroke among both men (558 g vs. 477 g, Pb.05) and women (442 g vs. 372 g, Pb.05). Patients who died from other CVDs had similar heart weights as those who died from heart disease (541 g and 416 g in men and women, respectively). In linear regression models adjusted for age, gender, and body weight, patients with valvular heart disease, hypertension, and old MI had significantly heavier hearts than those without these conditions in the total study population, as well as among patients who died from CVD (Table 3). Recent MI and coronary atherosclerosis were significantly associated with increased heart weight among all the patients but not among patients who died from CVD. There was no difference in mean heart weight between patients recorded with COPD and those without (regardless of the cause of death), although among women, standardized heart weight was significantly higher in the COPD group. Only 25 patients were recorded to die from COPD. Mean heart weight was slightly but nonsignificantly higher for both men and women compared to the reference group. 3.5. Duration of cancer Mean heart weight and heart weight deviance were slightly lower among patients whose cancer had been diagnosed more than 1 year

A

B

prior to death compared to those with a more recent cancer diagnosis, although the differences were not significant (Table 4). There were similar differences between patients who had lived with cancer for more than 5 years and less than 5 years, but these differences were smaller when expressed as heart weight deviance. In linear regression analyses adjusted for age and body weight, the same tendencies were present (not shown). In logistic regression analysis adjusted for age and body weight, patients with cancer for more than 1 year had less severe aortic atherosclerosis than patients with cancer for less than 1 year [odds ratio (OR) 0.54, Pb.05]. Although not statistically significant, coronary atherosclerosis also tended to be less severe in the former group of patients (OR: 0.59, P=.064). 4. Discussion In this autopsy study, we have described heart weights in patients with different medical conditions, with special emphasis on cancer and CVD. Patients who died from cancer had lower mean heart weight than the reference group, also when taking body weight and gender into consideration. Although not statistically significant, regression analyses also suggested a negative relationship between cancer survival time and heart weight. Not unexpectedly, we found higher heart weights among patients who died from CVD. Among all patients, heart valve disease, MI, hypertension, and coronary atherosclerosis

C

Fig. 3. Heart weight and body weight in age quartiles. (A) Mean heart weight; (B) mean body weight; (C) mean heart weight deviance. All figures relate to the reference group (n=328).

N.T. Kumar et al. / Cardiovascular Pathology 23 (2014) 5–11

9

Fig. 4. Scatter plots of heart weight (g) and body weight (kg) in women and men. ○ (black): reference group; ○ (green): death by CVD; ○ (blue): death by cancer. Fit line in corresponding color. Scale is logarithmic.

were associated with higher heart weights. We found heart weight to be slightly more strongly correlated with BSA than with body weight, both correlations being significant. Further, while absolute heart weight decreased with higher age, adjustment for body weight and gender suggested a relative increase in heart weight. Consistent with global trends, the prevalence of obesity has increased in Norway during the past decades [23–25]. Overweight and obesity are associated with arterial hypertension and other conditions [26] that may affect heart weight. However, Norway and the other Scandinavian countries have a lower prevalence of overweight and obesity compared to other industrialized countries [27]. Therefore, these populations should be well suited for studies of factors that influence heart weight independently of body fatness. As discussed previously [21], our study population is fairly representative of the general population, although there may be some selection bias, particularly among the oldest patients. Echocardiographically assessed left ventricular mass has been shown to be more closely related to BSA than to other measures of body size [28]. In a recent autopsy study, Gaitskell et al. [3] also found heart weight to correlate slightly better with BSA than with body weight, although the difference between the correlation coefficients was not statistically significant. By contrast, Kitzman et al. [29] found body weight to predict normal heart weight better than BSA. Thus, despite a slightly higher correlation coefficient when relating heart weight to BSA in our reference group, we decided to use body weight as a simple and robust measure of body size in the main analyses. Nomograms relating heart weight to body weight display differences between the reference group in our study populations

and the study population of Gaitskell et al. [3]; for a given body weight, patients in the present study population have a slightly higher heart weight. The difference in heart weight may be explained by the inclusion of younger patients (age range: 14–98 years) in the study by Gaitskell et al. [3], while in the present study the age range was 20–99 years. In addition, the former study did not exclude patients who died from cancer. In addition, relative heart weight was somewhat higher in our reference group than in the study population of Gaitskell et al. [3] (0.58% vs. 0.51%), reflecting a higher mean heart weight and a slightly lower mean body weight in the present study. Patients who died from cancer had lower mean heart weight but also lower mean body weight than patients in the other study groups. Although not significant for men when genders were analyzed separately, there was a tendency towards lower heart weights among cancer patients also after adjustment for body weight. Cancer is associated with cachexia, defined as loss of skeletal muscle mass with or without loss of fat mass [30]. Since the amount of epicardial fat correlates with various indices of adiposity [31]; at least part of the reduction of heart weight is probably caused by possible loss of fat. To our knowledge, very few studies have addressed cardiac changes in human cancer. However, in the 1950s, a study of 85 cases with atrophy of the heart [20] suggested malignancy as one important etiological factor, and in a case-based study published in 1968, Burch et al. [32] observed smaller hearts and electrocardiographic changes in cancer patients. Moreover, several animal studies [33–37] have found evidence of cancer-induced cardiac atrophy and remodeling. Thus, one may speculate that the lower heart weight observed in cancer patients in the present study could partly be caused by myocardial atrophy.

Table 2 Linear regression analysis with heart weight as the dependent variable and death cause, gender, age, and body weight as covariates All

Gender Age Body weight Cancer CVD Mixed

Men

Women

β (CI)

P

β (CI)

P

β (CI)

P

59 1.0 3.1 −20 95 62

b.001 b.001 b.001 .010 b.001 b.001

-

-

-

1.0 (0.4-1.6) 3.4 (3.0-3.9) −14 (−37 to 9) 107 (86–128) 59 (33–84)

.001 b.001 .241 b.001 b.001

1.1 (0.6-1.6) 2.8 (2.4-3.1) −26 (−44 to −8) 76 (59–94) 64 (45–84)

b.001 b.001 .005 b.001 b.001

(48–70) (0.6–1.4) (2.8–3.4) (−35 to −5) (81–110) (46–78)

Total N=1410. Cause of death (cancer, CVD, and mixed CVD–cancer–lung disease) compared to the reference group. Values are adjusted for age (years), gender (women=0 and men=1), and body weight (kg). Regression coefficients (β) equal change in heart weight (g).

10

N.T. Kumar et al. / Cardiovascular Pathology 23 (2014) 5–11

Table 3 Linear regression analysis with heart weight as the dependent variable and different CVDs as covariates All

Patients who died from CVD

β (CI) n Gender Age Body weight Recent MI Old MI Hypertension Coronary atherosclerosis Heart valve disease

1410 50 (39–61) 0.3 (−0.1 to 0.8) 3.3 (3.0–3.6) 29 (8–49) 54 (41–66) 33 (20–46) 36 (21–50) 120 (104–137)

P

β (CI)

P

b.001 .161 b.001 .006 b.001 b.001 b.001 b.001

470 69 0.2 3.5 −10 31 24 15 113

b.001 .625 b.001 .482 .008 .025 .498 b.001

(47–91) (−0.7 to 1.2) (2.9–4.1) (−39 to 18) (8–54) (3–45) (−28 to 57) (87–139)

Recent and old MI, hypertension, and heart valve disease are compared to the absence of these covariates. Moderate and severe coronary atherosclerosis is compared to no atherosclerosis. Values are adjusted for age (years), gender (women=0 and men=1), and body weight (kg). Regression coefficients (β) equal change in heart weight (g). CI, 95% confidence interval.

Patients suffering from carcinoma have previously been shown to have less pronounced atherosclerosis than patients without cancer [38]. In the present study, coronary and aortic atherosclerosis tended to be less severe in patients with a longer duration of cancer. These findings support the hypothesis that cancer is associated with regression of atherosclerosis [39]. Possible mechanisms may include anorexia, reduced body weight, and altered lipid metabolism. In the present study, heart valve disease, hypertension, MI, and coronary atherosclerosis were independently associated with higher heart weight. The first three of these conditions are well known to cause cardiac hypertrophy [10,40]. However, the association between coronary atherosclerosis and heart weight is somewhat controversial and is supported by some studies [13,15] but not by others [16,17]. Such an association might be explained by ischemia with secondary fibrosis and/or compensatory hypertrophy and by humoral factors [15,41]. Our study indicates some effect of atherosclerosis, but the predictive value is low within the group who died from CVD. Pulmonary hypertension and cor pulmonale may complicate chronic lung disease of which COPD is most important [18]. Among those patients dying from COPD in the present study, heart weight was slightly but nonsignificantly higher than in the reference group. However, the sample size is small, and we may not have had the statistical power to detect a significant effect on heart weight in the patients most severely affected by COPD. Age- and gender-related variations in normal heart weight have been attributed to various factors, including the amount of epicardial fat and cardiac connective tissue, depositions of senile amyloid, and factors related to physical exercise [29]. In our reference group, mean heart weight decreased with age. Similar findings were reported in a Japanese autopsy study [42] of organ weights in patients aged over 60 years. In addition, Kitzman et al. [29] found a decreased mean heart weight with increasing age in an autopsy material, but only during the

Table 4 Mean heart weight, heart weight deviance, and body weight according to survival time from a cancer diagnosis n

Heart weight (g)

233 98 304 27

359 343 357 326

Heart weight deviance

P Cancer Cancer Cancer Cancer

b1 year N1 year b5 years N5 years

.098 .054

Body weight (kg)

P −0.27 −0.50 −0.33 −0.45

.069 .569

P 64 65 65 56

.688 .001

Patients with cancer as the underlying cause of death (total N=331). P values based on t test.

7th through 10th decades of life. Other studies have found either no association between age and adult heart weight [1,5] or suggested the opposite, that is, higher heart weight with higher age [43,44]. This discrepancy may be partly explained by demographic differences. In the latter studies, one based on a forensic population [43] and the other on healthy individuals with radiographic measurements of cardiac size [44], the mean age was less than 50 years. As discussed by others [5,45], mean heart weight may increase during the early decades of life, but the age distribution in our study did not allow for such analyses. By multiple regression analysis and analysis of heart weight deviance, we found that heart weight relative to body weight increased with advancing age, in contrast to absolute heart weight. This may be partly explained by a higher prevalence of hypertension in older age groups. Due to the retrospective nature of the study, we were not able to obtain certain data that may have been of interest. For example, since left ventricular weight and wall thickness were not systematically measured, we could not assess left ventricular hypertrophy. Moreover, we did not systematically obtain information about chemo- and radiotherapy, both of which may have adverse cardiovascular effects associated with myocardial hypertrophy and contribute to coronary disease [46,47]. Such effects however, may have masked some of the findings in the cancer group. We did not have access to systematic information about previous blood pressure measurements. The prevalence of hypertension is approximately 40% in the Norwegian and other European adult populations [48,49] compared to 20% registered in the present study. Thus, the actual number of patients with a history of hypertension was probably higher than we were able to detect. The mean age was lower in the reference group than in the other groups, probably reflecting a higher proportion of deaths from less age-dependent causes (e.g., accidents, alcoholism, and infections). Approximately one third of the patients in the reference group were recorded with MI, heart valve disease, hypertension, or cancer or a combination of these diseases but not considered as causes of death. However, excluding these patients did not change the values of the key variables in this group notably. We have presented regression data and nomograms describing the relationship between body weight and heart weight in a Norwegian autopsy cohort. This study also demonstrates that the weight of the human heart is influenced by various disease processes, in addition to body weight, gender and, possibly, age. While the most prevalent types of CVD are associated with increased heart weight, patients who die from cancer have lower average heart weight than other patient groups. In part, this is attributed to lower body weight, but the finding should encourage further investigation into cardiac effects of human cancer. Acknowledgments This work was supported by the Department of Pathology, Oslo University Hospital, Oslo, Norway, and The Research Council of Norway (Norges forskningsråd). References [1] Smith LH. The relation of the weight of the heart to the weight of the body and of the weight of the heart to age. Am Heart J 1928;4:79–93. [2] Hanzlick R, Rydzewski D. Heart weights of white men 20 to 39 years of age. An analysis of 218 autopsy cases. Am J Forensic Med Pathol 1990;11:202–4. [3] Gaitskell K, Perera R, Soilleux EJ. Derivation of new reference tables for human heart weights in light of increasing body mass index. J Clin Pathol 2011;64: 358–62. [4] Lucas SB. Derivation of new reference tables for human heart weights in light of increasing body mass index. J Clin Pathol 2011;64:279–80. [5] Dadgar SK, Tyagi SP, Singh RP, Hameed S. Factors influencing the normal heart weight–a study of 140 hearts. Jpn Circ J 1979;43:77–82.

N.T. Kumar et al. / Cardiovascular Pathology 23 (2014) 5–11 [6] Gardin JM, Wagenknecht LE, Anton-Culver H, Flack J, Gidding S, Kurosaki T, et al. Relationship of cardiovascular risk factors to echocardiographic left ventricular mass in healthy young black and white adult men and women. The CARDIA study. Coronary Artery Risk Development in Young Adults. Circulation 1995;92:380–7. [7] de la Grandmaison GL, Clairand I, Durigon M. Organ weight in 684 adult autopsies: new tables for a Caucasoid population. Forensic Sci Int 2001;119:149–54. [8] He Q, Heshka S, Albu J, Boxt L, Krasnow N, Elia M, et al. Smaller organ mass with greater age, except for heart. J Appl Physiol 2009;106:1780–4. [9] Kumar V, Abbas AK, Fausto N, Aster JC. Robbins and Cotran Pathologic Basis of Disease. Elsevier: Saunders; 2010 533–7. [10] Reiner L, Mazzoleni A, Rodriguez FL, Freudenthal RR. The weight of the human heart. II. Hypertensive cases. Arch Pathol 1961;71:180–201. [11] Ishijima M. Histopathological study on myocardial hypertrophy associated with ischemic heart disease. Jpn Circ J 1990;54:616–23. [12] Roberts WC, Potkin BN, Solus DE, Reddy SG. Mode of death, frequency of healed and acute myocardial infarction, number of major epicardial coronary arteries severely narrowed by atherosclerotic plaque, and heart weight in fatal atherosclerotic coronary artery disease: analysis of 889 patients studied at necropsy. J Am Coll Cardiol 1990;15:196–203. [13] Zhang M, Shields J, Zhang Y, Li L, Fowler D, Zhao Z, et al. Correlation between coronary plaque burden and heart weight. Pathol Res Pract 2012;208:610–4. [14] Davis D, Blumgart HL. Cardiac hypertrophy: its relation to coronary arteriosclerosis and congestive heart failure. Ann Intern Med 1937;11:1024–38. [15] Dean JH, Gallagher PJ. Cardiac ischemia and cardiac hypertrophy. An autopsy study. Arch Pathol Lab Med 1980;104:175–8. [16] Correa P, Restrepo C, Montenegro MR, Solberg LA. Coronary atherosclerosis and heart weight. Lab Invest 1968;18:580–5. [17] Romppanen T, Seppa A, Roilas H. Ischemic heart disease and heart weight. Cardiology 1983;70:206–12. [18] Weitzenblum E, Chaouat A. Cor pulmonale. Chron Respir Dis 2009;6:177–85. [19] Packe GE, Cayton RM, Edwards CW. Comparison of right ventricular weight at necropsy in interstitial pulmonary fibrosis and in chronic bronchitis and emphysema. J Clin Pathol 1986;39:594–5. [20] Hellerstein HK, Santiago-Stevenson D. Atrophy of the heart; a correlative study of 85 proved cases. Circulation 1950;1:93–126. [21] Kumar NT, Liestol K, Loberg EM, Reims HM, Brorson SH, Maehlen J. The apolipoprotein E polymorphism and cardiovascular diseases-an autopsy study. Cardiovasc Pathol 2012;21:461–9. [22] SNOMED CT, International Health Terminology Standards Development Organisation Web site. http://www.ihtsdo.org/index.php?id=545. Accessed 4 March 2013. [23] James WP. The epidemiology of obesity: the size of the problem. J Intern Med 2008;263:336–52. [24] Sund ER, Jones A, Midthjell K. Individual, family, and area predictors of BMI and BMI change in an adult Norwegian population: findings from the HUNT study. Soc Sci Med 2010;70:1194–202. [25] Reas DL, Nygard JF, Svensson E, Sorensen T, Sandanger I. Changes in body mass index by age, gender, and socio-economic status among a cohort of Norwegian men and women (1990-2001). BMC Public Health 2007;7:269. [26] National Heart, Lung, and Blood Institute (NHLBI) Obesity Education Initiative. Lung, and Blood Institute (NHLBI) Obesity Education Initiative. Clinical Guidelines on the Identification, Evaluation, and Treatment of Overweight and Obesity in Adults: The Evidence Report: NIH Publication No. 98-4083. Bethesda, MD: U.S. Department of Health and Human Services, Public Health Service, NIH, NHBLI; 1998. [27] Ulset E, Undheim R, Malterud K. Has the obesity epidemic reached Norway? Tidsskr Nor Laegeforen 2007;127:34–7.

11

[28] Devereux RB, Lutas EM, Casale PN, Kligfield P, Eisenberg RR, Hammond IW, et al. Standardization of M-mode echocardiographic left ventricular anatomic measurements. J Am Coll Cardiol 1984;4:1222–30. [29] Kitzman DW, Scholz DG, Hagen PT, Ilstrup DM, Edwards WD. Age-related changes in normal human hearts during the first 10 decades of life. Part II (Maturity): a quantitative anatomic study of 765 specimens from subjects 20 to 99 years old. Mayo Clin Proc 1988;63:137–46. [30] Fearon K, Strasser F, Anker SD, Bosaeus I, Bruera E, Fainsinger RL, et al. Definition and classification of cancer cachexia: an international consensus. Lancet Oncol 2011;12:489–95. [31] Rabkin SW. Epicardial fat: properties, function and relationship to obesity. Obes Rev 2007;8:253–61. [32] Burch GE, Phillips JH, Ansari A. The cachetic heart. A clinico-pathologic, electrocardiographic and roentgenographic entity. Dis Chest 1968;54:403–9. [33] Cosper PF, Leinwand LA. Cancer causes cardiac atrophy and autophagy in a sexually dimorphic manner. Cancer Res 2011;71:1710–20. [34] Muhlfeld C, Das SK, Heinzel FR, Schmidt A, Post H, Schauer S, et al. Cancer induces cardiomyocyte remodeling and hypoinnervation in the left ventricle of the mouse heart. PLoS One 2011;6:e20424. [35] Tian M, Nishijima Y, Asp ML, Stout MB, Reiser PJ, Belury MA. Cardiac alterations in cancer-induced cachexia in mice. Int J Oncol 2010;37:347–53. [36] Tian M, Asp ML, Nishijima Y, Belury MA. Evidence for cardiac atrophic remodeling in cancer-induced cachexia in mice. Int J Oncol 2011;39:1321–6. [37] Xu H, Crawford D, Hutchinson KR, Youtz DJ, Lucchesi PA, Velten M, et al. Myocardial dysfunction in an animal model of cancer cachexia. Life Sci 2011;88: 406–10. [38] Wanscher O, Clemmesen J, Nielsen A. Negative correlation between atherosclerosis and carcinoma. Br J Cancer 1951;5:172–4. [39] Malinow MR, Senner JW. Arterial pathology in cancer patients suggests atherosclerosis regression. Med Hypotheses 1983;11:353–7. [40] Levy D, Anderson KM, Savage DD, Kannel WB, Christiansen JC, Castelli WP. Echocardiographically detected left ventricular hypertrophy: prevalence and risk factors. The Framingham Heart Study. Ann Intern Med 1988;108:7–13. [41] Okada R, Kawai S, Ishijima M. Relation between coronary stenosis and myocardial lesions determined by a semiquantitative approach to myocardial fibrosis and hypertrophy due to ischemia. Am J Cardiol 1989;63:2E–6E. [42] Inoue T, Otsu S. Statistical analysis of the organ weights in 1,000 autopsy cases of Japanese aged over 60 years. Acta Pathol Jpn 1987;37:343–59. [43] Sheikhazadi A, Sadr SS, Ghadyani MH, Taheri SK, Manouchehri AA, Nazparvar B, et al. Study of the normal internal organ weights in Tehran's population. J Forensic Leg Med 2010;17:78–83. [44] Sjogren AL. Left ventricular wall thickness determined by ultrasound in 100 subjects without heart disease. Chest 1971;60:341–6. [45] Mandal R, Loeffler AG, Salamat S, Fritsch MK. Organ weight changes associated with body mass index determined from a medical autopsy population. Am J Forensic Med Pathol 2012;33:382–9. [46] Yeh ET, Bickford CL. Cardiovascular complications of cancer therapy: incidence, pathogenesis, diagnosis, and management. J Am Coll Cardiol 2009;53:2231–47. [47] Darby SC, Cutter DJ, Boerma M, Constine LS, Fajardo LF, Kodama K, et al. Radiationrelated heart disease: current knowledge and future prospects. Int J Radiat Oncol Biol Phys 2010;76:656–65. [48] Klouman M, Asberg A, Wideroe TE. The blood pressure level in a Norwegian population–the significance of inheritance and lifestyle. Tidsskr Nor Laegeforen 2011;131:1185–9. [49] Wolf-Maier K, Cooper RS, Banegas JR, Giampaoli S, Hense HW, Joffres M, et al. Hypertension prevalence and blood pressure levels in 6 European countries, Canada, and the United States. JAMA 2003;289:2363–9.