Exercise training and blood lipids in hyperlipidemic and normolipidemic adults: A meta-analysis of randomized, controlled trials

European Journal of Clinical Nutrition (1999) 53, 514±522 ß 1999 Stockton Press. All rights reserved 0954±3007/99 $12.00 http://www.stockton-press.co....
Author: Drusilla Park
1 downloads 3 Views 104KB Size
European Journal of Clinical Nutrition (1999) 53, 514±522 ß 1999 Stockton Press. All rights reserved 0954±3007/99 $12.00 http://www.stockton-press.co.uk/ejcn

Exercise training and blood lipids in hyperlipidemic and normolipidemic adults: A meta-analysis of randomized, controlled trials JA Halbert1*, CA Silagy1, P Finucane2, RT Withers3 and PA Hamdorf4 1

Department of Evidence-Based Care and General Practice, School of Medicine, Flinders University of South Australia, Bedford Park, SA 5042, Australia; 2Department of Rehabilitation and Aged Care, Repatriation General Hospital, Daw Park, SA 5041, Australia; 3 Exercise Physiology Laboratory, School of Education, Flinders University of South Australia, Bedford Park, SA 5042, Australia; and 4 Centre for Physical Activity in Ageing, Hampstead Centre, Royal Adelaide Hospital, Adelaide, SA 5000, Australia

Objective: To determine the effectiveness of exercise training (aerobic and resistance) in modifying blood lipids, and to determine the most effective training programme with regard to duration, intensity and frequency for optimizing the blood lipid pro®le. Design: Trials were identi®ed by a systematic search of Medline, Embase, Science Citation Index (SCI), published reviews and the references of relevant trials. The inclusion criteria were limited to randomized, controlled trials of aerobic and resistance exercise training which were conducted over a minimum of four weeks and involved measurement of one or more of the following: total cholesterol (TC), high density lipoprotein (HDL-C), low density lipoprotein (LDL-C) and triglycerides (TG). Subjects: A total of 31 trials (1833 hyperlipidemic and normolipidemic participants) were included. Results: Aerobic exercise training resulted in small but statistically signi®cant decreases of 0.10 mmol=L (95% CI: 0.02, 0.18), 0.10 (95% CI: 0.02, 0.19), 0.08 mmol=L (95% CI: 0.02, 0.14), for TC, LDL-C, and TG, respectively, with an increase in HDL-C of 0.05 mmol=L (95% CI: 0.02, 0.08). Comparisons between the intensities of the aerobic exercise programmes produced inconsistent results; but more frequent exercise did not appear to result in greater improvements to the lipid pro®le than exercise three times per week. The evidence for the effect of resistance exercise training was inconclusive. Conclusions: Caution is required when drawing ®rm conclusions from this study given the signi®cant heterogeneity with comparisons. However, the results appear to indicate that aerobic exercise training produced small but favourable modi®cations to blood lipids in previously sedentary adults. Sponsorship: Public Health Research and Development Project Grant, National Health and Medical Research Council, Department of Health, Housing, Local Government and Community Services, 1995. Descriptors: exercise; cholesterol; aerobic-training; resistance-training

Introduction There is good epidemiological evidence that physical inactivity is an independent risk factor for coronary heart

*Correspondence: Ms. Julie Halbert (MSc), Department of EvidenceBased Care and General Practice, Flinders University of South Australia, Bedford Park, SA 5042, Australia. E-mail: julie.halbert@¯inders.edu.au Guarantor: Ms Julie Halbert Contributorship: Ms Julie Halbert: protocol for meta-analysis, collection of publications for inclusion, data extraction, draft and corrections to manuscript, guarantor for content. Professor Chris Silagy: protocol for meta-analysis, supervision of Ms Julie Halbert, interpretation of results, development of Discussion section, manuscript editing. Professor Paul Finucane: protocol for meta-analysis, supervision of Ms Julie Halbert, interpretation of results, development of Discussion section, manuscript editing. Dr Robert Withers: protocol for meta-analysis, supervision of Ms Julie Halbert, interpretation of results, manuscript editing. Dr Phil Hamdorf: protocol for meta-analysis, manuscript editing. Received: 25 October 1998; revised 1 January 1999; Accepted 25 January 1999

disease (CHD) (Paffenbarger et al, 1993; Powell et al, 1987). This association persists after adjustment for potential confounders such as gender, age, blood pressure and smoking status. However, there are no randomized controlled trials examining the effect of physical activity on morbidity and mortality from CHD. This is largely due to the methodological dif®culties associated with such trials. An interim approach is to hypothesize that physical activity might lead to a reduction in CHD mortality and morbidity by reducing known CHD risk factors such as hyperlipidemia, hypertension and obesity. Observational studies comparing active and inactive populations provide evidence that physical activity favourably modi®es blood lipids (Haskell, 1986). In several studies, trained endurance athletes had lower concentrations of total cholesterol (TC), low-density lipoprotein (LDL-C) and triglycerides (TG), and higher concentrations of high density lipoprotein (HDL-C) than untrained individuals. However, it is uncertain whether this effect persists after removal of confounders such as body composition and diet (Superko, 1991).

No. of subjects

45

34

40

97

88

53

37

78

42

36

20

90

22

113

37

300

16

45

22

Trial (First author, year)

Aellen 1993

Baker 1986

Binder 1996

Blumenthal 1991

Boyden 1993

Duncan 1991

Grandjean 1996

Hellenius 1993

Hersey 1994

Hinkleman 1993

Houmard 1994

Huttunen 1979

Johnson 1983

Juneau 1987

Kiens 1980

King 1991

Leon 1996

Lindheim 1994

Manning 1991

Males, aged 35 ± 60 y, hyperlipidemic Males and females, aged 70 ± 79 y, hyperlipidemic Females, mean age 34 y, normolipidemic Males, aged 40 ± 65 y, normolipidemic Males, aged 40 ± 45 y, hyperlipidemic Males, aged 24 ± 69 y, normolipidemic Males and females, mean age 48 y, hyperlipidemic Males only, aged 30 ± 44 y, normolipidemic Males and females, aged 50 ± 65 y, normolipidemic Males, aged 22 ± 44 y, normolipidemic Females, mean age 49 y, hyperlipidemic Females, aged 22 ± 57 y, obese but weight stable, normolipidemic

Females, normolipidemic

Males, mean age 24 y, normolipidemic Males, aged > 50 y, hyperlipidemic Females, aged 60 ± 72 y, normolipidemic Males and females, aged 60 ± 83 y, hyperlipidemic Females, aged 28 ± 39 y, normolipidemic Females, aged 20 ± 40 y, normolipidemic

Characteristics of subjects

Table 1 Characteristics of included exercise trials

Walking and jogging programme, 58% VO2 max, 50 min, 5 per week Leisure-time physical conditioning, 80 ± 85% VO2 max, 45 min 36per week Walking and jogging programme, 2 intensities; (1) 43% and (2) 76% VO2 max, 40 min, 36per week Brisk walking on treadmill and stair climbing, 40% VO2 max, 45 min, 56per week Walking and cycling programme, 52% VO2 max, 30 min, 3 6per week Resistance training programme, 36per week

Treadmill and walking programme, 68% VO2 max, 40=45 min, 46per week Walking and jogging programme, 64% VO2 max, 30 min, 3 6per week Resistance training programme, 45 ± 60 min, 36per week

Walking, jogging, cycling at 70 ± 80% VO2 max, 40 min, 3 6per week Walking and jogging programme, 50% VO2 max, 30 ± 40 min, 2 ± 36per week 2 groups; (1) walking programme at > 70% VO2 max, 45 min, 36per week (2) resistance programme 36per week Walking programme, 56% VO2 max, 45 min, 56per week

Walking and jogging programme, 3 intensities; (1) 30%, (2) 45% and (3) 76% VO2 max, 60, 45, 36 min, 56per week

Cycle ergometer exercise, 2 intensities; (1) 70% and (2) 80% VO2 max, 30 min, 46per week Walking and running programme, 80% VO2 max, 48 min, 3 6per week Walking, jogging and=or stair climbing programme, 60 ± 70% VO2 max, 30 min, 36per week Cycle and arm ergometer exercise, jogging programme, 84% VO2 max, 55 min, 36per week Resistance training programme, 60 min, 36per week

Description of exercise programme (type, intensity, duration and number of sessions per week)

±

90

225

120

70% VO2 max Training frequency 3 sessions per week > 3 sessions per week Blood lipid status hyperlipidemic normolipidemic

Training group n

Control group n

Change in blood lipid1 (mmol=L) (95%CI)

566 416

495 347

7 0.06 ( 7 0.10, 7 0.02) 7 0.04 ( 7 0.09, 0.00)

515 467

479 363

7 0.06 ( 7 0.09, 7 0.02) 7 0.05 ( 7 0.10, 0.00)

502 480

413 429

7 0.07 ( 7 0.10, 7 0.03) 7 0.04 ( 7 0.10, 0.02)

WMD2 (95%CI Random)

1

Change expressed as mean change in training group 7 mean change in control group, when change is baseline 7 ®nal lipid values. Weighted mean difference (95% CI Random).

2

Table 5 Effect of aerobic exercise training on LDL-C

Training intensity < 70% VO2 max > 70% VO2 max Training frequency 3 sessions per week > 3 sessions per week Blood lipid status hyperlipidemic normolipidemic

Training group n

Control group n

Change in blood lipid1 (mmol=L) (95%Cl)

479 364

427 315

0.07 ( 7 0.04, 0.19) 0.16 (0.06, 0.25)

491 352

466 276

0.16 (0.03, 0.29) 0.06 ( 7 0.07, 0.18)

422 421

362 380

0.05( 7 0.07, 0.16) 0.09 (0.00, 0.18)

WMD2 (95% Cl Random)

1

Change expressed as mean change in training group 7 mean change in control group, when change is baseline 7 ®nal lipid values. Weighted mean difference (95% CI Random).

2

trial is included in Table 1. It should be noted that in many cases, the score may re¯ect the quality of reporting rather than the methodological quality of the trials as information was often lacking about the randomization methods. Data analysis The effect of exercise on TC, TG, HDL-C and LDL-C was assessed independently. This effect was measured as the difference (mmol=L) between the mean change in blood lipid (baseline 7 ®nal value) in the training vs control groups. The variance of this difference is best calculated using the paired baseline and ®nal blood lipid measurements for each individual. However, none of the trials

presented suf®cient information to allow this. Consequently, we adopted a conservative approach and calculated the variance of the difference between means which required us to assume that the baseline and ®nal blood lipids measurements were unpaired. The pooled effect size is a weighted average of the individual effects, with the weightings being inversely proportional to the variance of each individual effect (Bracken, 1992). Results are reported as mean  s.d. Ninety-®ve percent con®dence intervals (CI) were calculated for the pooled effect size using random effect models. Tests of heterogeneity were performed using the MantelHaenszel method (Cochran, 1954).

The ef®cacy of exercise training in the modi®cation of blood lipids JA Halbert et al

519

Table 6 Effect of aerobic exercise training on TG

Training intensity < 70% VO2 max > 70% VO2 max Training frequency 3 sessions per week > 3 sessions per week Blood lipid status hyperlipidemic normolipidemic

1 2

Change in blood lipid1 (mmol=L) (95%Cl)

Training group n

Control group n

478 429

430 362

0.10 (0.01, 0.19) 0.07 (0.00, 0.13)

561 346

518 274

0.13 (0.06, 0.20) 0.03 ( 7 0.07, 0.13)

476 431

399 393

0.15 (0.07, 0.23) 0.03 ( 7 0.04, 0.11)

WMD2 (95% Cl Random)

Change expressed as mean change in training group 7 mean change in control group, when change in baseline 7 ®nal lipid values. Weighted mean difference (95% CI Random).

Results Descriptive data of included trials Forty-one trials met the inclusion criteria for the metaanalysis. After closer examination, 10 were excluded (see Appendix 1) due to insuf®cient data (either pre-and posttraining results or standard deviations=standard errors were not provided). Of the 31 trials (involving 1833 participants) which were included (Table 1), one trial (Leon et al, 1996) used a cross-over design while the remaining studies used a parallel design. Twenty-seven studies used aerobic training exclusively, three studies used resistance training (Boyden et al, 1993; Johnson et al, 1983; Manning et al, 1991) and one study (Hersey et al, 1994) employed both resistance and aerobic training groups (Table 1). Five trials included participants of both sexes, ten trials contained only females and sixteen contained only males. Of the studies that included both males and females, only two trials (Juneau 1987; King 1991) analysed the results according to gender. The mean number of participants per study was 59  52.4 (range 16 ± 300) and the ages of the participants ranged from 19 ± 83 y, with older participants ( > 60 y) represented in only six studies (Binder 1996; Blumenthal et al, 1991; Hersey et al, 1994; Motoyama et al, 1995; Nieman et al, 1993; Ready et al, 1995). The majority of the aerobic exercise programmes (21 trials) included walking, running and=or jogging. Three trials used cycle ergometer exercise (Aellen et al, 1993; Ready & Quinney, 1982; Stein et al, 1990), while three studies used a combination of training modes (Blumenthal et al, 1991; Grandjean et al, 1996; Kiens et al, 1980). One trial (Hersey et al 1994) contained both aerobic (walking) and resistance training groups. The average intensity of the aerobic programmes was 62.9  14.0% VO2 max (range 30 to 84%). The mean length of the exercise (aerobic and resistance) training programmes was 25.7 weeks (range 9 to 52 week) and frequency of training sessions averaged 3.9  1.1 sessions per week (range 2.5 to 7.0). There was a post-training decrease in body mass and body mass index of 0.95 kg (95% CI: 0.49, 1.40) and 0.62 (95% CI: 0.31, 0.94), respectively, in the exercise as compared with the control group. These changes in body weight were not signi®cantly correlated with the changes in TC (r ˆ 7 0.29), HDL-C (r ˆ 7 0.05), LDL-C (r ˆ 7 0.17) and TG (r ˆ 7 0.14).

Effect of aerobic training programmes Aerobic exercise programmes (Table 2) resulted in favourable modi®cations to blood lipids with statistically signi®cant (p < 0.05) changes to TC, TG and HDL-C. Tables 3, 4, 5 and 6 compare the features of the aerobic exercise programmes with regard to exercise intensity, frequency and lipid status of the participants. Overall, the comparison of exercise intensities showed that programmes at intensities greater than 70% VO2 max produced larger changes in TC and LDL-C while programmes at lower intensities modi®ed TG and HDL-C. Correlations between baseline lipid concentrations and the post-training changes were low; TC (r ˆ 0.10). TG (r ˆ 0.23), HDL-C (r ˆ 0.19) and LDL-C (r ˆ 0.17). Three exercise sessions per week resulted in greater modi®cations to all blood lipids than more frequent exercise. Signi®cant heterogeneity was found with all comparisons. The effect of exercise on blood lipids differed, with similar patterns of change occurring between HDL-C and triglycerides and between TC and LDL-C. The largest changes, as a percentage of baseline concentration, were for HDL-C (4% increase) and TG (6% decrease). There were no statistically signi®cant correlations between the changes in TC, HDL-C, LDL-C and TG and energy expenditure (intensity6duration of exercise sessions), energy expenditure per week (intensity6duration6frequency of exercise sessions) and total volume (duration6 frequency of exercise sessions). In addition, there did not appear to be a relationship between the baseline lipid pro®le and the characteristics of the prescribed exercise in terms of energy expenditure and total volume. Effect of resistance training programmes Resistance training programmes (Table 2) resulted in a statistically signi®cant decrease in LDL-C with no differences for TC, TG and HDL-C. Quality of included studies Only four studies provided any information about the method of random assignment (Duncan et al, 1991; King et al, 1991; Wood et al, 1983; Wood et al, 1988). Eleven studies included all randomized participants in the ®nal analysis with the remainder reporting results for only those participants who completed the follow-up. Twelve studies stated an adherence rate for session attendance (Baker et al,

The ef®cacy of exercise training in the modi®cation of blood lipids JA Halbert et al

520

1986; Blumenthal et al, 1991; Boyden et al, 1993; Duncan et al, 1991; Grandjean et al, 1996; Hersey et al, 1994; Houmard et al, 1994; Kiens et al, 1980; King et al, 1991; Leon et al, 1996; Ready et al, 1995; Stein et al, 1990). Four trials failed to provide information on the dietary instructions or the measurement of dietary intake (Johnson et al, 1983; King et al, 1991; Suter et al, 1994; Toriola, 1984). In 19 trials, dietary intake was approximated via the completion of a 24-h (Lindheim et al, 1994), 48-h (Grandjean et al, 1996), 3-d (Baker et al, 1986; Duncan et al, 1991; Hersey et al, 1994; Houmard et al, 1994; Leon et al, 1996; Manning et al, 1991; Ready & Quinney, 1982; Ready et al, 1995; Wood et al, 1983) or 7-d (Binder et al, 1996; Hellenius 1993; Hinkleman & Nieman, 1993; Kiens et al, 1980; Nieman et al, 1993; Stensel et al, 1993; Williams et al, 1994; Wood et al, 1988) diet record at the beginning and the end of the training period. Two trials (Santiago 1995; Stein 1990) reported that dietary intake was monitored but did not report for what period. Discussion Previous work by Superko (1991) demonstrated that aerobic exercise training resulted in increases to HDL-C, decreases in TG, and variable changes in LDL-C. A meta-analysis (Tran et al, 1983) of 66 controlled trials reported an overall decrease in LDL-C (0.13 mmol=L), TC (0.26 mmol=L) and TG (0.18 mmol=L), with an increase in HDL-C (0.03 mmol=L). In the current review, similar posttraining changes were seen with HDL-C (0.05 mmol=L) and LDL-C ( 7 0.10 mmol=L), with smaller decreases in TG ( 7 0.08 mmol=L) and TC (0.10 mmol=L). The reason for the discrepancy between the results of the metaanalyses probably re¯ects the inclusion of trials without random allocation of participants in the previous review by Tran et al, (1983). Recent methodological research has shown that the inclusion of non-randomized studies produces larger estimates of effect than are found in randomized trials (Schulz 1995). While the increase in HDL-C in the current review was small, it is worth examining the clinical signi®cance of this change. Previous work which has examined the relationship between HDL-C and CHD demonstrated that a 1 mg=dl (0.026 mmol=L) increase in HDL-C was associated with a signi®cant CHD risk decrement of 2% in men and 3% in women (Gordon 1989). Thus, the current HDL-C increase represents a 3.8% and 5.7% decrease in CHD risk for men and women, respectively. Training studies of the effect of resistive exercise programmes on blood lipid pro®les have reported mixed results. One review (Hurley, 1989) commented that while a number of trials had demonstrated improvements to lipid concentrations after resistive training, these studies were often ¯awed by poor methodological quality and failed to take into account confounders such as changes in body composition during the training period. In the current review, we were disappointed by the low number of trials of resistance exercise that met the inclusion criteria. The reduction in LDL-C after resistance training seen in the current meta-analysis should be interpreted with caution as this result was based on only three trials, involving a total of 131 participants, and requires con®rmation with a larger study. The descriptive reviews (Haskell, 1986; Superko, 1991; Goldberg & Elliot, 1985; Hurley, 1989) have commented

on the dif®culty in separating the effects of the exercise training programme from changes in body composition, dietary patterns, smoking status, alcohol consumption, use of exogenous hormones and menopausal status. It is possible that changes in body weight due to exercise training are less important than changes in body composition as decreases in percentage body fat result in increases in HDL-C (Superko, 1991). The majority of the included trials of this meta-analysis failed to measure body composition changes during the training period so we were unable to address this issue. To determine the best way to modify blood lipids, different training schedules need to be compared. The current meta-analysis found no clear relationship between the intensity of exercise and the change in blood lipids; however, a comparison of the frequency of exercise indicated that three exercise sessions per week were more effective in modifying blood lipids than more frequent activity. The inability of this review to identify signi®cant differences between exercise programmes probably re¯ects the relative homogeneity of the exercise programmes included. The majority of the trials compared aerobic exercise training with sedentary control groups and only seven trials directly compared exercise programmes of differing intensities and frequencies. In addition, comparisons between programmes were limited by the absence of exercise at lower intensities (less than 60% VO2 max). The literature frequently reports the presence of a `threshold' of exercise required to signi®cantly change blood lipids (Haskell, 1986; Superko, 1991). It is possible that it is not the intensity, frequency or duration of the exercise that elicits a change, but rather the volume (frequency6duration6intensity) of activity. While one reviewer reported that increasing the number of hours of training resulted in small increases in HDL-C (Tran et al, 1983), another stated that increases in HDL-C and decreases in TG are only obtained after 10 miles of jogging per week for 6 months (Superko, 1991). In the current review, there was no relationship between the changes in lipid concentrations over the training period and the energy expenditure and total volume of the exercise programmes included. It is also possible that the `threshold' is not dependent on the exercise programme but is related to the baseline lipid concentration. Previous reviews (Haskell, 1986; Tran et al 1983) have reported that the higher the lipid concentration prior to exercise training, the greater the reduction (or increase in the case of HDL-C) in the posttraining concentration. In contrast, the current meta-analysis found no relationship between the baseline lipid concentrations and the degree of change with exercise. In addition, baseline lipid concentrations were not related to the characteristics of the exercise programme, thereby discounting the possibility that the individuals with the less favourable baseline lipid pro®les were prescribed exercise training at a lower intensity and volume. The limitations of this review need to be considered when interpreting the ®ndings. As with other reviews, we found the quality of the published trials to be poor. The major defects were small sample sizes, failure to state the randomization methods and lack of information on how selection bias was controlled after entry. In addition, methodological differences in the measurement of blood lipids, laboratory and testing protocols resulted in trials of varying quality. This meta-analysis may also provide its own biases. In eight trials (Aellen et al, 1993; Duncan et al,

The ef®cacy of exercise training in the modi®cation of blood lipids JA Halbert et al

1991; Juneau et al, 1987; King et al, 1991; Stein et al, 1990; Suter et al, 1994; Toriola, 1984; Williams et al 1994) two or more intervention groups were directly compared with one control group. In these trials, the results of the control group were over represented in the analysis. One of the major limitations of meta-analyses based on smaller trials is the possibility of publication bias (Sacks et al, 1987). Funnel plots (not shown) did not indicate that publication bias may be a problem since the plots appeared to show an equal number of trials with no change, an increase, and decrease in blood lipids. Signi®cant heterogeneity was found between trials included in all of the comparisons. This ®nding was not surprising given the extremely diverse differences between the changes in blood lipids of individual trials. Possible contributors to the heterogeneity include: the variability in the age of the included subjects, differences in pre-training blood lipid concentrations, small sample sizes and varying exercise programmes, but there was inadequate data available from the individual trials to explore this further. In conclusion, we found that while aerobic exercise training did appear to alter blood lipids, the clinical importance of these changes is questionable. The bene®ts of exercise while small at an individual level, maybe of signi®cant public health bene®t if they can be applied broadly throughout the community. Such bene®ts need con®rmation in large prospective controlled trials. Until these are forthcoming, there is insuf®cient evidence to recommend exercising more than three times per week if the speci®c intention is to improve the blood lipid pro®le. References Aellen R, Hollmann W & Boutellier U (1993). Effects of aerobic and anaerobic training on plasma lipoproteins. Int. J. Sports Med. 14: 396 ± 400. American College of Sports Medicine (1990). The recommended quantity and quality of exercise for developing and maintaining cardiorespiratory and muscular ®tness in healthy adults. Position Stand. Med. Sci. Sports Exerc. 22, 265 ± 274. Baker TT, Allen D, Lei KY & Wilcox KK (1986). Alterations in lipid and protein pro®les of plasma lipoproteins in middle-aged men consequent to an aerobic exercise program. Metabolism 35, 1037 ± 1043. Binder EF, Birge SJ & Kohrt WM (1996). Effects of endurance exercise and hormone replacement therapy on serum lipids in older women. J. Am. Geriatr. Soc. 44, 231 ± 236. Blumenthal JA, Emery CF, Madden DJ, Coleman RE, Riddle MW, Schniebolk S, Cobb FR, Sullivan MJ & Higginbotham MB (1991). Effects of exercise training on cardiorespiratory function in men and women > 60 years of age. Am. J. Cardiol. 67, 633 ± 639. Boyden TW, Pamenter RW, Going SB, Lohman TG, Hall MC, Houtkooper LB, Bunt JC, Ritenbaugh C & Aickin M (1993). Resistance exercise training is associated with decreases in serum low-density lipoprotein cholesterol levels in premenopausal women. Arch. Intern. Med. 153, 97 ± 100. Bracken MB (1992). Statistical methods for analysis of effects of treatment in overviews of randomized trials. In: Effective Care of the Newborn Infant. JC Sinclair & MB Bracken, eds. Oxford University Press: Oxford, pp.13 ± 18. Cochran WG (1954). The combination of estimates from different experiments. Biometrics 10, 101 ± 129. Cochrane Collaboration (1995). Cochrane Collaboration Handbook, DL Sackett & AD Oxman eds. The Cochrane Collaboration Oxford, (updated July 1995). Duncan JJ, Gordon NF & Scott CB (1991). Women walking for health and ®tness: how much is enough? JAMA 266, 3295 ± 3299. Goldberg L & Elliot DL (1985). The effect of physical activity on lipid and lipoprotein levels. Med. Clin. of North America 69, 41 ± 53. Gordon DJ, Probst®eld JL, Garrison RJ, Neaton JD, Castelli WP, Knoke JD, Jacobs DR, Bangdiwala S & Tyroler HA (1989). High-density lipoprotein cholesterol and cardiovascular disease. Circulation 79, 8 ± 15.

Grandjean PW, Oden GL, Crouse SF, Brown JA & Green JS (1996). Lipid and lipoprotein changes in women following 6 months of exercise training in a worksite ®tness program. J. Sports Med. Phys. Fitness 36, 54 ± 59. Haskell WL (1986). The in¯uence of exercise training on plasma lipids and lipoproteins in health and disease. Acta. Med. Scan. 711: (Suppl 711) S25 ± S37. Hellenius M-L, de Faire U, Berglund B, Hamsten A & Krakau I (1993). Diet and exercise are equally effective in reducing risk for cardiovascular disease. Results of a randomized controlled trial in men with slightly to moderately raised cardiovascular risk factors. Atherosclerosis 103, 81 ± 91. Hersey WC, Graves JE, Pollock ML, Gingerich R, Shireman RB, Heath GW, Spierto F, McCole SD & Hagberg JM (1994). Endurance exercise training improves body composition and plasma insulin responses in 70- to 79-year-old men and women. Metabolism 43, 847 ± 854. Hinkleman LL & Nieman DC (1993). The effects of a walking program on body composition and serum lipids and lipoproteins in overweight women. J. Sports Med. Phys. Fitness 33, 49 ± 58. Houmard JA, McCulley C, Roy LK, Bruner RK, McCammon MR & Israel RG (1994). Effects of exercise training on absolute and relative measurements of regional adiposity. Int. J. Obesity 18, 243 ± 248. Hurley BF (1989). Effects of resistive training on lipoprotein-lipid pro®les a comparison to aerobic exercise training. Med. Sci. Sports Exerc. 21, 689 ± 693. Huttunen JK, Lansimies E, Voutilainen E, Ehnholm C, Hietanen E, Penttila I, Siitonen O & Rauramaa R (1979). Effect of moderate physical exercise on serum lipoproteins. A controlled clinical trial with special reference to serum high-density lipoprotein. Circulation 60, 1220 ± 1229. Johnson CC, Stone MH, Byrd RJ & Lopez-s A (1983). The response of serum lipids and plasma androgens to weight training exercise in sedentary males. J. Sports Med. 23, 39 ± 44. Juneau M, Rogers F, De Santos V, Yee M, Evans A, Bohn A, Haskell WL, Taylor CB & DeBusk RF (1987). Effectiveness of self-monitored, home-based, moderate-intensity exercise training in middle-aged men and women. Am. J. Cardiol. 60, 66 ± 70. Kiens B, Jorgensen I, Lewis S, Jensen G, Lithell H, Vessby B, Hoe S & Schnohr P (1980). Increased plasma HDL-cholesterol and apo A-1 in sedentary middle-aged men after physical conditioning. Eur. J. Clin. Invest. 10, 203 ± 209. King AC, Haskell WL, Taylor CB, Kraemer HC & De Busk RF (1991). Group- vs home-based exercise training in healthy older men and women. JAMA 266, 1535 ± 1542. Leon AS, Casal D & Jacobs Jr D (1996). Effects of 2,000 kcal per week of walking and stair climbing on physical ®tness and risk factors for coronary heart disease. J. Cardiopulmonary Rehabil. 16, 183 ± 192. Lindheim SR, Notelovitz M, Feldman EB, Larsen S, Khan FY & Lobo RA (1994). The independent effects of exercise and estrogen on lipids and lipoproteins in postmenopausal women. Obstet. Gynecol. 83, 167 ± 172. Manning JM, Dooly-Manning CR, White K, Kampa I, Silas S, Kesselhaut M & Ruoff M (1991). Effects of a resistive training program on lipoprotein ± lipid levels in obese women. Med. Sci. Sports Exerc. 23, 1222 ± 1226. Motoyama M, Sunami Y, Kinoshita F, Irie T, Sasaki J, Arakawa K, Kiyonaga A, Tanaka H & Shindo M (1995). The effects of long-term low intensity aerobic training and detraining on serum lipid and lipoprotein concentrations in elderly men and women. Eur. J. Appl. Physiol. 70, 126 ± 131. Nieman DC, Warren BJ, O'Donnell KA, Dotson RG, Butterworth DE & Henson DA (1993). Physical activity and serum lipids in elderly women. J. Am Geriatr. Soc. 41, 1339 ± 1344. Paffenbarger RS, Hyde RT, Wing AL, Lee I-M, Jung DL & Kampert JB (1993). The association of changes in physical-activity level and other lifestyle characteristics with mortality among men. N. Engl. J. Med. 328, 538 ± 545. Powell KE, Thompson PD, Caspersen CJ & Kendrick JS (1987). Physical activity and the incidence of coronary heart disease. Ann. Rev. Public Health 8, 253 ± 287. Ready AE & Quinney HA (1982). The response of serum lipids and lipoproteins to high intensity endurance training. Can. J. Appl. Spt. Sci. 7, 202 ± 208. Ready AE, Drinkwater DT, Ducas J, Fitzgerald DW, Brereton DG & Oades SC (1995). Walking program reduces elevated cholesterol in women postmenopause. Can. J. Cardiol. 11, 905 ± 912. Sacks HS, Berrier J, Reitman D, Ancona-Berk VA & Chalmers TC (1987). Meta-analyses of randomized controlled trials. N. Engl. J. Med. 316, 450 ± 455.

521

The ef®cacy of exercise training in the modi®cation of blood lipids JA Halbert et al

522

Santiago MC, Leon AS & Serfass RC (1995). Failure of 40 weeks of brisk walking to alter blood lipids in normolipidemic women. Can. J. Appl. Physiol. 20, 417 ± 428. Schulz KF, Chalmers I, Hayes RJ & Altman DG (1995). Empirical evidence of bias. Dimensions of methodological quality associated with estimates of treatment effects in controlled trials. JAMA 273, 408 ± 412. Stein RA, Michielli DW, Glantz MD, Sardy H, Cohen A, Goldberg N & Brown CD (1990). Effects of different exercise training intensities on lipoprotein cholesterol fractions in healthy middle-aged men. Am. Heart J. 119, 277 ± 283. Stensel DJ, Hardman AE, Brooke-Wavell K, Vallance D, Jones PRM, Norgan NG & Winder AF (1993). Brisk walking and serum lipoprotein variables in formerly sedentary men aged 42 ± 59 years. Clin. Sci. 85, 701 ± 708. Superko HR (1991). Exercise training, serum lipids, and lipoprotein particles is there a change threshold? Med. Sci. Sports Exerc. 23, 677 ± 685. Suter E, Marti B & Gutzwiller F (1994). Jogging or walking Ð comparison of health effects. Ann. Epidemiol. 4, 375 ± 381.

Toriola AL (1984). In¯uence of 12-week jogging on body fat and serum lipids. Br. J. Sports Med. 18, 13 ± 17. Tran ZV, Weltman A, Glass GV & Mood DP (1983). The effects of exercise on blood lipids and lipoproteins a meta-analysis of studies. Med. Sci. Sports Exerc. 15, 393 ± 402. Williams PT, Stefanick ML, Vranizan KM & Wood PD (1994). The effects of weight loss by exercise or by dieting on plasma high-density lipoprotein (HDL) levels in men with low, intermediate and normal-tohigh HDL at baseline. Metabolism 43, 917 ± 924. Wood PD, Haskell WL, Blair SN, Williams PT, Krauss RM, Lindgren FT, Albers JJ & Farquhar JW. (1983). Increased exercise level and plasma lipoprotein concentrations a one-year, randomized, controlled study in sedentary, middle-aged men. Metabolism 32, 31 ± 39. Wood PD, Stefanick ML, Dreon DM, Frey-Hewitt B, Garay SC, Williams PT, Superko HR, Fortmann SP, Albers JJ, Vranizan KM, Ellsworth NM, Terry RB & Haskell WL (1988). Changes in plasma lipids and lipoproteins in overweight men during weight loss through dieting as compared with exercise. New. Engl. J. Med. 319, 1173 ± 1179.

Suggest Documents