Table 3

Linear regression models predicting log of C-reactive protein

B
SE		B
SE		B
SE		B
SE
Model 1Model 2Model 3Model 4
Housing tenure (ref: owned with mortgage)
 Owned outright
 		0.22***−0.020.010.02
0.030.030.030.03
 Social rent
 		0.36***0.18***0.09*0.08
0.040.040.040.05
 Private rent
 		0.060.12*0.11*0.10*
0.050.050.050.05
Housing cost burden (yes)−0.10−0.15*−0.10−0.14*
0.070.070.070.07
Housing payment arrears (yes)0.050.060.010.00
0.060.060.050.05
Housing benefit (yes)0.26***0.090.01−0.05
0.040.050.040.05
Central heating (no)0.070.020.040.04
0.050.050.050.05
Warm enough (no)0.13*0.02−0.03−0.04
0.060.060.060.06
Overcrowding (yes)−0.030.050.010.00
0.050.060.050.05
Dwelling type (ref: detached)
 Semidetached
 		0.10***0.11***0.06*0.07*
0.030.030.030.03
 Terrace
 		0.15***0.19***0.13***0.13***
0.030.030.030.03
 Flat
 		0.11*0.20***0.18***0.18***
0.050.050.040.05
 Other
 		0.260.210.170.16
0.160.170.170.17
Prefer to stay0.12***0.040.06*0.05*
0.030.030.030.03
Expect to move−0.20***−0.06−0.03−0.06
0.040.050.040.05
Prefer to stay × expect to move–––0.15
0.10
 Demographics and SEP controls includedNYYY
 Health and health behaviour controls includedNNYY

  • Demographic and socioeconomic position controls: age, gender, cohabitation, ethnicity, region, employment status, highest educational qualification, income quartile (age standardised), subjective financial situation.

  • Health and health behaviour controls: smoking status, body mass index (categorised), long-standing illness or disability.

  • Regressions for housing characteristics run separately in models 1–3, but presented in a single column for ease of comparison and conciseness. All housing variables are included simultaneously in model 4.

  • n=9974.

  • *P<0.05; **P<0.01; ***P<0.005.