SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Obejctive: Weight loss and physical activity have shown favorable effects on risks associated with obesity. It is therefore of interest to evaluate exercise capacity and related co-morbidities in obese patients.

Design and Methods: We present data from obese subjects evaluated by the 6-minute walk test (6MWT) before and after a 7.3 (6.1-8.2) month weight reduction program.

Results: 251 subjects completed the test at baseline (BMI 40.6 [36.9-44.6] kg/m2) and 129 (51.4%) repeated the test after intervention (BMI 35.6 [31.2-38.5] kg/m2). The six minute walking distance (6MWD) at baseline (535 [480-580] m) and at follow up (599 [522-640] m) correlated to several cardiovascular risk markers. Age, weight, height, resting heart rate, smoking status, fP-glucose and use of β-blockers explained 43 % of the variance in predicted 6MWD at baseline. The effect of smoking status, fP-glucose, β-blockers, and resting heart rate lost significance at follow up. Presence of diabetes and the metabolic syndrome had a negative influence on 6MWD but did not affect the impact of intervention based on percentage increase in walking distance. Gender had no impact on 6MWD. Reported pain during the test was common but decreased after intervention (57.0% vs. 28.7%, P < 0.001).

Conclusions:The 6MWT may be used to evaluate intervention success beyond kilogram weight loss in obese subjects. We present formulas to predict 6MWD and the effect of weight loss on walking distance in clinical practice. Pain is a common problem which has to be considered when giving advice on exercise as a part of weight loss intervention.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The prevalence of obesity has increased rapidly during the last two decades (1, 2). Obesity is associated with significant morbidity, premature mortality, impaired quality of life, and rising healthcare costs (3–5). Both weight loss and increased physical activity have been shown to attenuate the risks associated with obesity; however, the relative impact is under debate. Observational studies have demonstrated an independent effect of weight loss on the risk for premature death for both obesity and physical inactivity (6-8). Other indications suggest that no excess mortality can be found in physically fit obese subjects (9). Impaired walking capacity is common in obese subjects not only because it is a weight-bearing activity. Obese subjects also have a lower level of exercise capacity and significant comorbidities such as musculoskeletal pain (10-12). There is a need for clinical tests evaluating obesity per se as well as the effect of intervention that go beyond the actual weight loss. The 6-min walk test (6MWT) is a simple, safe, and inexpensive test that evaluates physical performance and walking capacity (13-15). Test results are related to day-to-day physical activity and can also be assumed to represent a direct measure of impaired quality of life (16). Walking tests are suitable in patient groups who otherwise have difficulty in performing standard exercise tests. So far, the 6MWT has been used mainly in patients with chronic heart failure and pulmonary diseases to evaluate prognosis and treatment efficacy (16-19). Few studies have been performed in obese subjects (20-22). Studies in obese children, adolescents, and adults show a high degree of reproducibility, indicating that the test is suitable for obese individuals as a group (23, 24). Here, we performed a retrospective analysis of the 6MWT in subjects who conducted the test as part of their clinical evaluation prior to and after completing a standard group-based weight reduction program. Our aim was to identify factors that independently influence walking capacity in addition to weight loss.

Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Subjects

Patients attending the obesity outpatient unit at the Department of Endocrinology, Skåne University Hospital, Malmö, Sweden, between 2003 and 2009 were asked to participate in the current evaluation of physical performance and weight loss by means of a 6MWT before and after attending our standard weight reduction program. In total, 253 subjects (63 men and 190 women) performed the test prior to the intervention program (baseline). Blood samples were collected after an overnight fast and analyzed at the Department of Clinical Chemistry, Skåne University Hospital, Malmö, Sweden. Insulin sensitivity was estimated by the homeostasis model assessment of insulin resistance index (HOMA-IR (25)). Weight (nearest 0.1 kg) and height (nearest 0.5 cm) were measured in light clothing, and body mass index (BMI) was calculated (weight [kg]/height2 [m2]). Waist circumference (nearest 0.5 cm) was measured between the iliac crest and the rib cage in the standing position. Systolic and diastolic blood pressure (SBP and DBP) were measured after a 10-min rest. Complete data, allowing for evaluation of presence of the metabolic syndrome according to the US National Cholesterol Education Program Adult Treatment Panel III (26), were available for 195 subjects (77.7%). Subjects with both diabetes and the metabolic syndrome were classified as “metabolically at risk” subjects, whereas those with neither of the two conditions were classified as “metabolically healthy” subjects. The local ethics committee approved the study, and all subjects gave their written informed consent to participation.

Weight reduction program

The standard intervention program consisted of 14 group counseling sessions over the course of 6 months, and in most cases (n = 161) preceded by 12 weeks of low-calorie diet (LCD) based on milk protein (800-880 kcal/day [protein 25-30 E%, carbohydrates 49-55 E%, and fat 20-21 E%]). Subjects with a history of a recent myocardial infarction (within 6 months), malignancy, type 1 diabetes mellitus, or decompensated heart failure were excluded from LCD treatment and recommended alternative dietary strategies. Subjects with hypertension and type 2 diabetes were included and monitored during the LCD period. Each group typically consisted of 10 subjects, either male or female. All subjects met with a physiotherapist before the intervention and were encouraged to lead a generally healthy lifestyle including increased physical activity; however, the weight reduction treatment did not involve a specific training program.

Six-minute walk test

Subjects were instructed to walk at their own maximum walking speed in a 60-m-long hospital corridor with indicated turning points. Walking distance after 6 min (6-min walking distance, 6MWD) was measured. Heart rate (HR) was measured by a Polar HR monitor after 5 min of rest before the test and directly after the test. The degree of effort and the degree of activity-associated pain were recorded using the Borg CR10 scale (graded 1-10 (27)). Localization of reported pain was recorded. All tests were conducted by the same physiotherapist.

Statistics

Data are shown as mean ± SEM. Clinical characteristics are described either as median with interquartile range (25th-75th percentile) for numeric variables or as percentage (n [%]) for categorical variables. Data for HR and blood pressure include subjects prescribed β-blockers. For statistical analyses including HR or blood pressure, subjects prescribed β-blockers were excluded, if not otherwise stated, as this could be expected to influence the result. Subjects with diagnosed diabetes were not included in the presentation or analysis of fasting plasma insulin and HOMA-IR. The differences between the groups were tested using Wilcoxon's signed rank test for paired comparisons and Mann-Whitney U test for unpaired comparisons. Correlations between variables were tested using Spearman's test. Multiple regression analysis was used to identify independent variables explaining the variance in 6MWD and increase in 6MWD after weight loss, specifically to identify factors that contribute to increase in 6MWD independent of weight loss, which in itself was expected to do so. Categorical variables were analyzed with 1 for positive and 0 for negative and 0 for male and 1 for female gender. Diabetes diagnosis and fP-glucose were not added into the same analysis as they are strongly dependent of each other. The χ test was used when comparing frequencies. A P-value below 0.05 was considered statistically significant. All statistics were calculated using Number Cruncher Statistical System 2004 software (NCSS, Kaysville, UT).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Baseline

The baseline characteristics of the study subjects are shown in Table 1. In total, 253 subjects performed the 6MWT as part of their participation in the weight reduction program. About 99.2% of the subjects completed the baseline test. Two subjects terminated the test prematurely because of physical complaints (severe back and/or hip pain) and were excluded from further analyses (Figure 1). The median 6MWD was 535 (480–580) m, ranging between 290 and 688 m. This is equivalent to a walking pace of 1.49 m/s (range, 0.81-1.91 m/s). There was no significant difference between men and women (529 [480-567] vs. 539 [480-585] m, P = 0.480). Smokers and nonsmokers differed in 6MWD (508 [477-545] vs. 540 [483-583] m, P = 0.035). Those who used β-blockers performed worse than those who did not (509 [420-562] vs. 540 [485-584] m, P = 0.026).

thumbnail image

Figure 1. Flow diagram of the study. Two subjects did not complete the test at baseline because of severe restricting pain and were therefore excluded from further analysis. 6MWT, 6-min walk test.

Download figure to PowerPoint

Table 1. Clinical characteristicsa
 BaselineFollow-up cohortbPc
BaselineFollow-up
  • Abbreviations: SBP, systolic blood pressure; DBP, diastolic blood pressure; HOMA-IR index, homeostasis model assessment of insulin resistance index (25); HDL, high-density lipoprotein; LDL, low-density lipoprotein.

  • a

    Data are shown as median with interquartile range or as percentage.

  • b

    Follow-up cohort include those who repeated the 6-min walk test at follow-up. Baseline and follow-up data are shown. There were small but significant differences in height, triglycerides in women, and cholesterol at baseline between the follow-up cohort and those who performed the test only at baseline (data not shown).

  • c

    P-value after comparing follow-up cohort at baseline versus at follow-up (Wilcoxon's signed rank test).

  • d

    Data according to the US National Cholesterol Education Program Adult Treatment Panel III (NCEP (26)). Complete data allowing for evaluation were available for 195 subjects (77.7%).

  • e

    Borg CR10 scale (27).

n251129129 
Gender (M/F [%])63 (25.1%)/   
188 (74.9%)36 (27.9%)/   
93 (72.1%)NA   
Age (years)41.9 (33.6-51.6)41.4 (33.5-52.0)42.1 (34.0-52.7)NA
Weight (kg)117 (101-135)119 (105-135)103 (88-117)<0.001
Height (m)1.68 (1.63-1.77)1.69 (1.65-1.78)NA 
BMI (kg/m2)    
 Male42.0 (39.0-47.0)41.9 (39.0-45.1)36.1 (33.3-39.9)<0.001
 Female40.0 (36.4-44.3)40.5 (36.6-44.4)34.8 (30.1-38.1)<0.001
Waist (cm)    
 Male136 (130-145)137 (129-145)118 (110-133)<0.001
 Female119 (111-130)119 (110-130)103 (93-117)<0.001
SBP (mmHg)130 (120-140)130 (120-140)123 (110-131)<0.001
DBP (mmHg)80 (70-82)80 (70-85)70 (60-80)0.020
fP-glucose (mmol/l)5.3 (4.9-6.1)5.2 (4.8-5.8)4.9 (4.5-5.4)<0.001
fP-insulin (mIE/l)12 (8-16)11 (7.3-14.8)6.1 (3.9-11.3)<0.001
HOMA-IR index2.7 (1.8-3.9)2.6 (1.7-3.6)1.3 (0.8-2.4)<0.001
Cholesterol (mmol/l)4.8 (4.2-5.4)4.7 (4.1-5.3)4.5 (3.9-5.2)0.301
HDL (mmol/l)    
 Male0.92 (0.70-1.09)1.01 (0.82-1.13)1.10 (0.89-1.21)0.131
 Female1.07 (0.92-1.26)1.11 (0.89-1.28)1.20 (0.93-1.49)<0.001
LDL (mmol/l)3.2 (2.6-3.7)3.1 (2.5-3.6)3.0 (2.3-3.5)0.182
Triglycerides (mmol/l)    
 Male1.5 (0.9-2.2)1.4 (0.9-2.1)1.2 (0.6-2.1)0.005
 Female1.2 (0.8-1.6)1.1 (0.7-1.5)0.7 (0.5-1.2)<0.001
Diabetes mellitus (n [%])    
 Type 14 (1.6%)2 (1.6%)NA 
 Type 253 (21.5%)24 (18.9%)NA 
Metabolic syndrome (n [%])d98 (50.3%)51 (46.8%)NA 
Smokers (n [%])    
 Current31 (13.6%)16 (13.0%)NA 
 Former59 (25.9%)28 (22.8%)NA 
 Nonsmokers138 (60.5%)79 (64.2%)NA 
Prescribed β-blockers (n [%])28 (11.2%)20 (15.5%)16 (12.4%)0.451
Distance (m)535 (480-580)548 (487-590)599 (522-640)<0.001
Walking pace (m/s)1.49 (1.33-1.61)1.52 (1.35-1.63)1.66 (1.45-1.78)<0.001
Heart rate before (bpm)79 (71-89)78 (70-88)75 (65-82)<0.001
Heart rate after (bpm)130 (119-142)130 (119-145)127 (118-146)0.139
Degree of effort (1-10)e3.0 (2.5-4.0)3.0 (2.0-4.0)2.5 (1.0-3.0)<0.001
Degree of pain (1-10)e2 (0-4)2 (0-3)0 (0-1)<0.001

Table 2 shows variables that were significantly associated with 6MWD in a univariate correlation analysis. There was a significant correlation with age, weight, and BMI, but none with height (r = 0.081, P = 0.201). 6MWD was also correlated to waist circumference, SBP, DBP, and fP-glucose (r = −0.185, P = 0.016 when analyzed without subjects with diabetes) but not to other surrogate markers for insulin resistance such as fP-insulin or HOMA-IR (r = −0.017, P = 0.842 and r = −0.061, P = 0.490, respectively). Multiple regression analysis using variables known before the test (Table 3) yielded the following formula explaining 43% of the variance in 6MWD (r2 = 0.43):

Formula 1:

6MWDprior1 (m) = 1561.80 − 2.08 × (age) − 545.19 × log(weight) − 0.68 × (HR before 6MWT) − 32.67 × (β-blockers) − 2.49 × (gender) + 1370.41 × log(height) − 24.54 × (smoking status) − 89.44 × log(fP-glucose)

Table 2. Variables associated to 6MWD in a univariate correlation analysis
 r
BaselineFollow-up
  • Abbreviations: HR, heart rate; 6MWT, 6-min walk test; SBP, systolic blood pressure; DBP, diastolic blood pressure; LDL, low-density lipoprotein; 6MWD, 6-min walking distance.

  • *

    P < 0.001.

  • **

    0.001 < P < 0.01.

  • ***

    0.01 < P < 0.05.

Weight−0.258*−0.227**
Height0.0810.193***
BMI−0.395*−0.407*
Age−0.401*−0.403*
Waist−0.317*−0.290**
HR before 6MWT−0.126−0.188***
HR after 6MWT0.512*0.627*
ΔHR during 6MWT0.563*0.742*
SBP−0.289*−0.462**
DBP−0.207**−0.177
Degree of pain−0.137***−0.189***
fP-Glucose−0.295*−0.264***
Cholesterol−0.071−0.262***
LDL−0.052−0.260***
6MWD at baselineNA0.883*
Table 3. Stepwise multiple regression analysis with 6MWD as the dependent variable
 BaselineFollow-up
Formula 1Formula 2Formula 3Formula 4
r2 incrementr2 incrementr2 incrementr2 increment
  • Abbreviations: HR, heart rate; 6MWT, 6-min walk test.

  • a

    Knee pain reported during the test.

  • b

    Gender was added for adjustment but did not change the cumulative r2.

  • *

    P < 0.001.

  • **

    0.001 < P < 0.01.

  • ***

    0.01 < P < 0.05.

Age0.081*0.0138**0.128*0.0158**
Weight0.201*0.190*0.155*0.141*
Height0.099*0.104*0.128*0.164*
Diabetes 0.015**  
fP-glucose0.016***   
β-Blockers0.016***   
HR before 6MWT0.012***0.052* 0.092*
HR after 6MWT 0.186* 0.393*
Smoke0.012***   
Knee paina 0.013**  
Genderb<0.0010.002<0.0010.004
Cumulative r20.430.590.340.79

To identify other variables influencing 6MWD, we also performed the analysis with all known variables, that is, including data obtained during the test, which yielded the following formula explaining 59% of the variance in 6MWD (r2 = 0.59):

Formula 2:

6MWDtest1 (m) = 1124.95 − 0.83 × (age) − 505.50 × log(weight) − 1.42 × (HR before 6MWT) + 2.42 × (HR after 6MWT) − 28.68 × (knee pain) − 1.27 × (gender) + 1281.76 × log(height) − 23.29 × (diabetes)

Gender was not a significant variable and did not change the r2 in any of the described analyses but was included for adjustment.

Either presence of diabetes or the metabolic syndrome was associated with a shorter 6MWD (500 [420-540] vs. 547 [489-590] m, P < 0.001 and 516 [489-569] vs. 546 [504-583] m, P = 0.022, respectively). Those diagnosed with both diabetes and the metabolic syndrome (at-risk obese subjects) performed worse than the obese but otherwise healthy subjects (488 [409-529] vs. 545 [502-583] m, P < 0.001; Figure 2). The two groups were comparable in weight (114 [100-134] vs. 115 [101-129] kg, P = 0.970); however, at-risk obese subjects were older than the metabolically healthy obese subjects (48 [39-56] vs. 40 [32-49] years, P = 0.005). Having the metabolic syndrome without the presence of diabetes did not affect 6MWD (P = 0.949). As there were very few subjects with diabetes not fulfilling the criteria for the metabolic syndrome (n = 7 and n = 4 before and after weight loss, respectively), this group was not included in the analysis.

thumbnail image

Figure 2. The influence of the metabolic syndrome (MS) and diabetes mellitus (DM) on 6MWD at baseline (□) and after weight loss (▪) shown as mean ± SEM. ***P < 0.001 (Wilcoxon's signed rank test).

Download figure to PowerPoint

One hundred forty-three (57.0%) of the subjects reported any kind of pain (graded 4 [3-5]) during the test. The frequency of reported pain was the same in men and women (P = 0.2). The four most common localizations were tibia, knee, foot, and lower back (Table 4). There was no difference in 6MWD between those reporting pain and those who did not (538 [475-568] vs. 534 [495-590] m, P = 0.200). Instead, experiencing tibial pain was associated with a greater 6MWD (567 [536-598] vs. 524 [471-567] m, P < 0.001). There was no difference in the degree of reported effort between those who reported pain over tibia and those who did not (data not shown); however, the increase in HR during the test (ΔHR) was significantly greater in those who reported tibial pain (73.5 [55.3-91.0]% vs. 61.0 [43.0-82.0]%; P = 0.012). Subjects who reported hip or knee pain performed worse when compared with those who did not (469 [402-507] vs. 540 [485-581] m, P = 0.003 and 485 [407-528] vs. 540 [486-584] m, P < 0.001, respectively). The degree of reported pain correlated to 6MWD in the univariate analysis (Table 2). This correlation was attenuated in the multiple regression analysis when adjusting for weight or height (r2 = 0.019, P = 0.03 vs. r2 = 0.003, P = 0.320).

Table 4. Activity-related pain as reported during the 6MWTa
 Baseline (n = 251)Follow-up cohortbPc
Baseline (n = 129)Follow-up (n = 129)
  • a

    Data are shown as median with interquartile range or as percentage.

  • b

    Follow-up cohort include those who repeated the 6-min walk test at follow-up. Baseline and follow-up data are shown. There was no difference in the frequency of reported pain, pain localization, or degree of pain between the follow-up cohort and those who performed the test only at baseline.

  • c

    P-value after comparing follow-up cohort at baseline versus at follow-up (χ2 test when comparing frequencies and Mann-Whitney U test when comparing degree of reported pain).

Subjects reporting pain (n [%])143 (57.0%)73 (56.9%)37 (28.7%)<0.001
Degree of pain (1-10)4 (3-5)3 (3-5)3 (2-5)0.190
Localization    
 Tibia52 (20.7%)28 (21.7%)15 (11.6%)0.032
 Knee22 (8.8%)11 (8.5%)7 (5.4%)0.216
 Foot20 (8.0%)8 (6.2%)5 (3.9%)0.401
 Lower back14 (5.6%)9 (7.0%)3 (2.3%)0.078

Follow-up

One hundred seventy subjects (67.2%) completed the weight reduction program (Figure 1). Adherence to the program was lower among those who initiated the treatment with LCD (62.7% vs. 76.7%, P = 0.001). The three most common reasons for abandoning among those who reported a reason were lack of time, lack of resources, and not being able to cope with the diet.

The median weight loss was 12.4 kg (4.2-26.0) ranging from a 12.0 kg gain to a 66.0 kg loss of weight. For those who initiated weight reduction with LCD, the median weight loss was 22.0 kg (13.0-30.0). Weight loss relative to baseline weight was associated with decrease in resting HR (r = 0.390, P < 0.001), decrease in waist circumference (r = 0.807, P < 0.001), and decrease in HOMA-IR (r = 0.395, P = 0.009); all comparisons were made to baseline.

One hundred twenty-nine subjects repeated the test after the program (Figure 1). Baseline data for those who repeated the test (follow-up cohort) and those who did not was compared to identify any factors predictive for noncompliance. We found small but significant differences in triglycerides (P = 0.009), height (P = 0.032), and cholesterol (P = 0.022) in women. The frequency in reported pain at baseline did not differ in those who performed the test at baseline only and those who repeated the test after intervention (57.0% vs. 56.9%, P = 0.840).

6MWD increased significantly after the weight reduction program (42 [16-72] m, P < 0.001, ranging from 46 m decrease to 168 m increase). In a univariate analysis, 6MWD at follow-up was associated with 6MWD at baseline, age, weight, height, and BMI (Table 2). There was also a significant correlation to waist, resting HR, SBP, fP-glucose (r = −0.204, P = 0.085 excluding subjects with diabetes), low-density lipoprotein (LDL), and cholesterol, but none to HOMA-IR or fP-insulin (r = −0.011, P = 0.942; and r = 0.026, P = 0.865, respectively). Multiple regression analysis including variables known before the test (Table 3) yielded the following formula explaining 34% of the variance in 6MWD at follow-up (r2 = 0.34):

Formula 3:

6MWDprior2 (m) = 1227.50-2.76 × (age) − 3.98 × log(weight) + 9.58 × (gender) + 1535.96 × log(height)

Using all known variables including those obtained during the test analogous to Formula 2 yielded the following formula explaining 79% of the variance in 6MWD (r2 = 0.79):

Formula 4:

6MWDtest2 (m) = 949.82 − 0.98 × (age) − 463.02 × log(weight) − 2.37 × (HR before 6MWT) + 3.03 × (HR after 6MWT) − 15.22 × (gender) + 1713.90 × log(height)

Gender was not a significant variable and did not change the r2 in any of the described analyses, but was added for adjustment.

Fewer subjects reported test-related pain after the intervention (Table 4). Again, there was no difference in 6MWD between those who reported pain and those who did not (567 [518-617] vs. 601 [521-647] m, P = 0.195), and knee pain specifically was still significantly associated with a shorter 6MWD (487 [372-560] vs. 600 [530-640] m, P = 0.005). Those who reported knee pain during the test already at baseline showed a tendency toward a lesser increase in 6MWD (4.0 [1.0-10.0]% vs. 9.0 [3.0-13.0]%, P = 0.051) and a significantly less pronounced weight loss after intervention (4.6 [1.1-12.2]% vs. 11.4 [5.0-21.2]%, P = 0.040). The reported degree of effort decreased significantly after intervention (3 [3-4] vs. 2 [1-3], P < 0.001).

The increase in 6MWD relative to baseline (Δ6MWD) correlated to decrease in weight (Figure 3A; r = 0.354, P < 0.001), decrease in BMI (r = 0.345, P < 0.001), decrease in waist circumference (r = 0.372, P < 0.001), increase in HR after the test (r = 0.321, P < 0.001), and increase in ΔHR during the test (Figure 3B; r = 0.434, P < 0.001) in a univariate analysis. There was no difference in percentage increase in walking distance after intervention between the metabolically at-risk obese subjects and the metabolically healthy obese subjects (8.0% vs. 9.0%, P = 0.781; Figure 2). A multiple regression analysis with Δ6MWD as the dependent variable generated the following formula explaining 38% of the variance (r2 = 0.38):

Formula 5:

Δ6MWD (m) = 77.34 − 0.17 × (age at follow-up) − 0.92 × (BMI at follow-up) − 33.05 × (knee pain at follow-up) − 1.23 × (Δweight) + 1.57 × (Δ[HR after 6MWT] during intervention) − 8.55 × (gender)

thumbnail image

Figure 3. Correlation between percentage change in 6MWD and percentage change in weight (A) and percentage change in increased heart rate (HR) during the test (B) after the weight reduction intervention program.

Download figure to PowerPoint

Gender as well as BMI and age at follow-up were not significant variables, but were added for adjustment as these factors were expected to influence the walking distance.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

In this observational intervention study, we have shown that the 6MWT is well tolerated by obese subjects and can be used to evaluate individual physical capacity and the expected effects of weight loss in both men and women. Using easily obtainable clinical data, it is possible to predict the expected performance during the test, both prior to and after intervention, and also to predict the improvement in walking distance in relation to weight loss. Gender had no impact on walking distance either before or after intervention. The metabolically at-risk obese subjects performed worse during the 6MWT when compared with the metabolically healthy subjects; however, both groups had a comparable effect of intervention based on increase in walking distance. Reported pain during the test was common at baseline, and it had a significant impact on the walking distance that was dependent on location.

Before the program, the median 6MWD was 535 m, equivalent to a median walking pace of 1.49 m/s. Restricting our analysis to BMI > 35 kg/m2 gave a similar walking pace of 1.52 m/s, which is comparable with a previous study in obese (BMI > 35 kg/m2) and lean (BMI ≤ 26 kg/m2) women who walked 540 (1.5 m/s) and 720 m (2.0 m/s), respectively, during 6 min (20). 6MWD in normal weight subjects is influenced by weight, height, age, and gender, as well as cardiovascular and pulmonary diseases and musculoskeletal factors (13, 28, 29). Some practical factors such as the length of the walkway, encouragement, and familiarity with the test may also influence the result (30, 31). We were able to show that weight, height, age, and pain reported during the test also influenced the 6MWD in obese subjects; however, in contrast to lean subjects, gender had no impact.

We used available data in an attempt to generate a formula to be used to calculate an expected 6MWD in clinical practice. Including factors known prior to performing the test in this analysis, age, use of β-blocker, fP-glucose, height, weight, resting HR, and smoking status, proved useful to predict 6MWD. The three factors with the greatest impact were weight, height, and age, accounting for 35% of the variance in 6MWD. Including all known variables (also those registered during the test), additional significant factors such as increase in HR during the test, presence of diabetes, and knee pain were identified. In this formula, HR after the test was the second most influencing variable, by itself explaining 19% of the variance in 6MWD. Additionally, we found that a shorter 6MWD at baseline was associated with cardiovascular risk markers such as increased waist circumference, fP-glucose, and blood pressure. The presence of metabolic syndrome alone did not affect the walking capacity, whereas diabetes (with or without the metabolic syndrome) and increased fP-glucose levels had a negative influence. This implicates that impaired glucose homeostasis and associated comorbidities contribute more to the impaired walking capacity than does the other parameters included in the metabolic syndrome such as hypertension, dyslipidemia, and increased waist circumference. All factors influencing 6MWD at follow-up were also seen at baseline; however, the effect of resting HR, smoking status, diabetes, and activity-related pain were attenuated after weight loss. This suggests that with decreasing weight, the impact of the factors known to influence the 6MWD in normal weight subjects become relatively more influential. As was the case at baseline, a shorter 6MWD at follow-up was associated with several cardiovascular risk markers, including higher levels of total cholesterol and LDL.

The greatest increase in 6MWD was seen in those with the greatest weight loss, those who improved their ability to increase HR during the test, and those who were not restricted by knee pain during physical activity. Although we lack a more precise and objective measure of physical fitness, the increase in HR after the test seen as a result of the intervention is a proxy for increased physical fitness secondary to the weight loss itself or to other associated unknown factors. Meanwhile, changes in other markers that may be associated with physical fitness such as triglycerides, high-density lipoprotein, cholesterol, and HOMA-IR index were not correlated to change in 6MWD. Both increased physical activity and weight loss may improve metabolic outcome; however, it is still not clear which contribute more than the other. A previous study has shown that both aerobic and resistance training increases the 6MWD of obese women irrespective of weight loss (22). The subjects here showed an improvement in walking capacity after weight loss in a program without special emphasis on physical activity other than general encouragement.

Gender has been shown to explain 9% of the variance in walking distance in healthy elderly subjects where men had a greater walking distance when compared with women (29). In contrast to the findings in normal weight subjects, we found no difference in 6MWD between men and women. Men had a higher BMI when compared with women at baseline; however, gender had no significant independent effect on 6MWD even after adjusting for weight and height by multiple regression. As far as we know, only one previous study on 6MWT in obese adults has assessed the difference in walking distance between men and women and was unable to find an independent effect of gender (23). It may be hypothesized that weight loss would increase 6MWD more in men than in women, restoring the relation found in normal weight subjects. However, despite comparable weight loss in men and women (11.8% vs. 11.5%), there was still no difference regarding walking distance after the intervention. There are several possible explanations for this finding. Factors explaining the inability for the obese men in our study to restore the expected gender difference in walking distance in response to weight loss may be important contributors to why they became obese in the first place. Such factors would be important to address in future studies. Alternatively, the male inability may suggest that the females in this study benefited relatively more from the intervention when compared with the men.

More than half of the subjects reported pain during the first test. This proportion was decreased by half after the intervention. In general, experiencing pain did not affect the walking distance. It has been previously shown that obese women experienced significantly more pain than lean women during the 6MWT but that this did not necessarily translate into a shorter 6MWD (20, 23). Here, pain over the tibia was the most common localization reported by one of every five subjects. Tibial pain was associated with a longer 6MWD, something we interpret as a consequence of physical exertion rather than a restricting factor for walking. In either case, the results clearly illustrate the multifaceted deconditioning of obese subjects. Under these conditions, tibial pain is probably due to mechanical stress on the tibial periosteum associated with both a low level of physical fitness in this group as well as workload. The decrease in reported pain over tibia after the program might be due to weight loss and decreased workload. As the subjects were encouraged to increased physical activity as a part of the intervention program, it might alternatively be due to increased physical fitness with an increased lactate threshold. Pain from the knee, foot, and lower back was also common, similar to previous studies. Pain from the knee was a restricting factor that influenced walking capacity negatively both prior to and after weight loss. More importantly, those who experienced knee pain already at baseline did not benefit as much from the intervention program based on change in walking capacity and weight loss. As knee pain can be assumed to negatively affect most physical training activities, it might be hypothesized that either the pain itself or the lack of exercise impairs the motivation to keep to the instructions during the intervention. Although activity-related pain in general was common in obese subjects, we found a prompt decrease in activity-related pain after weight loss. As pain can be expected to have an impact on compliance to advice regarding physical activity, this is important to consider during different stages of a weight reduction intervention program.

Our study has some limitations. Even though compliance with the program was relatively high (67.2%), the dropouts may constitute a nonrandom selection. As there were no major differences regarding clinical characteristics or 6MWT results between those who performed the test only at baseline and those who repeated the test at follow-up, we feel that this is a minor problem. It is known that familiarity to the 6MWT increases the walking distance and that this effect can remain for up to 2 months (31). In our study, the subjects performed the test once at baseline and then again at follow-up 6 to 9 months later. The impact of familiarity should therefore be minimal, which is also supported by the guidelines for 6MWT provided by the American Thoracic Society (30). As lean subjects have been studied previously, both alone and in comparison with obese subjects, we chose not to include a control group.

In conclusion, the 6MWT is a test for assessing walking capacity or physical performance and it is well tolerated by obese subjects. The test is easy to use and can easily be implemented in a clinical setting to evaluate obese subjects and how they might respond to weight loss in terms of physical activity. It is encouraging for the patient to be able to predict an increase in exercise capacity with weight reduction intervention, and physicians can guide the discussion when results differ from the expected. Activity-related pain was seen in more than half of the subjects but responded promptly to weight loss. This is important to consider when giving advice on physical activity as a part of weight loss programs. As seen now in two studies, male obese subjects underperform when compared with females when evaluated by the 6MWT. Whether men also benefit less from the kind of interventional program instituted here or have a relative inability to perform physical activity to begin with remains to be investigated.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The authors thank dietician Rosita Lindholm for her cooperation.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References