Skip to main content

Table 5 Multiple linear regression model: Determinants of greater self-care

From: Self-care in people with long term health problems: a community based survey

Step

Variable

R2

R2 change

Final β

95% CI

1

Age group

0.088

0.088

  
 

   18 - 40

  

0.79*

0.20 to 1.38

 

   41 - 64

  

0.50*

0.02 to 0.97

 

   65+

  

0.00

 
 

Sex:

    
 

   Male

  

0.96**

0.54 to 1.37

 

   Female

  

0.00

 

2

Primary long-term health problem

0.135

0.047

  
 

   Headaches or migraine

  

-1.12**

-1.82 to -0.42

 

   Tiredness or fatigue

  

-0.84*

-1.45 to -0.23

 

   Tummy or bowel problems

  

-1.40**

-1.97 to -0.82

 

   Feeling stressed or anxious

  

-0.62*

-1.23 to -0.02

 

   Menstrual problems

  

-2.87**

-3.92 to -1.83

 

   Back pain

  

0.00

 

3

Bothersomeness of problem

0.217

0.082

0.47**

0.23 to 0.72

4

Having a diagnosis

0.243

0.026

0.54*

0.10 to 0.98

5

Have seen a specialist

0.252

0.009

0.49*

0.04 to 0.94

6

Number of long term health conditions

0.265

0.013

0.23*

0.05 to 0.40

7

Sources of health information

0.396

0.131

  
 

   GP

  

0.01

-0.45 to 0.48

 

   Nurse

  

0.80*

0.21 to 1.39

 

   Pharmacist

  

0.71*

0.01 to 1.41

 

   Therapist

  

0.71

-0.03 to 1.44

 

   Complementary and alternative medicine

  

0.89*

0.23 to 1.55

 

   Friend

  

1.09**

0.52 to 1.66

 

   Family member

  

0.94**

0.44 to 1.45

 

   Internet chat room/forum

  

0.50

-1.12 to 2.12

 

   Support group

  

0.48

-0.93 to 1.89

 

   Newspaper or magazine

  

0.79*

0.03 to 1.54

 

   TV programme

  

0.62

-0.38 to 1.61

 

   Web sites

  

0.18

-0.34 to 0.71

8

Sources of information trusted

0.400

0.004

  
 

   Formal (GP, nurse, pharmacist, therapist, CAM)

  

0.20*

0.04 to 0.36

 

   Informal (Friend, family, forum, support group)

  

0.06

-0.22 to 0.34

 

   Media (Paper, TV, website)

  

-0.10

-0.44 to 0.24

  1. β = regression coefficient
  2. * p < 0.05, ** p < 0.005
  3. Regression coefficients (betas) for variables with mutually exclusive categories (including age, sex and presenting problem) are expressed relative to a reference category. For variables, such as sources of information used, where respondents can select more than one category, there is no reference category: regression coefficients indicate the mean number of additional self-care practices for respondents selecting a particular category compared to respondents not choosing that category