On this page:

Characteristics of Adults in Scotland with Long-Term Health Conditions: An Analysis of Scottish Household and Scottish Health Surveys

« Previous | Contents | Next »

Listen

MULTIVARIATE ANALYSIS OF SCOTTISH HEALTH SURVEY DATA

Tables 8.1 to 8.6 and 8.9 to 8.14 are based on logistic regression analysis of a dichotomous outcome (dependent) variable. They report the odds ratios ( OR), which represent the ratio of the probability of an event occurring versus it not occurring. For example, Table 8.1 displays the ORs of the probability of having some period of acute illness in the fortnight preceding the SHeS versus the probability of no period of acute illness. An OR of greater than one means the relative likelihood of an event happening is increased (as compared to the reference or base category); a lower relative likelihood is represented by an OR of less than one. All variables which are statistically significant at p<0.05 or greater are depicted in bold text.

Tables 8.7 and 8.8 report the results of univariate analysis of variance ( ANOVA), used to determine whether samples come from populations with equal means. The F statistic is the ratio of 'within-groups variance' to 'between-groups variance', and is a measure of how much variance is attributable to the different independent variables, versus the variance expected from random sampling. To determine if the F statistic is sufficiently large to reject the null hypothesis of no difference between groups, the F distribution is referred at a specified level, usually 0.05 or below. Post-hoc tests were then conducted to examine the actual differences in means.

Table 8.1 Logistic regression analysis of acute illness

Odds ratio 1

95% confidence intervals

Significance

(p value)

Age

0.001

16-24 (base)

1

25-34

1.44

1.04-2.01

0.029

35-44

1.43

1.04-1.97

0.026

45-54

1.48

1.08-2.04

0.016

55-64

1.07

0.77-1.50

0.683

65-74

0.92

0.65-1.30

0.628

75+

1.14

0.80-1.64

0.468

Sex

Male (base)

1

Female

1.16

1.00-1.35

0.047

Socio-economic status ( NS-Sec 5)

0.054

Managerial and professional (base)

1

Intermediate

0.71

0.55-0.92

0.009

Small employers/own account workers

0.72

0.52-1.01

0.060

Lower supervisory & technical

0.96

0.74-1.25

0.766

Semi-routine & routine

0.87

0.72-1.06

0.172

Equivalised income

0.511

£32500+ (base)

1

£21511-£32500

1.05

0.82-1.33

0.717

£14322 - £21511

0.99

0.78-1.28

0.973

£9100 - £14332

1.20

0.93-1.54

0.164

<£9100

1.06

0.81-1.38

0.683

Index of Multiple Deprivation (Simd5)

0.341

Least deprived (5393-7.7347)

1

7.7354 - 13.5231

0.92

0.72-1.17

0.479

13.5303 - 21.0301

1.05

0.83-1.35

0.670

21.0421 - 33.5214

1.10

0.86-1.40

0.437

Most deprived (33.5277 - 87.5665)

1.17

0.91-1.50

0.220

Long-term conditions

No or non-limiting long-standing condition (base)

1

Limiting long-standing condition

7.37

6.31-8.62

0.000

Model fit (-2 log likelihood) = 4997.3

0.000

N in model = 6628

Note 1: Relative probability of some acute illness in previous fortnight versus no acute illness in previous fortnight.

Table 8.2 Logistic regression analysis of Body Mass Index ( BMI)

Odds ratio 1

95% confidence intervals

Significance

(p value)

Age

0.000

16-24 (base)

1

25-34

1.99

1.45-2.73

0.000

35-44

2.96

2.20-3.98

0.000

45-54

3.27

2.42-4.42

0.000

55-64

4.05

2.98-5.49

0.000

65-74

3.96

2.88-5.46

0.000

75+

2.17

1.51-3.12

0.000

Sex

Male (base)

1

Female

1.21

1.05-1.38

0.007

Index of Multiple Deprivation (Simd5)

0.000

Least deprived (5393-7.7347)

1

7.7354 - 13.5231

1.21

0.96-1.50

0.104

13.5303 - 21.0301

1.57

1.26-1.95

0.000

21.0421 - 33.5214

1.32

1.06-1.65

0.015

Most deprived (33.5277 - 87.5665)

1.76

1.41-2.20

0.000

Long-term conditions

0.000

No long-standing condition (base)

1

Non-limiting long-standing condition

1.61

1.33-1.95

0.000

Limiting long-standing condition

1.91

1.63-2.25

0.000

Model fit (-2 log likelihood) = 5032.50

0.000

N in model = 4789

Note 1: Relative probability of bmi of 30 or over versus bmi of less than 30.

Table 8.3 Logistic regression analysis of waist-hip ratio (men)

Odds ratio 1

95% confidence intervals

Significance

(p value)

Age

0.000

16-24 (base)

1

25-34

4.39

3.03-6.36

0.000

35-44

12.31

8.57-17.69

0.000

45-54

16.80

11.52-24.49

0.000

55-64

26.49

17.62-39.84

0.000

65-74

21.64

14.01-33.42

0.000

75+

27.72

16.64-46.18

0.000

Index of Multiple Deprivation (Simd5)

0.148

Least deprived (5393-7.7347)

1

7.7354 - 13.5231

1.00

0.76-1.32

0.993

13.5303 - 21.0301

1.26

0.95-1.67

0.112

21.0421 - 33.5214

1.06

0.80-1.40

0.709

Most deprived (33.5277 - 87.5665)

1.36

1.01-1.83

0.045

Long-term conditions

0.014

No long-standing condition (base)

1

Non-limiting long-standing condition

1.37

1.05-1.78

0.021

Limiting long-standing condition

1.32

1.04-1.66

0.021

Model fit (-2 log likelihood) = 2804.20

0.000

N in model = 2533

Note 1: Relative probability of high waist-hip ratio versus normal/low waist-hip ratio.

Table 8.4 Logistic regression analysis of waist-hip ration (women)

Odds ratio 1

95% confidence intervals

Significance

(p value)

Age

0.000

16-24 (base)

1

25-34

1.38

0.97-1.95

0.075

35-44

2.03

1.47-2.82

0.000

45-54

2.44

1.75-3.41

0.000

55-64

2.90

2.06-4.08

0.000

65-74

3.81

2.69-5.57

0.000

75+

3.87

2.69-5.57

0.000

Index of Multiple Deprivation (Simd5)

0.000

Least deprived (5393-7.7347)

1

7.7354 - 13.5231

1.17

0.90-1.52

0.239

13.5303 - 21.0301

1.23

0.94-1.60

0.745

21.0421 - 33.5214

1.55

1.19-2.02

0.001

Most deprived (33.5277 - 87.5665)

2.11

1.62-2.76

0.000

Long-term conditions

0.000

No long-standing condition (base)

1

Non-limiting long-standing condition

1.15

0.90-1.46

0.259

Limiting long-standing condition

1.61

1.32-1.96

0.000

Model fit (-2 log likelihood) = 3320.14

0.000

N in model = 2679

Note 1: Relative probability of high waist-hip ratio versus normal/low waist-hip ratio..

Table 8.5 Logistic regression analysis of self-assessed general health

Odds ratio 1

95% confidence intervals

Significance

(p value)

Age

0.000

16-24 (base)

1

25-34

1.87

1.35-2.58

0.000

35-44

1.90

1.40-2.59

0.000

45-54

2.91

2.13-3.97

0.000

55-64

3.35

2.45-4.58

0.000

65-74

3.21

2.32-4.44

0.000

75+

3.61

2.57-5.06

0.000

Sex

Male (base)

1

Female

0.97

0.84-1.12

0.962

Socio-economic status ( NS-Sec 5)

0.000

Managerial and professional (base)

1

Intermediate

1.22

0.96-1.57

0.109

Small employers/own account workers

1.43

1.06-1.94

0.019

Lower supervisory & technical

1.40

1.09-1.81

0.008

Semi-routine & routine

1.61

1.33-1.95

0.000

Equivalised income

0.000

£32500+ (base)

1

£21511-£32500

1.39

1.08-1.78

0.009

£14322 - £21511

2.10

1.64-2.68

0.000

£9100 - £14332

2.78

2.16-3.56

0.000

<£9100

2.72

2.09-3.54

0.000

Index of Multiple Deprivation (Simd5)

0.000

Least deprived (5393-7.7347)

1

7.7354 - 13.5231

1.01

0.80-1.29

0.908

13.5303 - 21.0301

1.08

0.85-1.37

0.538

21.0421 - 33.5214

1.38

1.09-1.75

0.007

Most deprived (33.5277 - 87.5665)

1.62

1.27-2.07

0.000

Long-term conditions

No or non-limiting long-standing condition (base)

1

Limiting long-standing condition

11.07

9.62-12.74

0.000

Model fit (-2 log likelihood) = 5326.56

0.000

N in model = 6632

Note 1: Relative probability of reporting health as 'fair/poor or very poor' versus 'good/very good'.

Table 8.6 Logistic regression analysis of GHQ scores

Odds ratio 1

95% confidence intervals

Significance

(p value)

Age

0.000

16-24 (base)

1

25-34

1.16

0.86-1.57

0.330

35-44

1.19

0.90-1.59

0.229

45-54

1.24

0.92-1.66

0.156

55-64

0.80

0.59-1.10

0.174

65-74

0.60

0.43-0.85

0.004

75+

0.62

0.43-0.89

0.010

Sex

Male (base)

1

Female

1.26

1.08-1.47

0.003

Socio-economic status ( NS-Sec 5)

0.361

Managerial and professional (base)

1

Intermediate

1.08

0.84-1.39

0.561

Small employers/own account workers

0.82

0.58-1.17

0.267

Lower supervisory & technical

0.83

0.62-1.10

0.194

Semi-routine & routine

1.02

0.83-1.25

0.836

Equivalised income

0.000

£32500+ (base)

1

£21511-£32500

0.98

0.75-1.26

0.850

£14322 - £21511

1.35

1.04-1.74

0.022

£9100 - £14332

1.42

1.09-1.84

0.010

<£9100

1.83

1.39-2.40

0.000

Index of Multiple Deprivation (Simd5)

0.210

Least deprived (5393-7.7347)

1

7.7354 - 13.5231

0.93

0.72-1.20

0.596

13.5303 - 21.0301

1.18

0.92-1.52

0.197

21.0421 - 33.5214

1.20

0.93-1.54

0.156

Most deprived (33.5277 - 87.5665)

1.11

0.86-1.45

0.420

Long-term conditions

No or non-limiting long-standing condition (base)

1

Limiting long-standing condition

4.34

3.70-5.10

0.000

Model fit (-2 log likelihood) = 4779.13

0.000

N in model = 6301

Note 1: Relative probability of high GHQ score versus low GHQ score.

Table 8.7 Analysis of variance of PCS-12

F statistic

Significance

(p value)

Means

Age

111.09

0.000

16-24

52.98

25-34

52.37

35-44

51.49

45-54

49.62

55-64

46.73

65-74

43.25

75+

39.31

Sex

1.72

0.189

Male

49.18

Female

48.49

Index of Multiple Deprivation (Simd5)

27.61

0.000

Least deprived (5393-7.7347)

51.44

7.7354 - 13.5231

49.96

13.5303 - 21.0301

48.76

21.0421 - 33.5214

47.56

Most deprived (33.5277 - 87.5665)

45.50

Long-term conditions

2041.71

0.000

No or non-limiting long-standing condition

52.72

Limiting long-standing condition

39.96

Model fit (adjusted R 2) = 0.505

N in model = 6330

Table 8.8 Analysis of variance of MCS-12

F statistic

Significance

(p value)

Means

Age

30.24

0.000

16-24

51.67

25-34

51.25

35-44

50.76

45-54

50.86

55-64

52.26

65-74

53.42

75+

52.85

Sex

28.97

0.000

Male

52.41

Female

50.93

Index of Multiple Deprivation (Simd5)

11.36

0.000

Least deprived (5393-7.7347)

52.81

7.7354 - 13.5231

52.64

13.5303 - 21.0301

51.67

21.0421 - 33.5214

50.94

Most deprived (33.5277 - 87.5665)

49.59

Long-term conditions

295.17

0.000

No or non-limiting long-standing condition

53.14

Limiting long-standing condition

48.19

Model fit (adjusted R 2) = 0.126

N in model = 6330

Table 8.9 Logistic regression analysis of GP consultations

Odds ratio 1

95% confidence intervals

Significance

(p value)

Age

0.210

16-24 (base)

1

25-34

1.06

0.84-1.33

0.640

35-44

0.82

0.65-1.03

0.092

45-54

0.99

0.78-1.24

0.904

55-64

0.98

0.77-1.24

0.857

65-74

0.93

0.72-1.20

0.571

75+

1.09

0.85-1.42

0.495

Sex

Male (base)

1

Female

1.29

1.14-1.46

0.000

Index of Multiple Deprivation (Simd5)

0.541

Least deprived (5393-7.7347)

1

7.7354 - 13.5231

1.14

0.94-1.39

0.171

13.5303 - 21.0301

1.03

0.85-1.26

0.745

21.0421 - 33.5214

1.10

0.91-1.34

0.316

Most deprived (33.5277 - 87.5665)

1.14

0.94-1.39

0.172

Acute illness in previous fortnight

No (base)

1

Yes

2.76

2.39-3.20

0.000

Self-assessed general health

Very good/good (base)

1

Fair/poor/very poor

1.98

1.70-2.31

0.000

Long-term conditions

0.000

No long-standing condition (base)

1

Non-limiting long-standing condition

1.42

1.19-1.70

0.000

Limiting long-standing condition

1.61

1.36-1.91

0.000

Model fit (-2 log likelihood) = 6985.81

0.000

N in model = 7720

Note 1: Relative probability of consulting with GP in previous fortnight versus not consulting.

Table 8.10 Logistic regression analysis of attendance at outpatients

Odds ratio 1

95% confidence intervals

Significance

(p value)

Age

0.481

16-24 (base)

1

25-34

0.86

0.72-1.03

0.096

35-44

1.01

0.85-1.20

0.925

45-54

1.01

0.85-1.22

0.877

55-64

0.98

0.81-1.18

0.793

65-74

0.93

0.76-1.13

0.458

75+

Sex

Male (base)

1

Female

0.94

0.85-1.03

0.196

Index of Multiple Deprivation (Simd5)

0.323

Least deprived (5393-7.7347)

1

7.7354 - 13.5231

1.00

0.86-1.16

0.998

13.5303 - 21.0301

1.16

0.99-1.35

0.064

21.0421 - 33.5214

1.03

0.89-1.20

0.701

Most deprived (33.5277 - 87.5665)

1.03

0.88-1.20

0.752

Acute illness in previous fortnight

No (base)

1

Yes

0.53

0.46-0.60

0.000

Self-assessed general health

Very good/good (base)

1

Fair/poor/very poor

0.64

0.57-0.73

0.000

Long-term conditions

0.000

No long-standing condition (base)

1

Non-limiting long-standing condition

0.52

0.45-0.60

0.000

Limiting long-standing condition

0.42

0.36-0.48

0.000

Model fit (-2 log likelihood) = 9763.03

0.000

N in model = 8083

Note 1: Relative probability of 'not attending' versus 'attending' hospital as an outpatient

Table 8.11 Logistic regression analysis of attendance as an inpatient

Odds ratio 1

95% confidence intervals

Significance

(p value)

Age

0.000

16-24 (base)

1

25-34

0.74

0.56-0.97

0.027

35-44

1.28

0.97-1.70

0.081

45-54

1.46

1.09-1.97

0.011

55-64

1.54

1.14-2.09

0.005

65-74

0.93

0.69-1.24

0.607

75+

0.80

0.59-1.09

0.154

Sex

Male (base)

1

Female

0.75

0.65-0.87

0.000

Index of Multiple Deprivation (Simd5)

0.004

Least deprived (5393-7.7347)

1

7.7354 - 13.5231

1.02

0.80-1.30

0.885

13.5303 - 21.0301

1.10

0.86-1.41

0.442

21.0421 - 33.5214

0.86

0.68-1.09

0.222

Most deprived (33.5277 - 87.5665)

0.74

0.59-0.93

0.010

Acute illness in previous fortnight

No (base)

1

Yes

0.63

0.53-0.75

0.000

Self-assessed general health

Very good/good (base)

1

Fair/poor/very poor

0.57

0.47-0.69

0.000

Long-term conditions

0.000

No long-standing condition (base)

1

Non-limiting long-standing condition

0.81

0.64-1.02

0.074

Limiting long-standing condition

0.49

0.40-0.60

0.000

Model fit (-2 log likelihood) = 5172.17

0.000

N in model = 8082

Note 1: Relative probability of 'not attending' versus 'attending' hospital as an inpatient

Table 8.12 Logistic regression analysis of medication

Odds ratio 1

95% confidence intervals

Significance

(p value)

Age

0.000

16-24 (base)

1

25-34

1.27

0.97-1.64

0.077

35-44

1.33

1.04-1.71

0.025

45-54

2.29

1.78-2.95

0.000

55-64

4.78

3.67-6.24

0.000

65-74

11.57

8.49-15.77

0.000

75+

22.77

15.50-33.45

0.000

Sex

Male (base)

1

Female

1.79

1.55-2.06

0.000

Index of Multiple Deprivation (Simd5)

0.330

Least deprived (5393-7.7347)

1

7.7354 - 13.5231

1.01

0.82-1.25

0.905

13.5303 - 21.0301

1.10

0.89-1.36

0.396

21.0421 - 33.5214

0.86

0.69-1.08

0.192

Most deprived (33.5277 - 87.5665)

0.98

0.78-1.24

0.878

Acute illness in previous fortnight

No (base)

1

Yes

1.81

1.46-2.24

0.000

Self-assessed general health

Very good/good (base)

1

Fair/poor/very poor

2.71

2.23-3.29

0.000

Long-term conditions

0.000

No long-standing condition (base)

1

Non-limiting long-standing condition

6.24

5.15-7.57

0.000

Limiting long-standing condition

6.26

5.12-7.65

0.000

Model fit (-2 log likelihood) = 4915.20

0.000

N in model = 5441

Note 1: Relative probability of currently taking medication versus no taking medication

Table 8.13 Logistic regression analysis of smoking status

Odds ratio 1

95% confidence intervals

Significance

(p value)

Age

0.000

16-24 (base)

1

25-34

1.43

1.20-1.71

0.000

35-44

1.22

1.03-1.45

0.025

45-54

0.98

0.82-1.18

0.849

55-64

0.76

0.63-0.93

0.006

65-74

0.53

0.43-0.66

0.000

75+

0.31

0.24-0.40

0.000

Sex

Male (base)

1

Female

0.93

0.84-1.03

0.170

Index of Multiple Deprivation (Simd5)

0.000

Least deprived (5393-7.7347)

1

7.7354 - 13.5231

1.29

1.08-1.53

0.004

13.5303 - 21.0301

1.70

1.44-2.02

0.000

21.0421 - 33.5214

2.41

2.04-2.84

0.000

Most deprived (33.5277 - 87.5665)

3.89

3.30-4.59

0.000

Long-term conditions

0.000

No long-standing condition (base)

1

Non-limiting long-standing condition

0.92

0.79-1.08

0.314

Limiting long-standing condition

1.35

1.19-1.52

0.000

Model fit (-2 log likelihood) = 9031.61

0.000

N in model = 7537

Note 1: Relative probability of currently smoking versus non or ex-smoker

Table 8.14 Logistic regression analysis of alcohol consumption

Odds ratio 1

95% confidence intervals

Significance

(p value)

Age

0.000

16-24 (base)

1

25-34

0.87

0.72-1.06

0.169

35-44

0.75

0.62-0.90

0.002

45-54

0.87

0.72-1.05

0.149

55-64

0.71

0.58-0.87

0.001

65-74

0.49

0.39-0.63

0.000

75+

0.26

0.19-0.36

0.000

Sex

Male (base)

1

Female

0.45

0.40-0.50

0.000

Index of Multiple Deprivation (Simd5)

0.811

Least deprived (5393-7.7347)

1

7.7354 - 13.5231

0.96

0.81-1.14

0.676

13.5303 - 21.0301

0.91

0.76-1.08

0.281

21.0421 - 33.5214

0.95

0.80-1.14

0.597

Most deprived (33.5277 - 87.5665)

1.00

0.84-1.20

0.971

Long-term conditions

0.002

No long-standing condition (base)

1

Non-limiting long-standing condition

1.03

0.87-1.21

0.760

Limiting long-standing condition

0.78

0.68-0.91

0.001

Model fit (-2 log likelihood) = 7699.84

0.000

N in model = 7988

Note 1: Relative probability of drinking over weekly limits versus drinking below weekly limits

« Previous | Contents | Next »

Page updated: Monday, October 29, 2007