Weight loss and progression of early atherosclerosis in the ... - Nature

International Journal of Obesity (1999) 23, 948±956 ß 1999 Stockton Press All rights reserved 0307±0565/99 $15.00 http://www.stockton-press.co.uk/ijo

Weight loss and progression of early atherosclerosis in the carotid artery: a four-year controlled study of obese subjects K Karason1*, J Wikstrand2, L SjoÈstroÈm3 and I Wendelhag2 1

Department of Cardiology, 2Wallenberg Laboratory for Cardiovascular Research, and 3Department of Medicine, Sahlgrenska University Hospital, Gothenburg, Sweden

OBJECTIVE: To investigate the extent of carotid artery atherosclerosis in obese subjects and to examine the possible effects of weight loss on atherosclerotic development. DESIGN: Controlled 4 y intervention study. SUBJECTS: 20 obese patients treated with weight-reducing gastroplasty, 19 obese patients treated with dietary recommendations and 35 lean subjects. MEASUREMENTS: Body weight, blood pressure, blood lipids, glucose and insulin were measured. A B-mode ultrasound was recorded to determine the intima-media thickness (IMT) and lumen diameter (LD) of the carotid artery. Study groups were investigated at baseline and re-examined after 3 to 4 y of follow-up. RESULTS: At baseline, obese patients had higher blood pressure, serum total cholesterol, triglycerides, glucose and insulin compared with lean subjects; they also had a larger IMT in the carotid artery bulb (P < 0.05) and a larger LD in the common carotid artery (P < 0.01). After 4 y of follow-up, obese patients treated with surgery displayed a mean weight loss of 22 kg (19%), while the average weight in the obese control group remained unchanged (P < 0.001). The weight loss group showed improvements in blood pressure, HDL-cholesterol, triglycerides and insulin compared with the obese control group (P < 0.05). The progression rate of carotid bulb IMT in the weight loss group was similar to that observed in the lean control group (0.024 vs 0.025 mm=y, n.s.), whereas the IMT progression rate was almost three times higher in the obese control group (0.068 mm=y, P < 0.05 compared with lean controls). CONCLUSION: Obese people have an unfavourable risk factor pro®le and signs of premature carotid artery atherosclerosis. Weight loss is followed by an improvement in several risk factors and may reduce the progression rate of atherosclerotic changes in the carotid artery bulb. Keywords: weight loss; intima-media thickness; atherosclerosis; carotid artery; ultrasound

Introduction Obesity is associated with multiple risk factors for atherosclerosis, including hypertension, dyslipidaemia and impaired glucose tolerance.1 ± 3 Furthermore, several epidemiological studies have demonstrated that obesity is associated with an increase in the incidence of cardiovascular disease, in particular myocardial infarction and stroke.4 ± 6 Despite this evidence, cross-sectional studies based on angiography and autopsy have not been able to demonstrate a consistent relationship between obesity and the degree of atherosclerosis.7,8 In an attempt to explain this controversy, it has been suggested that obesity may be related more to the precipitation of cardiovascular events, e.g. thrombus formation, than to the underlying atherosclerotic process. *Correspondence: Kristjan Karason, Department of Cardiology, Sahlgrenska University Hospital, SE-413 45 Gothenburg, Sweden. Email: [email protected] Received 26 October 1998; revised 2 February 1999; accepted 13 April 1999

Several studies have shown that weight loss is associated with a reduction in cardiovascular risk factors.9 ± 11 However, the bene®t of weight loss with respect to cardiovascular complications remains undetermined. Controlled intervention studies addressing this issue are lacking and observational studies, which are limited by their inability to differentiate between voluntary and involuntary weight loss, have reported either a decrease or an increase in cardiovascular mortality following weight reduction.12,13 The possible effects of weight loss on the progress of atherosclerosis have not been investigated. B-mode ultrasound has been applied in clinical research in order non-invasively to study the atherosclerotic process in the carotid artery. It is assumed that atherosclerotic changes in the carotid artery re¯ect general atherosclerosis, and ultrasound measurements of intima-media thickness in the carotid artery have been used as a surrogate measurement of coronary atherosclerosis.14 ± 16 In the present study, we used B-mode ultrasound to examine the extent of carotid artery atherosclerosis in obesity and to investigate the possible effects of weight loss on the progression rate of atherosclerotic changes.

Weight loss and carotid artery atherosclerosis K Karason et al

Subjects and methods Study groups

In all, 74 subjects from the city of Gothenburg and the surrounding areas, comprising 56 males and 18 females with ages ranging from 28 to 63 years, were enrolled in the study. At baseline, 39 obese patients with a body mass index (BMI) of 30 ± 40 kg=m2 and 35 non-obese subjects with a body mass index of 20 ± 28 kg=m2 were evaluated. The obese and non-obese study groups were matched with respect to age, gender and height. After baseline assessment, the patients with obesity were assigned to one of two groups: the `obese operation' group including 20 patients referred to weight-reducing gastric surgery and the `obese control' group, consisting of 19 obese subjects who received conventional dietary recommendations. These two groups were matched with respect to age, gender, height and weight. Nineteen subjects from the obese operation group and 17 subjects from the obese control group returned for re-examination after 4 y of follow-up. Three-year follow-up data were available for 29 subjects in the non-obese control group. The obese patients were recruited from the ongoing Swedish Obese Subjects (SOS) study, which is a nationwide trial designed to determine whether the mortality rates among obese people who lose weight by surgical means differ from those in an obese reference sample.17 The non-obese study group was chosen from a sample of adults who had been randomly selected from the population registry of Gothenburg and had served as a normal reference group in our laboratory. Anthropometry and blood pressure

Body weight was measured with the subjects wearing light clothing and no shoes and rounded off to the nearest 0.1 kg. Height measurements were rounded off to the nearest 0.01 m. BMI was calculated as the weight (kg) divided by the square of the height (m2). Systolic and diastolic (phase V) blood pressure were measured in the right arm using a mercury sphygmomanometer after 10 min of supine rest. An appropriate cuff, with a width of at least 40% of the circumference of the subject's arm, was applied. Biochemical analysis

Blood samples were obtained in the morning after a 10 ± 12 h fasting period. Total cholesterol and triglyceride concentrations in serum were determined using enzymatic techniques. High-density lipoprotein (HDL)-cholesterol was determined after the precipitation of low-density lipoprotein (LDL) with heparin and manganese chloride. LDL-cholesterol was calculated according to the equation developed by Friedewald et al.18 Blood glucose was measured

enzymatically, whereas plasma insulin was determined by radioimmunoassay. Ultrasonography

Ultrasound studies of the right carotid artery were performed with the subject in the supine position. A commercially available ultrasound system (Acuson 128 XP, Mountain View, CA) with a 7 MHz linear transducer was used. Images for intima-media thickness (IMT) were recorded from the common carotid artery and the carotid artery bulb. Three different images at the position of the thickest part of the far wall (visually judged) were obtained from each arterial segment and recorded on video tape. In order to minimize variability due to the cardiac cycle, images were captured by electrocardiographic triggering on the top of the R wave. Immediately after the examination, a study protocol was ®lled in, including a description of the probe angle, as well as a drawing of the vessel topography. This protocol, which did not contain any information relating to the degree of atherosclerosis, was used as a guide at the follow-up investigation.19 Intima-media thickness and lumen diameter

The ultrasound images from the video tape were analysed in a computerized analysing system based on automated detection of the echo structures, with the option of making manual corrections via the operator. IMT was de®ned as the distance from the leading edge of the lumen ± intima interface to the leading edge of the media ± adventitia interface of the far wall and lumen diameter (LD) was de®ned as the distance between the leading edges of the intima ± lumen interface of the near wall and the lumen ± intima interface of the far wall. The measurement of IMT in the carotid artery was made in two separate segments: along a 10-mm long segment of the common carotid artery; and along a 10-mm long segment in the carotid artery bulb. The program gives the mean and the maximum thickness of the intima ± media complex, as well as the mean lumen diameter in the common carotid artery.20 The progression rate of IMT expressed in mm=y was calculated in order to permit comparisons between study groups with different observation intervals. The axial resolution of the ultrasound system, as well as the measurement precision obtained with our analysing procedure, are illustrated in Figure 1. The theoretical resolution of the ultrasound system is about 0.3 mm, meaning that if the intima media complex is thinner than this, it cannot be measured. This was not the case in any of our study subjects. The measurement precision, on the other hand, depends on the analysing system and for IMT is approximately 0.005 mm, when a mean value of three measured images is used.21 Most of the recordings were performed by a single laboratory technologist with vast experience of ultra-

949


950

Figure 1 Axial resolution of the ultrasound system and measurement precision for IMT illustrated schematically. Upper panel: The theoretical axial resolution of a 7-MHz transducer is approximately 0.3 mm. This means if intima plus media (IMT) is thinner than 0.3 mm the two echo interfaces cannot be separated and measurement of the intima ± media complex is not possible (situation A). If IMT is thicker than 0.3 mm two separate echoes may be recorded and IMT measured (situation B). Lower panel: The measurement error of IMT over a 10-mm long segment is approximately 0.009 mm when 10 independent pairs of measurement points are used. A mean value of three images reduces the measurement error to approximately 0.005 mm. Adapted with permission from ref. 21.

sonography (IW) and the images were stored digitally. Measurements were made afterwards by the same examiner, now blinded with respect to the study subjects. Estimation of intra-observer variability for the actual examiner, including both registration and reading variability, has given coef®cients of variation of 10.6% for the common carotid artery and 13.2% for

the carotid artery bulb.22 Furthermore, variation coef®cients for re-reading images in the present study were 1.2% and 3.8% for the respective arterial segments. Only good quality readings with clearly delineated echo interfaces were included in analyses. Because obese subjects often have an accumulation of adipose


tissue in the neck area their carotid vessel anatomy can be dif®cult to visualize. Due to drop out (three obese patients and six lean subjects) and these technical problems, the availability of good quality readings from both baseline and follow-up was reduced. Complete data from the common carotid artery and carotid artery bulb were available for 69% and 59% of the obese subjects respectively, and the corresponding ®gures for lean subjects were 83% and 71%. There were no baseline differences between subjects for whom IMT data were available and those for whom they were not. Statistical analyses

Statistical analyses were performed with the Statview and SAS software packages (SAS Institute, Cary, NC). Data are given as the mean s.d. Comparisons between groups at baseline and follow-up were made using the w2 test (categorical variables) and the MannWhitney U test (continuous variables) and comparisons within groups were made using Wilcoxon's signed rank test. Comparisons of the IMT progression rate between the three groups were made using analysis of covariance with adjustment for baseline IMT. The t distribution was used to calculate a 95% con®dence interval for differences between groups. All the probability values were derived from twotailed tests and a P-value of < 0.05 was considered signi®cant.

Results At baseline, obese and non-obese subjects were similar with respect to gender, age, height and smoking habits (Table 1). By de®nition, obese patients had a higher body weight and BMI than lean subjects; they also had higher systolic and diastolic blood pressure. Obese subjects had higher serum total cholesterol, triglycerides, glucose and insulin compared with lean subjects, while LDL-cholesterol and HDL-cholesterol did not differ between the two groups. As shown in Table 2, obese patients had an 8% larger lumen diameter in the common carotid artery compared with lean subjects (P < 0.01) and a 19% larger

951 Table 1 Clinical characteristics of obese and lean subjects at baseline Obese Lean (n 39) (n 35) No. (%) of men Smoking status, no. (%) Current smoker Former smoker Never smoker No. (%) on antihypertensive treatment Age (y) Height (cm) Weight (kg) Body mass index (kg=m2) Systolic blood pressure (mmHg) Diastolic blood pressure (mmHg) Total cholesterol (mmol=l) LDL-cholesterol (mmol=l) HDL-cholesterol (mmol=l) Triglycerides (mmol=l) Glucose (mmol=l) Insulin (mU=l)

31 (79)

27 (77)

9 19 11 8

12 10 13 0

(23) (49) (28) (21)

(34) (29) (37) (0)

49 5 49 8 176 8 176 8 116 14 76 11 37 4 24 2 141 18 124 14 87 11

74 9

5.9 1.1 5.5 0.7 3.8 1.0 1.2 0.3 2.2 1.2 5.6 2.2 19 10

mean (mm) maximum (mm) mean (mm) maximum (mm)

95% CI 95% con®dence interval. Values are given as the mean s.d. *P < 0.05, **P < 0.01.

0(73 ± 3) 0(74 ± 4) 40(34 ± 46) 13(11 ± 15) 17(9 ± 25)*** 13(8 ± 18)*** 0 .4 (0.1 ± 0.7)*

3.5 0.7 0 .3 (70.2 ± 0.8) 1.3 0.4 70 .1 (70.3 ± 0.1) 1.4 0.6 0 .8 (0.3 ± 1.3)*** 4.2 0.6 1 .4 (0.6 ± 2.2)*** 11 4 8(3 ± 13)**

95% CI 95% con®dence interval; LDL low density lipo-protein; HDL high density lipoprotein. Values are given as the mean s.d. *P < 0.05, **P < 0.01, ***P < 0.001.

intima±media thickness in the carotid artery bulb (P < 0.05). In the common carotid artery, however, the difference in IMT between obese and lean subjects was not statistically signi®cant. At baseline, the obese operation and obese control groups were similar in terms of clinical variables and ultrasound measurements, except for slightly higher diastolic blood pressure in the obese operation group (Tables 3 and 4). During follow-up, three former smokers in the obese operation group started to smoke again and one patient in the obese control group quit smoking. Obese patients who were receiving antihypertensive medication at baseline were still on the same therapy at follow-up, however, the dosage in the obese operation group had been reduced (data not shown). After 4 y of follow-up, the obese operation group displayed a mean weight loss of 22 kg (19%), while the average weight in the obese control group remained unchanged (Table 3). The weight loss group displayed reductions in systolic and diastolic

Table 2 Intima-media thickness and lumen diameter of obese and lean subjects at baseline

Common carotid artery Intima-media thickness Intima-media thickness Lumen diameter (mm) Carotid artery bulb Intima-media thickness Intima-media thickness

Difference (95% CI)

Obese

Lean

n 31 0.81 0.16 1.02 0.18 6.71 0.90 n 26 1.01 0.44 1.45 0.60

n 35 0.76 0.18 0.94 0.20 6.20 1.12 n 34 0.85 0.25 1.19 0.34

Difference (95% CI) 0.05 (70.03 ± 0.13) 0.08 (70.01 ± 0.17) 0.51 (0.07 ± 0.94)** 0.16 (0.01 ± 0.31)* 0.26 (0.02 ± 0.50)*


952 Table 3 Clinical characteristics of obese operation and obese control groups at baseline and change at four-year follow-upa Obese operation (n 19)

Obese control (n 17)

118 15 722 10{{{

115 14 0 13

38.0 3.6 76.8 3.5{{{

37.6 4.9 0.2 4.2

146 20 718 21{{

139 15 4 17

7(75 ± 19) 722(734 ± 78)**

91 10 710 12{{

83 10 1 8

8(2 ± 14)* 711(718 ± 74)**

6.1 1.1 70.5 0.7{{

5.9 1.2 70.5 0.6{

0.2 (70.4 ± 0.6) 0.0 (70.4 ± 0.4)

3.9 1.2 70.4 0.7{

3.7 0.9 70.4 0.7{

0.2 (70.7 ± 0.9) 0.0 (70.5 ± 0.5)

1.2 0.3 0.2 0.3{

1.2 0.3 70.1 0.3

0.0 (70.2 ± 0.2) 0.3 (0.1 ± 0.5)*

2.5 1.4 70.9 0.9{{{

2.0 1.1 70.2 0.9

0.5 (70.4 ± 1.4) 70.7 (71.3 ± 70.1)*

5.2 1.7 70.3 1.0

6.4 2.6 70.2 2.6

71.2 (72.7 ± 0.3) 70.1 (71.3 ± 1.2)

Weight (kg) Baseline Change at follow-up Body mass index (kg=m2) Baseline Change at follow-up Systolic blood pressure (mmHg) Baseline Change at follow-up Diastolic blood pressure (mmHg) Baseline Change at follow-up Total cholesterol (mmol=l) Baseline Change at follow-up LDL-cholesterol (mmol=l) Baseline Change at follow-up HDL-cholesterol (mmol=l) Baseline Change at follow-up Triglycerides (mmol=l) Baseline Change at follow-up Glucose (mmol=l) Baseline Change at follow-up Insulin (mU=l) Baseline Change at follow-up

21 10 78 11{{

19 11 0 7

Difference (95% CI) 3(77 ± 13) 722(729 ± 714)*** 0.4 (72.2 ± 3.3) 77.0 (79.7 ± 74.3)***

2(74 ± 8) 78(715 ± 71)*

a Clinical characteristics of the lean control group showed no signi®cant changes between baseline and follow-up. 95% CI 95% con®dence intervals. Values are given as the mean s.d. *P < 0.05; **P < 0.01; ***P < 0.001. {P < 0.05; {{P < 0.01; {{{P < 0.001 compared with baseline values.

Table 4 Intima-media thickness and lumen diameter in the obese operation and obese control groups at baseline and change at four-year follow-upa

Common carotid artery Intima-media thickness Baseline value Change at follow-up Intima-media thickness Baseline values Change at follow-up Lumen diameter (mm) Baseline value Change at follow-up Carotid artery bulb Intima-media thickness Baseline value Change at follow-up Intima-media thickness Baseline value Change at follow-up

mean (mm) maximum (mm)

mean (mm) maximum (mm)

Obese operation

Obese control

Difference (95% CI)

n 14

n 13

0.78 0.13 0.06 0.20

0.83 0.13 0.04 0.09

70.05 (70.15 ± 0.05) 0.02 (70.10 ± 0.14)

0.99 0.11 0.02 0.19

1.04 0.19 0.04 0.13

70.05 (70.17 ± 0.07) 70.02 (70.15 ± 0.11)

6.58 0.94 70.03 0.46 n 14

6.87 0.94 70.03 0.31 n9

70.29 (71.03 ± 0.45) 0.00 (70.31 ± 0.31)

0.93 0.26 0.10 0.25

0.91 0.17 0.26 0.27*

0.02 (70.18 ± 0.22) 70.16 (70.39 ± 0.07)

1.40 0.47 0.09 0.42

1.25 0.24 0.36 0.41*

0.15 (70.20 ± 0.50) 70.27 (70.64 ± 0.10)

a

Only subjects with good quality ultrasound recordings both at baseline and follow-up are included in these analyses. 95% CI 95% con®dence intervals. Values are given as the mean s.d. *P < 0.05 compared with baseline values.

blood pressure compared with the obese control group and showed decreases in serum triglycerides and insulin and an increase in HDL-cholesterol. As shown in Table 4, the obese control group displayed a 29% increase in both the mean and the maximum

IMT in the carotid artery bulb (P < 0.05), while the corresponding values for the weight loss group were only 11% and 6% (n.s.). However, when changes were compared between obese groups, the differences did not reach statistical signi®cance. In the common


953

Figure 2 Mean ( s.e.) progression rate in intima±media thickness in the carotid artery bulb in the three study groups.

carotid artery, no signi®cant differences in IMT or LD were observed within or between the obese groups at follow-up. After 3 y of follow-up the lean control subjects did not display any signi®cant changes in body weight, blood pressure or metabolic variables. One lean subject stopped smoking (data not shown). The mean and maximum IMT in the carotid artery bulb increased by 9% and 11% respectively (P < 0.05). No signi®cant increase was observed in common carotid IMT or LD. The rate of progression of IMT (mm=y) in the carotid bulb for all three study groups is shown in Figure 2. The progression rate was almost three times higher in the obese control group compared with both the lean group and the weight loss group. However, only the difference between the obese control and the lean group reached statistical signi®cance.

Discussion The main ®ndings in the present study are that obese subjects have a thicker intima ± media complex in the carotid artery bulb and a larger lumen diameter in the common carotid artery compared with lean subjects. Furthermore, obese subjects who lose weight appear to have a lower progression rate of IMT compared with subjects with persistent obesity. These observations are of importance since ultrasound measurements of carotid IMT are not only an estimation of

atherosclerosis in the carotid artery but also a surrogate measurement of coronary atherosclerosis.16 Despite the accumulation of cardiovascular risk factors in obese subjects, previous studies have not been able to identify a clear relationship between obesity and atherosclerosis. In 1983, Solberg and Strong reviewed autopsy studies from 24 geographical locations and concluded that there were no consistent associations between obesity and atherosclerotic disease.7 Furthermore, in a review article from 1985, Barrett-Connor stated that, despite biological plausibility, the literature did not support a causal relationship between obesity and atherosclerosis.23 In accordance with this, neither cross-sectional nor longitudinal angiographic studies have been able to demonstrate a consistent association between obesity and coronary atherosclerosis.8,24 ± 28 However, some of these studies have demonstrated a signi®cant relationship between visceral adipose tissue and coronary artery disease.8,25 ± 27 Since autopsy and angiography studies have not been able to demonstrate a consistent relationship between obesity and the degree of atherosclerosis, it has been speculated that the increased frequency of cardiovascular events in obese subjects may be related to disease processes other than atherosclerosis, such as endothelial dysfunction and thrombus formation.29 This is supported by the presence of abnormal ®brinolytic activity in obese patients.30,31 However, it should be noted that angiography only provides information about relatively advanced atherosclerotic lesions leading to the narrowing of the vessel lumen


954

and that autopsy studies, despite permitting a more accurate estimation of atherosclerotic disease, are limited because of their cross-sectional nature and selective study samples. B-mode ultrasound, on the other hand, offers an opportunity non-invasively to quantify early atherosclerotic disease in the carotid artery and to detect small changes in wall thickness occurring over time. Because of these advantages the ultrasound technique has been used to estimate atherosclerotic disease in several longitudinal population and intervention studies.19,32 ± 34 In accordance with previous studies, obese patients in this study had an adverse risk factor pro®le with increments in blood pressure, serum total cholesterol, triglycerides, glucose and insulin. One novel ®nding was that obese patients had a thicker intima ± media complex in the carotid artery bulb compared with lean subjects and a higher IMT progression rate during 4 y of follow-up. These observations strongly suggest that obesity predisposes subjects to the development of atherosclerosis and that the disease process is likely to be mediated by haemodynamic and metabolic risk factors. The results are consistent with the ®ndings of the Honolulu Heart Program, in which obesity predicted the degree of coronary artery disease assessed angiographically,35 and the more recent Bogalusa Heart Study, where body fatness was related to the extent of atherosclerotic lesions found at autopsy in young people who had died of trauma.36 In the present study, obese subjects had more pronounced intima ± media thickening in the carotid artery bulb than in the common carotid artery. This is consistent with previous experience that atherosclerotic lesions occur more frequently in arterial regions with ¯ow separation and low shear stress, which is typical of the carotid bifurcation area. The importance of atherosclerosis in the carotid bifurcation is further emphasized in a recent study by Hulthe et al, showing that IMT measured in the carotid bulb correlates better with coronary atherosclerosis than IMT measured in the common carotid artery.16 As expected, weight loss in our obese subjects was associated with improvements in several risk factors including reductions in blood pressure, triglycerides and insulin and an increase in HDL-cholesterol. One important observation was that the obese control group displayed a signi®cant increase in carotid bulb IMT at follow-up, while the weight loss group did not. The progression rate of IMT in the weight loss group was only about one-third of that found in the obese control group and was similar to that observed in lean subjects. Although the difference between the weight loss group and the obese control group did not reach statistical signi®cance, this could be due to small sample sizes (beta error). Taken together our ®ndings suggest that weight loss, along with concomitant risk factor improvements, may reduce the progression rate of atherosclerosis. However, it should be noted that factors other than weight reduction, such as changes in dietary habits and enhanced physical activity, also

could contribute to retardation of the atherosclerotic process. Another interesting observation in our study was that obese subjects had a larger lumen diameter in the common carotid artery compared with lean subjects. Previous studies have shown that obesity is associated with an elevation in blood volume and cardiac output and an enlargement in the arterial lumen may therefore be an adaptation to augmented haemodynamic load. Furthermore, it can be speculated that increments in LD may lead to a compensatory thickening of the arterial wall in order to maintain normal wall stress. However, haemodynamic adaptation of this kind is unlikely to be the sole explanation of obesityrelated intima ± media thickening in the carotid bulb, a region known to be prone to atherosclerosis. Although weight reduction has been shown to be associated with a regression of haemodynamic load,37 ± 39 no changes were observed in lumen diameter in our weight loss subjects at follow-up. This suggests that obesity-related increments in arterial lumen may be irreversible, or my need longer time to normalize. Limitations of the study

The present study has several limitations, some of which are addressed below. Firstly, the study was neither blinded nor randomized due to practical and ethical reasons. However, in order to compensate for this, the two obese groups were matched to one another with respect to several clinical variables. Secondly, dif®culties involved in the ultrasound assessment of carotid vessel anatomy in obese subjects resulted in missing data. Nonetheless, we feel that our conclusions are valid, since obese patients who could not be included in the analyses did not differ signi®cantly from patients included with respect to gender distribution or any other measured clinical variable. Thirdly, the power of our study for detecting the true effects of weight loss on IMT progression rate was low due to small sample sizes. This may explain why the large difference in IMT progression rate between obese operation and obese control subjects did not reach statistical signi®cance. Further largescale studies are thus needed to con®rm our ®ndings.

Conclusions We conclude that middle-aged obese people have an unfavourable risk factor pro®le and signs of premature carotid artery atherosclerosis. The carotid artery lumen is enlarged in obese subjects and this may be related to an increase in haemodynamic load in these patients. Weight loss is followed by an improvement in several risk factors and may also reduce the progression rate of atherosclerotic changes in the carotid artery bulb.


Acknowledgements

This study was supported by grants from Sahlgrenska University Hospital, the GoÈteborg Medical Society, the Swedish Heart Lung Foundation, the Swedish Medical Research Council (grant no. 05239 and 10880) and Hoffman-La Roche. We thank Caroline Schmidt and Johannes Hulthe for their help with data collection.

References

1 Chiang BN, Perlman LV, Epstein FH. Overweight and hypertension. A review. Circulation 1969; 39: 403 ± 421. 2 Kannel WB, Gordon T, Castelli WP. Obesity, lipids, and glucose intolerance. The Framingham Study. Am J Clin Nutr 1979; 32: 1238 ± 1245. 3 SjoÈstroÈm CD, Hakangard AC, Lissner L, SjoÈstroÈm L. Body compartment and subcutaneous adipose tissue distribution Ð risk factor patterns in obese subjects. Obes Res 1995; 3: 9 ± 22. 4 Hubert HB, Feinleib M, McNamara PM, Castelli WP. Obesity as an independent risk factor for cardiovascular disease: a 26year follow-up of participants in the Framingham Heart Study. Circulation 1983; 67: 968 ± 977. 5 Manson JE, Colditz GA, Stampfer MJ, Willett WC, Rosner B, Monson RR, et al. A prospective study of obesity and risk of coronary heart disease in women. N Engl J Med 1990; 322: 882 ± 889. 6 SjoÈstroÈm LV. Mortality of severely obese subjects. Am J Clin Nutr 1992; 55: 516S ± 523S. 7 Solberg LA, Strong JP. Risk factors and atherosclerotic lesions. A review of autopsy studies. Arteriosclerosis 1983; 3: 187 ± 198. 8 Hodgson JM, Wahlqvist ML, Balazs ND, Boxall JA. Coronary artherosclerosis in relation to body fatness and its distribution. Int J Obese Relat Metab Disord 1994; 18: 41 ± 46. 9 Ashley FW, Jr., Kannel WB. Relation of weight change to changes in atherogenic traits: the Framingham Study. J Chronic Dis 1974; 27: 103 ± 114. 10 Reisin E, Abel R, Modan M, Silverberg D, Eliahou H, Modan B. Effect of weight loss without salt restriction on the reduction of blood pressure in overweight hypertensive patients. N Engl J Med 1978; 298: 1 ± 6. 11 SjoÈstroÈm CD, Lissner L, SjoÈstroÈm L. Relationships between changes in body composition and changes in cardiovascular risk factors: the SOS Intervention Study. Swedish Obese Subjects. Obes Res 1997; 5: 519 ± 530. 12 Williamson DF, Pamuk ER. The association between weight loss and increased longevity. A review of the evidence. Ann Intern Med 1993; 119: 731 ± 736. 13 Pamuk ER, Williamson DF, Serdula MK, Madans J, Byers TE. Weight loss and subsequent death in a cohort of U.S. adults. Ann Intern Med 1993; 119: 744 ± 748. 14 Margitic SE, Bond MG, Crouse JR, Furberg CD, Probst®eld JL. Progression and regression of carotid atherosclerosis in clinical trials. Arterioscler Thromb 1991; 11: 443 ± 451. 15 Wikstrand J, Wiklund O. Frontiers in cardiovascular science. Quantitative measurements of atherosclerotic manifestations in humans. Arterioscler Thromb 1992; 12: 114 ± 119. 16 Hulthe J, Wikstrand J, Emanuelsson H, Wiklund O, de Feyter PJ, Wendelhag I. Atherosclerotic changes in the carotid artery bulb as measured by B-mode ultrasound are associated with the extent of coronary atherosclerosis Stroke 1997; 28: 1189 ± 1194. 17 SjoÈstroÈm L, Larsson B, Backman L, Bengtsson C, Bouchard C, Dahlgren S, et al. Swedish obese subjects (SOS). Recruitment for an intervention study and a selected description of the obese state. Int J Obes Relat Metab Disord 1992; 16: 465 ± 479.

18 Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972; 18: 499 ± 502. 19 Wendelhag I, Wiklund O, Wikstrand J. Intima-media thickness after cholesterol lowering in familial hypercholesterolemia. A three-year ultrasound study of common carotid and femoral arteries. Atherosclerosis 1995; 117: 225 ± 236. 20 Wendelhag I, Liang Q, Gustavsson T, Wikstrand J. A new automated computerized analyzing system simpli®es readings and reduces the variability in ultrasound measurement of intima-media thickness. Stroke 1997; 28: 2195 ± 2200. 21 Schmidt C, Wendelhag I. How can the variability in ultrasound measurement of intima-media thickness be reduced? Studies of interobserver variability in carotid and femoral arteries. Clin Physiol 1999; 19: 45 ± 55. 22 Wendelhag I. Ultrasound measurement of intima-media thickness and atherosclerotic plaques in carotid and femoral arteries. Methodological Studies and a Five-year Observational Study in Familial Hypercholesterolemia. Thesis. GoÈteborg: GoÈteborg University, 1995. 23 Barrett Connor EL. Obesity, atherosclerosis, and coronary artery disease. Ann Int Med 1985; 103: 1010 ± 1019. 24 Kramer JR, Matsuda Y, Mulligan JC, Aronow M, Proud®t WL. Progression of coronary atherosclerosis. Circulation 1981; 63: 519 ± 526. 25 Zamboni M, Armellini F, Sheiban I, De Marchi M, Todesco T, Bergamo Andreis IA, et al. Relation of body fat distribution in men and degree of coronary narrowings in coronary artery disease. Am J Carotiol 1992; 70: 1135 ± 1138. 26 Hartz A, Grubb B, Wild R, Van Nort JJ, Kuhn E, Freedman D, et al. The association of waist hip ratio and angiographically determined coronary artery disease. Int J Obes 1990; 14: 657 ± 665. 27 Clark LT, Karve MM, Rones KT, Chang DeMoranville B, Atluri S, Feldman JG. Obesity, distribution of body fat and coronary artery disease in black women. Am J Cardiol 1994; 73: 895 ± 896. 28 Adams Campbell LL, Peniston RL, Kim KS, Mensah E. Body mass index and coronary artery disease in African-Americans. Obes Res 1995; 3: 215 ± 219. 29 Alexander JK. Obesity and coronary heart disease. In: Alpert MA, Alexander JK (eds). The Heart and Lung in Obesity. New York: Futura Publishing Company, 1998; pp 213 ± 238. 30 Korsan-Bengtsen K, Stenberg J, SjoÈstroÈm L, BjoÈrntorp P, Sullivan L. Blood coagulation, ®brinolysis and platelet function in obese subjects at rest and after maximal exercise. Thrombosis Res 1972; 1: 389 ± 406. 31 Sundell IB, Nilsson TK, Ranby M, Hallmans G, Hellsten G. Fibrinolytic variables are related to age, sex, blood pressure, and body build measurements: a cross-sectional study in Norsjo, Sweden. J Clin Epidemiol 1989; 42: 719 ± 723. 32 Blankenhorn DH, Selzer RH, Crawford DW, Barth JD, Liu CR, Liu CH, et al. Bene®cial effects of colestipol-niacin therapy on the common carotid artery. Two- and four-year reduction of intima-media thickness measured by ultrasound. Circulation 1993; 88: 20 ± 28. 33 Furberg CD, Adams HP, Jr, Applegate WB, Byington RP, Espeland MA, Hartwell T, et al. Effect of lovastatin on early carotid atherosclerosis and cardiovascular events. Asymptomatic Carotid Artery Progression Study (ACAPS) Research Group. Circulation 1994; 90: 1679 ± 1687. 34 Borhani NO, Mercuri M, Borhani PA, Buckalew VM, Canossa Terris M, Carr AA, et al. Final outcome results of the Multicenter Isradipine Diuretic Atherosclerosis Study (MIDAS). A randomized controlled trial. JAMA 1996; 276: 785 ± 791. 35 Reed D, Yano K. Predictors of arteriographically de®ned coronary stenosis in the Honolulu Heart Program. Comparisons of cohort and arteriography series analyses. Am J Epidemiol 1991; 134: 111 ± 122.

955


956

36 Berenson GS, Srinivasan SR, Bao W, Newman WPr, Tracey RE, Wattigney WA. Association between multiple cardiovascular risk factors and atherosclerosis in children and young adults. The Bogalusa Heart Study. N Engl J Med 1998; 338: 1650 ± 1656. 37 Alexander JK, Peterson KL. Cardiovascular effects of weight reduction. Circulation 1972; 45: 310 ± 318.

38 Backman L, Freyschuss U, Hallberg D, Melcher A. Reversibility of cardiovascular changes in extreme obesity. Effects of weight reduction through jejunoileostomy. Acta Med Scand 1979; 205: 367 ± 373. 39 Reisin E, Frohlich ED, Messerli FH, Dreslinski GR, Dunn FG, Jones MM, et al. Cardiovascular changes after weight reduction in obesity hypertension. Ann Int Med 1983; 98: 315 ± 319.