GROWING UP IN SCOTLAND: THE CIRCUMSTANCES AND EXPERIENCES OF 3 YEAR OLD CHILDREN LIVING IN SCOTLAND IN 2007/08 AND 2013

Chapter 7 Parental Health

This chapter presents findings on parental health including general health, and longstanding illness as well as measures of parental physical and mental wellbeing.

Parental health and wellbeing is an important influence on parenting behaviours and on children in their early years. Previous analysis of GUS data has shown that among children born into the most disadvantaged backgrounds, the likelihood of experiencing negative outcomes over their first four years (including poor cognitive, social, emotional and behavioural outcomes as well as poor physical health) was reduced if their mother had not experienced a long-term health problem or disability during that period (Bromley and Cunningham-Burley, 2010). Analysis of GUS data has also shown that children whose mothers were emotionally well during their first four years had better social, emotional and behavioural development than those whose mothers had brief mental health problems. Children of mothers with brief mental health problems, in turn, had better development than those whose mothers had repeated mental health problems. These relationships remained significant even after taking account of other family characteristics and socio-economic circumstances (Marryat and Martin, 2010).

7.1 General health

All respondents were asked to assess their general health as either 'excellent' 'very good, 'good', 'fair' or 'poor'. Self-assessed health is a useful measure of how individuals regard their own overall health status. Assessments have been shown to be strongly related to the presence of chronic and acute disease, as well as being a good predictor of hospital admission and mortality (Idler et al., 1997; Hanlon et al., 2007).

The vast majority of parents reported that their health was at least 'good' (86% in both BC1 and BC2, Table 7.1). There was a slight decrease in the proportion of parents reporting 'excellent' or 'very good health' between 2007/08 and 2013. In BC1, 21% of parents said their health was 'excellent' compared with 17% of parents in BC2. This small difference was statistically significant. The proportion of parents reporting 'poor' health decreased slightly from 3% in BC1 to 2% in BC2.

Table 7.1 Parental general health by cohort

	BC1	BC2
	%	%
Excellent	21	17
Very good	39	37
Good	26	32
Fair	11	12
Poor	3	2
Unweighted bases	4171	4975

Tested on category 'excellent', differences between cohort p <.001

Table 7.2 shows that there is a linear and statistically significant relationship between parental health and income: as income increases so too does the proportion of parents reporting 'excellent' health. In BC1, 29% of parents in the highest household income quintile reported excellent health compared with 14% of parents in the lowest household income quintile. This trend prevails amongst BC2 parents with 22% of parents from the highest income quintile households reporting excellent health compared with 14% of parents from households in the lowest income quintile. The decline in self-reported excellent health is seen across all income groups in BC2 with the exception of the lowest income group. However, there has been no statistically significant change in the nature of the relationship between health and household income between the two cohorts.

As well as there being a linear relationship between income and 'excellent' health, there is also a relationship between income and the proportion of parents reporting 'poor' health. Amongst BC1 parents, 6% of those in the lowest income quintile reported poor health compared with 1% of parents in the highest income quintile. There was no statistically significant difference, between BC1 and BC2 in the proportion of parents assessing their health as poor.

Table 7.2 Parental general health by equivalised household income (quintile) and cohort

	Lowest quintile	2nd quintile	3rd quintile	4th quintile	Highest quintile
	%	%	%	%	%
BC1
Excellent	14	19	20	28	29
Very good	29	41	40	42	44
Good	32	27	27	22	21
Fair	19	11	11	7	5
Poor	6	2	2	1	1
BC2
Excellent	14	13	16	22	22
Very good	28	37	38	39	44
Good	35	32	34	31	26
Fair	19	15	10	7	6
Poor	4	3	2	1	1
Unweighted bases - BC1	773	802	759	857	719
Unweighted bases - BC2	968	774	809	774	1020

Tested on category 'excellent': differences by income p < .001; differences by cohort p < .000; cohort*income p < .0286

Table 7.3 shows that parental general health is also related to the level of parental education. Whilst this relationship is not strictly linear, parents educated to degree level reported higher levels of excellent health than all the other groups. For example, amongst BC1 parents, 27% of parents educated to degree level reported excellent health compared with 19% of parents with no qualification. Amongst BC2 parents, 21% of parents educated to degree level said their health was excellent compared with 11-15% of parents with lower qualifications. Differences in self-reported health by levels of parental education were statistically significant in both cohorts. The decrease in the proportion of parents reporting 'excellent' or 'very good health' between 2007/08 and 2013 was seen across all levels of parental education. As such, as with household income, there was no statistically significant difference in the relationship between self-reported health and level of education between BC1 and BC2.

Focusing on the proportion of parents with poor health, the table shows that this has declined between the two cohorts amongst those with no qualifications. This difference is not statistically significant.

Table 7.3 Parental general health by parental level of education and cohort

	No qualifications	Lower Standard Grades or VQs or Other	Upper level SGs or Intermediate VQs	Higher grades and upper level VQs	Degree level academic and vocational qualifications
	%	%	%	%	%
BC1
Excellent	19	15	17	19	27
Very good	31	32	37	39	41
Good	21	35	30	28	22
Fair	22	13	13	11	7
Poor	6	5	3	2	2
BC2
Excellent	15	11	11	15	21
Very good	29	28	35	35	41
Good	32	40	33	33	28
Fair	22	17	16	13	8
Poor	2	5	4	3	1
Unweighted bases - BC1	195	204	799	1371	1595
Unweighted bases - BC2	155	211	709	1420	2311

Tested on category 'Excellent': differences by education p< .001; differences by cohort p< .001; cohort* education p=NS.

Table 7.4 shows that mothers aged 40 or older at the birth of their child were more likely to report excellent health than younger mothers. In BC1, 26% of mothers aged over 40 reported excellent health compared with 15% of mothers aged under 20 years old. In BC2, there is a similar pattern though reports of excellent health have decreased amongst all age groups. As such, whilst the proportion of each age group who reported excellent health has reduced, there have been no statistically significant changes in the relationship between maternal age and self-reported health between the cohorts. With regard to self-reported 'poor' health, whilst amongst the youngest mothers this appears to have decreased between BC1 and BC2 (4% to 3% respectively), this difference is not statistically significant.

Table 7.4 Parental general health by maternal age at child's birth and cohort

	Under 20 years old	20 to 29 years old	30 to 39 years old	40 or older
	%	%	%	%
BC1
Excellent	15	21	23	26
Very good	38	36	41	34
Good	33	27	25	24
Fair	11	13	9	13
Poor	4	3	2	2
BC2
Excellent	13	16	18	19
Very good	32	35	38	39
Good	38	31	31	29
Fair	13	15	10	10
Poor	3	3	2	2
Unweighted bases - BC1	220	1555	2219	155
Unweighted bases - BC2	214	1956	2558	225

Tested on category 'Excellent': differences by maternal age p< .01; differences by cohort p< .01; cohort* education p=NS.

Parents from the least deprived areas were more likely to report excellent health than parents from other areas. Differences by area deprivation are statistically significant in both cohorts (Table 7.5). For example, looking at BC2 parents, 20% of parents from the least deprived area reported excellent health compared with 14% of parents from the most deprived area. Conversely, if we look at poor health, we see that it was correlated to living in a more deprived area, a trend also evident in both cohorts. There has been no statistically significant change in the nature of the association between parental health and area deprivation between the cohorts.

Table 7.5 Parental general health by area deprivation (quintiles) and cohort

	1 Most deprived	2	3	4	5 Least deprived
	%	%	%	%	%
BC1
Excellent	16	19	22	25	27
Very good	34	36	41	40	42
Good	31	29	24	24	23
Fair	16	13	10	8	7
Poor	4	3	3	2	1
BC2
Excellent	14	15	16	19	20
Very good	32	32	37	41	43
Good	32	36	33	30	27
Fair	16	15	12	9	9
Poor	5	2	2	1	1
Unweighted bases - BC1	824	693	872	878	904
Unweighted bases - BC2	935	924	1018	1059	1006

Tested on category 'Excellent': differences by area deprivation p< .001; differences by cohort p< .001; cohort* education p=NS.

7.2 Parental longstanding illness or disability

All parents were asked whether they had any health problems or disabilities that had lasted or were expected to last for more than a year. There has been no change in the prevalence of parental longstanding illness between the cohorts. Overall, 17% of parents in BC1 and 16% of parents in BC2 reported a longstanding illness or disability when the child was aged three, and this difference is not statistically significant (Table 7.6).

Table 7.6 Parental longstanding illness or disability by cohort

	BC1	BC2
	%	%
Yes	17	16
No	83	84
Unweighted bases	3908	4344

Tested on 'Yes': differences by cohort p=NS

Parents who said they had a longstanding illness or disability were also asked about the extent to which this affected their ability to carry out day-to-day activities. It is important to note that whilst parents in BC1 and BC2 were asked the same question, they had different response categories to choose from. In BC1, parents could answer either 'Yes' or 'No', whilst BC2 parents could answer 'Yes, a lot'; 'Yes, a little' or 'No'.^[21] In BC1, 40% of parents said that their longstanding illness or disability affected them compared with 51% of parents in BC2 (9% a little, 42% a lot).

This increase in the impact of the longstanding illness /disability between BC1 and BC2 is likely to be due, at least in part, to the change in answer categories.

Table 7.7 Whether longstanding illness or disability affects parent in ability to carry out day-to-day activities by cohort

	BC1	BC2
	%	%
Yes	40	-
Yes, a little	-	9
Yes, a lot	-	42
No	60	49
Unweighted bases	661	741

Tested on category No: differences by cohort p<.001

Table 7.8 shows the proportion of parents reporting a longstanding illness/disability by equivalised household income and cohort. Parents from lower income households were more likely to report a longstanding illness or disability than other parents and this trend was evident in both cohorts. Approximately one in ten parents in the highest income quintile (11% BC1, 12% BC2) reported a longstanding illness compared with over two in ten in the lowest income quintile (25% in BC1, 20% in BC2). Whilst there appears to have been a decline in the prevalence of longstanding illness amongst parents in the lowest income group there were no statistically significant differences by cohort nor in the nature of the relationship between household income and longstanding illness between the cohorts.

Table 7.8 % of parents with longstanding illness/disability by equivalised household income and cohort

	Lowest quintile	2nd quintile	3rd quintile	4th quintile	Highest quintile
	%	%	%	%	%
BC1	25	17	15	11	11
BC2	20	16	14	12	12
Unweighted bases - BC1	772	802	758	857	719
Unweighted bases - BC2	968	774	808	774	1020

Tested on category 'has longstanding illness': differences by equivalised income p< .001; differences by cohort p=NS; cohort* equivalised income p=NS.

Table 7.9 shows that there are particular differences in the prevalence of a longstanding illness or disability according to parental level of education. The more educated the parent, the less likely he/she is to have a longstanding illness or disability. For both cohorts, around one in ten parents educated to degree level reported a longstanding illness or disability (13% in both BC1 and BC2) compared with three in ten of parents with no qualification (31% in BC1 and 30% in BC2). There were no differences in the relationship between parental level of education longstanding illness or disability between the cohorts.

Table 7.9 Parental longstanding illness/disability by parental level of education and cohort

	No qualifications	Lower Standard Grades or VQs or Other	Upper level SGs or Intermediate VQs	Higher grades and upper level VQs	Degree level academic and vocational qualifications
	%	%	%	%	%
BC1	31	18	17	17	13
BC2	30	19	15	15	13
Unweighted bases - BC1	195	204	789	1371	1594
Unweighted bases - BC2	155	211	709	1420	2310

Tested on category 'has longstanding illness': differences by education p< .001; differences by cohort p=NS; cohort* equivalised income p=NS.

Table 7.10 shows differences in the prevalence of longstanding illness and disability by maternal age at the birth of the cohort child. Whilst the data suggest that older mothers were more likely to report a longstanding illness or disability in both BC1 and BC2, differences across the age groups are not statistically significant in either cohort.

Table 7.10 Parental longstanding illness/disability by maternal age at child's birth and cohort

	Under 20 years old	20 to 29 years old	30 to 39 years old	40 or older
	%	%	%	%
BC1	15	17	16	18
BC2	10	16	16	16
Unweighted bases - BC1	220	1555	2217	155
Unweighted bases - BC2	214	1956	2557	225

Tested on category 'has longstanding illness': differences by maternal age p=NS; differences by cohort p=NS; cohort* equivalised maternal age p=NS.

Area deprivation is associated with a parent reporting a longstanding illness or disability (Table 7.11). Two in ten parents living in the most deprived areas (22% BC1; 20% BC2) had a longstanding illness or disability compared with one in ten (10% for both BC1 and BC2) in the least deprived areas. The table shows that the trend has remained the same for both cohorts and there has been no change in the nature of the relationship between parental longstanding illness/disability and area deprivation between the cohorts.

Table 7.11 Parental longstanding illness/disability by area deprivation (quintiles) and cohort

	1 Most deprived	2	3	4	5 Least deprived
	%	%	%	%	%
BC1	22	18	16	14	10
BC2	20	17	15	14	10
Unweighted bases - BC1	903	878	872	692	824
Unweighted bases - BC2	1006	1058	1018	924	935

Tested on category 'has longstanding illness': differences by area deprivation p< .001; differences by cohort p=NS; cohort* equivalised area deprivation p=NS.

7.3 Parental wellbeing

Health-related quality of life was measured by the Medical Outcomes Study 12-Item Short Form (SF-12). This measure has been used at previous interviews in GUS and is also widely used on other large population surveys such as the Scottish Health Survey (SHeS). The SF-12 gives two summary scale scores: a physical component score (PCS) and a mental component score (MCS). The physical wellbeing component measures the extent to which the parent's physical health limits their ability to perform a range of daily activities (such as climbing stairs). The mental wellbeing component measures the parent's recent mood and energy levels and the extent to which emotional problems limit their ability to accomplish things. Both are measures of health related quality of life. Higher summary scale scores in both PCS and MCS are indicative of better health-related quality of life. However, as the results are based on the respondents' self-reports of their own physical and mental functioning they are subjective and may lead to differential reporting between respondents with an equivalent health status.

7.4 Parental physical wellbeing

The average mean score of physical wellbeing was 52.6 amongst BC1 parents and 52.3 amongst BC2 parents. This small difference was not statistically significant. As such, there has been no change in physical wellbeing mean score between the cohorts.

Table 7.12 Parental physical wellbeing mean score by cohort

	BC1	BC2
	%	%
Physical wellbeing mean score	52.6	52.3
Unweighted bases	4170	4972

Tested on mean physical wellbeing mean score p = NS

Table 7.13 shows mean physical wellbeing scores by equivalised household income. Physical wellbeing was related to household income with mean wellbeing score increasing as household income increased. This pattern was evident in both cohorts. For example in BC2, the mean physical wellbeing score amongst parents in the lowest income group was 51.1 compared with 53.7 amongst parents in the highest income group. The table shows that there has been no notable change in the nature of the relationship between physical wellbeing and household income between the cohorts.

Table 7.13 Physical wellbeing mean scores by equivalised household income quintile and cohort

	Lowest quintile	2nd quintile	3rd quintile	4th quintile	Highest quintile
	%	%	%	%	%
BC1	50.5	52.5	52.8	53.9	53.8
BC2	51.1	52.0	52.8	53.6	53.7
Unweighted bases - BC1	773	802	759	857	718
Unweighted bases - BC2	967	773	809	774	1020

Tested on mean score: differences by equivalised household income p< .001; differences by cohort p=NS; cohort* equivalised household income p=NS.

Table 7.14 Physical wellbeing mean scores by parental level of education and cohort

	No qualifications	Lower Standard Grades or VQs or Other	Upper level SGs or Intermediate VQs	Higher grades and upper level VQs	Degree level academic and vocational qualifications
	%	%	%	%	%
BC1	49.8	51.4	52.3	52.7	53.3
BC2	51.5	50.9	51.3	52.3	53.2
Unweighted bases - BC1	195	204	799	1371	1594
Unweighted bases - BC2	154	209	709	1420	2311

Tested on mean score: differences by education p< .001; differences by cohort p=NS; cohort* education p=NS.

Table 7.15 shows that there is no discernible pattern in physical wellbeing by maternal age. However, in both cohorts, mothers aged between 30 and 39 at the birth of their child scored the highest physical wellbeing mean score (53.1 in BC1 and 52.5 in BC2). The slight differences in scores between the cohorts were not statistically significant.

Table 7.15 Physical wellbeing mean scores maternal age at child's birth and cohort

	Under 20 years old	20 to 29 years old	30 to 39 years old	40 or older
	%	%	%	%
BC1	51.8	52.1	53.1	52.3
BC2	52.3	52.3	52.5	52.1
Unweighted bases - BC1	220	1555	2218	155
Unweighted bases - BC2	213	1955	2557	225

Tested on mean score: differences by maternal age p< .01; differences by cohort p=NS; cohort* education p=NS.

Table 7.16 shows a linear relationship between area deprivation and physical wellbeing. Parents from the lowest area of deprivation reported the lowest mean scores of physical wellbeing. There has been no change in the nature of the relationship between area deprivation and parental physical wellbeing between the two cohorts.

Table 7.16 Parental Physical wellbeing mean scores by area deprivation (quintiles) and cohort

	1 Most deprived	2	3	4	5 Least deprived
	%	%	%	%	%
BC1	51.2	52.1	53.1	53.2	53.6
BC2	51.2	51.8	52.5	53.1	53.4
Unweighted bases - BC1	824	693	871	878	904
Unweighted bases - BC2	933	923	1018	1059	1006

Tested on mean score: differences by area deprivation p< .001; differences by cohort p=NS; cohort* area deprivation p=NS.

7.5 Parental mental wellbeing

The average mean mental wellbeing score was 49.6 amongst BC1 parents. This had increased slightly to 50.8 amongst BC2. This change was statistically significant (Table 7.17).

Table 7.17 Parental mental wellbeing mean score by cohort

	BC1	BC2
	%	%
Mental wellbeing mean score	49.6	50.8
Unweighted bases	4170	4972

Tested on mean wellbeing score; differences by cohort p<.001

Table 7.18 shows that there is an association between mental wellbeing and household income: as income increases so too does the mean mental wellbeing score. For example, in BC2 the mean mental wellbeing score amongst parents in the highest income quintile was 52.0 compared with a score of 50.6 among parents in the second income quintile and 49.1 amongst parents in the lowest income quintile. This relationship exists in both cohorts.

There has been a statistically significant increase in the mental wellbeing score of parents in all household income groups but particularly amongst those in the lowest income group. As a result, the difference in mean scores between those in the lowest income quintile and those in the highest income quintile has narrowed between BC1 and BC2. In BC1, the difference in mean score between those in the highest and lowest income quintile was 5.5 whilst for BC2 parents it was 2.9. These differences are statistically significant. Whilst this is an encouraging trend, it will be interesting to see if this pattern continues in future years as the children get older, or with future cohorts.

Table 7.18 Mental wellbeing mean score by equivalised household income (quintile) and cohort

	Lowest quintile	2nd quintile	3rd quintile	4th quintile	Highest quintile
	%	%	%	%	%
BC1	46.4	49.3	50.2	51.1	51.9
BC2	49.1	50.6	51.4	51.5	52.0
Unweighted bases - BC1	773	802	759	857	718
Unweighted bases - BC2	967	773	809	774	1020

Tested on mean score: differences by equivalised household income p< .001; differences by cohort p< .001; cohort* equivalised household income p <0.01

Table 7.19 shows that mental wellbeing increased with level of education. In both cohorts, parents educated to degree level had the highest mental wellbeing scores (BC1 = 50.7 and BC2 = 51.4) and parents with no qualifications had the lowest scores (BC1 = 45.5) and (BC2 = 49.1). There has been a slight and statistically significant increase in mean mental wellbeing scores across all levels of parental education between BC1 and BC2. However, there is no statistically significant difference in the nature of the relationship between education and mental wellbeing between the cohorts.

Table 7.19 Mental wellbeing mean scores by parental level of education and cohort

	No qualifications	Lower Standard Grades or VQs or Other	Upper level SGs or Intermediate VQs	Higher grades and upper level VQs	Degree level academic and vocational qualifications
	%	%	%	%	%
BC1	45.5	47.5	48.6	50.1	50.7
BC2	49.1	49.4	50.3	50.8	51.4
Unweighted bases - BC1	195	204	799	1371	1594
Unweighted bases - BC2	154	209	709	1420	2311

Tested on mean score: differences by education p< .001; differences by cohort p< .001; cohort* education p=NS.

Table 7.20 shows mental wellbeing mean scores by maternal age at the birth of the cohort child. There was no particular relationship between mental wellbeing and age although as with physical wellbeing, mothers aged 30-39 years had higher mean scores of mental wellbeing for both BC1 (50.0) and BC2 (51.1) than mothers in all other age groups. At the other end, younger mothers had the lowest scores (BC1 = 48.0, BC2 = 49.7). The increase in mental wellbeing score between BC1 and BC2 affected all age groups of mothers and there has been no statistically significant change in the nature of this relationship between the cohorts.

Table 7.20 Mental wellbeing mean scores maternal age at child's birth and cohort

	Under 20 years old	20 to 29 years old	30 to 39 years old	40 or older
	%	%	%	%
BC1	48.0	49.3	50.0	49.8
BC2	49.7	50.6	51.1	50.6
Unweighted bases - BC1	220	1555	2218	155
Unweighted bases - BC2	213	1955	2557	225

Tested on mean score: differences by maternal age p< .01; differences by cohort p< .001; cohort* education p=NS.

There was a linear relationship between area deprivation and mental wellbeing, as deprivation increases, mental wellbeing decreases (Table 7.21). This trend is evident in both cohorts. The increase in mean mental wellbeing scores between BC1 and BC2 has affected parents from all areas of deprivation. Furthermore, we can also see that there has been a narrowing of the gap between mean scores of mental wellbeing between parents from households in the lowest income quintile and parents in the highest income quintile between BC1 and BC2.

Table 7.21 Mental wellbeing mean scores by area deprivation (quintiles) and cohort

	1 Most deprived	2	3	4	5 Least deprived
	%	%	%	%	%
BC1	51.2	52.1	53.1	53.2	53.6
BC2	51.2	51.8	52.5	53.1	53.4
Unweighted bases - BC1	824	693	871	878	904
Unweighted bases - BC2	933	923	1018	1059	1006

Tested on mean score: differences by area deprivation p< .001; differences by cohort p< .001; cohort* area deprivation p< .05

7.6 Parental smoking habits

During the interview parents were asked if they currently smoked cigarettes. The comparison included here is simply on whether or not the parent smokes and not on number of cigarettes smoked or smoking in the household.

There has been a reduction in levels of smoking amongst parents between BC1 and BC2. In BC1, 28% of parents smoked. This figure had decreased to 24% in BC2 (see Table 7.22). This decrease is statistically significant. This reflects a wider trend showing a decline in smoking levels amongst the whole Scottish population over the same period^[22].

Table 7.22 Proportion of parents who smoke by cohort

	BC1	BC2
	%	%
Parent smokes	28	24
Unweighted bases	4171	4975

Tested on whether parent smokes: differences by cohort: p<0.01

Smoking rates varied according to household income with parents from lower income households more likely to smoke than parents in higher income households (Table 7.23). Amongst BC1 parents, 47% in the lowest income group smoked compared with 11% in the highest income group. Although the overall decrease in smoking levels is seen in all income groups, proportionate decreases are higher among those in the higher income groups. For example, smoking levels amongst parents in the lowest income quintile has fallen from 47% in BC1 to 40% in BC2 and amongst those in the highest income quintile from 11% in BC1 to 7% in BC2. There has however, been no statistically significant change in the strength of the relationship between household income and smoking levels between the cohorts.

Table 7.23 Parental smoking by equivalised household income (quintile) and cohort

	Lowest quintile	2nd quintile	3rd quintile	4th quintile	Highest quintile
	%	%	%	%	%
BC1	47	34	25	19	11
BC2	40	31	22	14	7
Unweighted bases - BC1	904	878	872	693	824
Unweighted bases - BC2	1006	1059	1018	924	935

Tested whether parent smokes: differences by equivalised household income p< .001; differences by cohort p< .001; cohort* equivalised household income p=NS.

Table 7.24 shows that there is an association between smoking and parental education. Parents with higher qualifications were less likely to smoke than those with lower or no qualifications. For example, amongst BC1 parents, nearly 6 in 10 (59%) parents with no qualifications smoked compared with nearly one in ten (9%) parents educated to degree level. This trend is also seen in BC2 parents where 53% of parents with no qualifications smoked compared with 8% of parents educated to degree level. Differences by cohort are not statistically significant.

Table 7.24 Parental smoking by parental level of education and cohort

	No qualifications	Lower Standard Grades or VQs or Other	Upper level SGs or Intermediate VQs	Higher grades and upper level VQs	Degree level academic and vocational qualifications
	%	%	%	%	%
BC1	59	49	47	26	9
BC2	53	50	40	26	8
Unweighted bases - BC1	195	204	799	1371	1595
Unweighted bases - BC2	155	211	709	1420	2311

Tested whether parent smokes: differences by education p< .001; differences by cohort p< .05; cohort* education p=NS.

Levels of parental smoking are related to maternal age (Table 7.25). Younger mothers were more likely to smoke than older mothers. In BC1, 59% of mothers aged 20 or under smoked compared with 33% of mothers aged 20-29 years old, 19% of mothers aged 30-39 years old and 14% of mothers aged 40 or older. The difference between smoking rates amongst the youngest and oldest mothers has reduced between cohorts mainly as a result of a considerable decline in the proportion of mothers aged under 20 smoking (59% in BC1 to 42% in BC2) relative to other age groups. Nevertheless, the considerable social gradient is such that there has therefore been no statistically significant change in the relationship between smoking and age between the cohorts.

Table 7.25 Parental smoking by maternal age at child's birth and cohort

	Under 20 years old	20 to 29 years old	30 to 39 years old	40 or older
	%	%	%	%
BC1	48.0	49.3	50.0	49.8
BC2	49.7	50.6	51.1	50.6
Unweighted bases - BC1	220	1555	2218	155
Unweighted bases - BC2	213	1955	2557	225

Tested whether parent smokes: differences by maternal age p< .001; differences by cohort p< .05; cohort* maternal age p=NS.

Table 7.26 shows levels of parenting smoking by area deprivation. Smoking was associated with deprivation in both cohorts. Parents in more deprived areas were more likely to smoke than those in less deprived areas. For example, in BC1 52% of parents living in areas in the most deprived are quintile smoked compared with 9% of parents in the least deprived area quintile. As we know levels of smoking amongst parents of three years olds has dropped between 2007/08 and 2013, with slightly greater reductions amongst parents living in the more deprived areas. However, there has been no notable nor statistically significant change in the relationship between area deprivation and smoking between the cohorts.

Table 7.26 Parental smoking by area deprivation (quintiles) and cohort

	1 Most deprived	2	3	4	5 Least deprived
	%	%	%	%	%
BC1	52	33	23	13	9
BC2	43	27	16	10	7
Unweighted bases - BC1	773	802	759	857	719
Unweighted bases - BC2	968	774	809	774	1020

Tested whether parent smokes: differences by area deprivation p< .001; differences by cohort p< .001; cohort* maternal age p=NS.