Growing up in Scotland: overweight obesity and activity

CHAPTER 4 PHYSICAL ACTIVITY AND SEDENTARY BEHAVIOUR

4.1 Key findings

• 15% of children exercised for less than the recommended level of 60 minutes daily, and 31% of children had 3 or more hours of screen time on a typical weekday.
• Children who were less physically active at age 6 were likely to have been less active when they were younger. Similarly, high sedentary behaviour as measured by screen time at age 6 reflected earlier high levels of screen use.
• However, it did not appear that high levels of screen time led to lower physical activity, or vice versa. The two behaviours appeared to have different sets of associated factors at age 6.
• Factors associated with low physical activity included winter season, child characteristics (being an only child, minority ethnic group), maternal characteristics (lower concern for and modelling of physical activity), neighbourhood characteristics (no swimming pool) and a less warm mother-child relationship.
• High screen time was more common in boys, in families with lower socio-economic status (low maternal education, mother smoking) and in neighbourhoods with poor quality local green spaces. High screen time was also associated with mothers' own screen use and a TV in the child's bedroom. General parenting was also involved, as screen use was greater in families where mothers did fewer activities with their children or had fewer rules for the child's behaviour.
• Being overweight or obese at age 4 was not a risk factor for either low physical actvity or high screen time at age 6.

4.2 Levels of physical activity and sedentary behaviour at age 6

According to mothers' reports, most children (85%) at age 6 met UK recommended targets for at least 60 minutes of moderate to vigorous physical activity every day (Table 4.1). 15% of the children were exercising for less than the recommended target . There was only a small increase in this figure (to 16%) when school-based activities were excluded. The proportion of children exercising for less than 60 minutes daily in the GUS data set was lower than found in the 2010 Scottish Health Survey ( SHeS). Taking school-based activity into account, the SHeS found that 27% of 5-7 year-olds exercised for less than the recommended amount. Excluding school-based actvities, this figure rose to 37% of 5-7 year olds (Marryat 2011) ¹³ .

There are no firm guidelines for sedentary behaviour in children, except that it should be limited. Most children (61%) watched TV and/or used a computer or games console for between 1 and 3 hours on a weekday, with just under a third using screens for 3 hours or more (Table 4.1). At the weekend, levels of children's screen use were higher, with around half the sample using screens for 3 hours or more on a Saturday. The SHeS did not have a measure of sedentary behaviour to compare with the GUS measure.

Table 4.1 Levels of physical activity and screen time in the age 6 sample

		%	Bases
			Wtd	Uwtd
Physical activity including school-based activity	60+ min on all 7 days	86	2512	2581
	30-59 min on all 7 days	8	230	225
	Lower level of activity	7	191	184
	Total	100	2933	2990
Physical activity excluding school-based activity	60+ min on all 7 days	83	2432	2497
	30-59 min on all 7 days	8	245	245
	Lower level of activity	8	243	236
	Total	100	2920	2978
Weekday screen time	Under 1 hour	8	190	212
	1 hour to under 2 hours	30	734	803
	2 hours to under 3 hours	31	744	757
	3 hours plus	31	763	698
	Total	100	2431	2470
Saturday screen time	Under 1 hour	8	198	201
	1 hour to under 2 hours	16	394	421
	2 hours to under 3 hours	24	575	608
	3 hours plus	52	1248	1228
	Total	100	2415	2458

Note: Due to rounding, percentages may not sum to 100.

4.3 Are sedentary behaviour and physical activity interrelated at age 6?

Children with higher levels of screen time were less likely to meet recommended targets for 60 minutes a day of moderate to vigorous physical activity at age 6 (see Figure 4.1 A and Figure 4.1 B). However, even among those children with screen time of 3 hours or more on a weekday, the majority (83%) still met the target.

Figure 4.1 Associations between weekday and Saturday screen time and physical activity level at age 6

A Weekday screen time

B Saturday screen time

4.4 Associations between activity levels and children's health

Physical activity did not vary systematically according to the child's general health (as reported by mothers at age 6), see Figure 4.2. However, children with poorer health had higher levels of screen time, see Figure 4.3.

Figure 4.2 Percentage of children with low (< 60 minutes daily) physical activity according to child's health at age 6

Figure 4.3 Percentage of children with high (3+ hours weekday) screen time according to their health at age 6

4.5 Associations between activity levels and season of the year

Children were physically active for longer periods in the second and third quarters of the year, when longer daylight hours permit more outdoor activity (Figure 4.4).

Use of screens showed less seasonal variation than physical activity, but was highest in the first, third and fourth quarters of the year (Figure 4.5). High screen time in the third quarter is likely to reflect increased leisure time during the school summer holidays in July and August.

Figure 4.4 Percentage of children with low (< 60 minutes daily) physical activity according to quarter of the year surveyed at age 6

Figure 4.5 Percentage of children with high (3+ hours weekday) screen time according to quarter of the year surveyed

4.6 Associations between age 6 activity levels and overweight or obesity

Current physical activity levels did not show clear associations with overweight or obesity at age 6 (Table 4.2). Further investigation with more sensitive measures (total minutes per week in all activities) also failed to find an association with overweight and obesity.

Screen time was not associated with overweight and obesity. Again, further investigation of weekend screen time allowing two further extended exposure categories of between 5 and 7 hours (13% of the sample) and 7 or more hours (5% of the sample) also failed to find an association with overweight and obesity.

Although there were no clear associations between either individual activity measure and overweight/obesity, it is possible that children who were less physically active AND used screens for long periods were most at risk. However, further investigation did not support the view that a combination of low activity and high sedentary behaviour, at age 6, is associated with greater risk of being overweight and/or obese (not shown here).

Table 4.2 Physical activity and sedentary behaviour at age 6: associations with age 6 overweight and obesity

		% Overweight	% Obese	Bases
				Wtd	Uwtd
Physical activity including school-based	60+ min on all 7 days	23	9	2512	2581
	30-59 min on all 7 days	20	7	230	225
	Lower level of activity	20	9	191	184
Physical activity excluding school-based	60+ min on all 7 days	23	9	2432	2497
	30-59 min on all 7 days	20	8	245	245
	Lower level of activity	21	9	243	236
Weekday screen time	Under 1 hour	19	8	190	212
	1 hour to under 2 hours	19	7	734	803
	2 hours to under 3 hours	24	9	744	757
	3 hours plus	25	11	763	698
Saturday screen time	Under 1 hour	18	8	198	201
	1 hour to under 2 hours	23	8	394	421
	2 hours to under 3 hours	20	9	575	608
	3 hours plus	24	10	1248	1228

Note: No differences between groups were statistically significant. Note that associations are measured separately for each factor and do not control for other influences.

A recent UK study found that being overweight led to lower physical activity in children (Metcalf et al. 2011). We also explored whether overweight and obesity at age 4 was associated with later physical activity and sedentary behaviour at age 6. We did not find that children in the GUS sample who were overweight or obese at age 4 adopted different activity patterns from their healthy weight peer group two years later (not shown).

4.7 Exploration of factors associated with age 6 physical activity and sedentary behaviour

This analysis focused on two outcomes:

1. Low physical activity (defined as not meeting recommended guidelines of at least 60 minutes every day, including school-based activity) ¹⁴

2. High weekday sedentary behaviour of 3 hours or more screen time ¹⁵ .

Associations explored in this section do not take account of possible confounding influences (other factors that simultaneously may play a role).

4.7.1 Child and family characteristics associated with physical activity and sedentary behaviour

Girls and boys were equally likely to meet physical activity targets at age 6, but boys were more likely than girls to spend longer periods using screens. Lower levels of physical activity were seen in children from minority ethnic groups, but there were no ethnic differences in screen time. Children without any siblings had lower physical activity than those from larger families. There appeared to be a U-shaped relationship between the number of children in the family and screen use. The only child, and those in larger families had greater levels of screen time than children with one or two siblings (Table 4.3).

Table 4.3 Child and family characteristics associated with physical activity and sedentary behaviour at age 6

		Not active for 60 mins/day (including school activity)				Screen time 3+hours weekday
		%	p	Bases		%	p	Bases
				Wtd	Uwtd			Wtd	Uwtd
Gender	Male	14	NS	1500	1519	34	**	1293	1306
	Female	14		1433	1471	28		1137	1164
Child's ethnic group	White	14	***	2812	2890	31	NS	2333	2390
	Minority	30		120	99	35		97	79
Children in household	One	19	*	584	551	35	*	482	454
	Two	13		1537	1642	29		1293	1373
	Three	14		621	635	31		507	519
	Four +	12		192	162	39		148	124
Birth order	First born	16	*	1471	1444	32	NS	1194	1162
	Other children	13		1462	1546	30		1236	1308

Notes: p=probability based on adjusted F where NS=not significant, † p<0.1, * p<0.05, ** p<0.01, *** p<0.001. Note that associations are measured separately for each factor and do not control for other influences.

4.7.2 Parental factors associated with physical activity and sedentary behaviour

Parental modelling

Parental modelling of specific behaviours appeared to be important. There were strong associations between mothers' physical activity at sweep 5 and children's physical activity at age 6. There were also strong associations between between mothers' screen time at sweep 5 and children's screen time at age 6 (Table 4.4). However, mothers' screen time did not predict children's physical activity, nor conversely did mothers' physical activity predict children's screen time.

Mothers' overweight or obesity at age 6 was not associated with children's physical activity or sedentary behaviour (not shown).

Table 4.4 Associations between parental modelling and children's physical activity and sedentary behaviour at age 6

		Not active for 60 mins/day (including school activity)				Screen time 3+hours weekday
		%	p	Bases		%	p	Bases
				Wtd	Uwtd			Wtd	Uwtd
Mother's weekly physical activity sweep 5 (quintiles)	q1 - Bottom 20%	21	***	550	584	31	NS	462	484
	q2	17		623	646	29		512	533
	q3	11		452	478	28		369	391
	q4	10		528	541	28		445	455
	q5 - Top 20%	9		574	552	35		472	448
Mother's weekly screen time sweep 5 (tertiles)	Low	13	NS	926	1005	21	***	724	784
	Medium	13		978	1021	27		818	856
	High	17		1027	961	43		885	827

Notes: p=probability based on adjusted F where NS=not significant, † p<0.1, * p<0.05, ** p<0.01, *** p<0.001. Note that associations are measured separately for each factor and do not control for other influences.

Health-related practices

Earlier measures of children's physical activity gathered at previous sweeps were associated with greater physical activity at age 6, but not with screen time at age 6. Frequency of visiting local green spaces (measured at age 6) was associated with both higher children's physical activity and reduced screen time (Table 4.5).

Children who had a TV in their bedroom at earlier sweeps and who spent a lot of time watching TV and/or playing with a computer or games console at earlier sweeps were more likely to use screens for 3 hours or more at age 6. However, there were no clear associations between earlier screen time and age 6 physical activity (Table 4.6).

Table 4.5 Associations between health practices related to active behaviour and children's physical activity and sedentary behaviour at age 6

		Not active for 60 mins/day (including school activity)				Screen time 3+hours weekday
		%	p	Bases		%	p	Bases
				Wtd	Uwtd			Wtd	Uwtd
Weekly duration child's active behaviour (sweep 3)	Less than 1 hr	41	***	58	55	36	NS	46	44
	1 to under 3 hrs	25		251	253	35		211	209
	3 to under 5 hrs	21		544	537	36		454	445
	5 to under 7 hrs	15		604	620	29		505	519
	7 hrs plus	9		1446	1495	30		1191	1230
Physical activity level sweep 5 (quintiles)	q1 - Lowest 20%	28	***	60 1	593	35	NS	507	503
	q2	19		557	565	33		462	467
	q3	11		598	616	31		512	519
	q4	8		592	606	27		476	490
	q5 - Top 20%	6		579	601	29		469	484
Frequency mother and child visit local green spaces/month (age 6)	Every day	8	***	488	522	27	**	391	421
	Several times a week	12		747	784	30		615	644
	Once a week	16		569	578	30		478	482
	Several times a month	13		515	523	30		442	443
	Once a month	19		266	263	36		216	216
	Less often	22		217	200	40		179	163
	Not at all	27		130	119	45		109	100

Notes: p=probability based on adjusted F where NS=not significant, † p<0.1, * p<0.05, ** p<0.01, *** p<0.001. Note that associations are measured separately for each factor and do not control for other influences.

Table 4.6 Associations between earlier screen use and its related factors and children's physical activity and sedentary behaviour at age 6

		Not active for 60 mins/day (including school activity)				Screen time 3+ hours weekday
		%	p	Bases		%	p	Bases
				Wtd	Uwtd			Wtd	Uwtd
TV in bedroom (sweep 5)	Yes	14	NS	1511	1367	40	***	1288	1169
	No	15		1413	1612	21		1136	1294
TV in bedroom (age 4)	Yes	15	NS	1331	1179	41	***	1118	996
	No	14		1572	1782	23		1293	1458
Hours watching TV on weekday (sweep 3)	None	10	NS	15	18	<1	***	11	13
	Up to 30 minutes	14		299	311	23		229	236
	30 minutes to 1 hour	11		464	491	22		365	387
	1 to 2 hours	15		1015	1063	25		849	882
	2 to 3 hours	16		692	689	40		595	595
	3 to 4 hours	15		188	180	49		169	163
	4 to 5 hours	16		82	71	59		72	65
	5 hours or more	11		69	57	53		64	54
Hours watching TV on weekday (age 4)	None	13	NS	152	165	17	***	111	119
	Up to 30 minutes	11		78	84	11		59	63
	30 minutes to 1 hour	10		155	170	7		114	126
	1 to 2 hours	14		1120	1189	22		917	972
	2 to 3 hours	15		841	851	38		715	729
	3 to 4 hours	16		318	296	46		291	267
	4 to 5 hours	24		130	115	63		112	102
	5 hours or more	17		104	86	57		88	71
Screen time weekday (sweep 5)	Under 1 hour	11	NS	279	314	7	***	188	211
	1 to under 2 hours	14		875	955	14		684	746
	2 to under 3 hours	16		903	917	30		776	791
	3 to under 5 hours	15		691	656	50		618	590
	5 hours or more	11		157	125	65		145	116

Notes: p=probability based on adjusted F where NS=not significant, † p<0.1, * p<0.05, ** p<0.01, *** p<0.001. Note that associations are measured separately for each factor and do not control for other influences.

General patterns of parenting

Aspects of connection, control and conflict were each associated with screen time at age 6. Connection and conflict were associated with physical activity (Table 4.7).

Mothers who had a warm relationship with their child, who engaged in many joint activities and who experienced low levels of conflict were more likely to have physically active children with relatively low levels of screen use. In addition, mothers who had rules to guide their children's behaviour and a more organised home environment were less likely to have children who spent long periods watching TV or playing electronic games.

Parental supervision did not have a clear association with physical activity, although there was a trend for low supervision to be associated with higher levels of screen use.

Table 4.7 Associations between general patterns of parenting and children's physical activity and sedentary behaviour at age 6

			Not active for 60 mins/day (including school activity)				Screen time 3+ hours weekday
			%	p	Bases		%	p	Bases
					Wtd	Uwtd			Wtd	Uwtd
Connection	Pianta warmth (quartiles) a	q1 - low	17	**	730	686	38	***	609	561
		q2 and 3	16		871	896	33		725	743
		q4 - high	12		1314	1392	26		1084	1154
	Mother-child joint activities (quartiles)	q1 - low	16	**	770	721	43	***	620	579
		q2	20		291	290	29		248	247
		q3	15		1089	1109	30		900	910
		q4 - high	10		783	870	23		663	734
Control	Supervision (quintiles) a	q1 - low	17	NS	568	534	36	?	471	445
		q2	14		921	906	31		758	741
		q3 and 4	14		1015	1065	31		853	897
		q5 - high	12		430	485	26		348	387
	Number of rules (quintiles) a	q1 and q2 - low	15	NS	1055	1017	40	***	897	867
		q3	15		759	788	30		613	633
		q4	13		846	904	25		694	741
		q5 - high	13		274	281	20		226	229
	Home chaos (quintiles)	q1 - low	15	NS	766	820	26	***	639	681
		q2	16		676	701	31		558	578
		q3	14		474	490	29		393	406
		q4	12		628	622	32		518	511
		q5 - high	16		390	357	46		323	294
Conflict	Pianta conflict (quintiles)	q1 - low	13	**	751	798	27	***	634	668
		q2	17		409	425	27		328	350
		q3	11		581	600	27		485	500
		q4	13		633	637	34		521	521
		q5 - high	18		540	513	42		449	419

Notes: p=probability based on adjusted F where NS=not significant, † p<0.1, * p<0.05, ** p<0.01, *** p<0.001. Note that associations are measured separately for each factor and do not control for other influences. ^aDue to lack of subdivisions in measures, some percentiles are combined.

4.7.3 Distal family constraints on children's physical activity and sedentary behaviour

There appeared to be no clear social inequalities in the patterning of children's physical activity, according to the distal family constraints examined. There was a trend (bordering on statistical significance) for families on lower incomes and families where the mother had not worked to have lower physical activity levels.

In contrast, children's screen time varied strongly by factors at the family level. These suggest social inequalities in the patterning of children's sedentary behaviour. Children were more likely to use screens for 3 or more hours if their mother did not work, had a relatively low educational level, was a lone parent and had lower levels of household income. Children were also more likely to use screens for longer if their mother was in poor mental or physical health, or if their mother smoked when pregnant. However, mother's working hours were not clearly associated with greater screen use (Table 4.8).

Table 4.8 Family distal constraints associated with children's physical activity and sedentary behaviour at age 6

		Not active for 60 mins/day (including school activity)				Screen time 3+ hours weekday
		%	p	Bases		%	p	Bases
				Wtd	Uwtd			Wtd	Uwtd
Maternal employment per week (sweep 5)	35 hrs or more	11	NS	446	498	25	***	367	409
	16 to less than 35 hrs	14		1183	1271	27		994	1064
	Less than 16 hrs	14		307	351	24		251	284
	Not working	16		993	865	42		814	708
Mother's education	Up to Standard Grades or equivalent	15	NS	1102	934	44	***	923	782
	Higher Grades or equivalent	15		649	669	31		536	555
	Upper level VQs ( e.g. HNC, HND)	12		370	400	25		316	340
	Degree or equivalent	13		805	980	17		648	787
Biological father in household	No	14	NS	849	639	43	***	682	507
	Yes	14		2085	2351	27		1749	1963
Household income (quintiles)	Top 60%	13	?	1499	1784	21	***	1242	1476
	Bottom 40%	15		1418	1192	40		1176	983
Mother's occupational class	Managerial and professional occupations	13	?	1066	1269	19	***	872	1035
	Intermediate occupations	16		570	604	30		475	507
	Small employers and own account workers	12		118	134	24		97	110
	Lower supervisory and technical occupations	9		163	155	36		137	127
	Semi-routine and routine occupations	16		883	746	45		734	623
	Never worked	22		132	81	47		115	67
Mother's age at birth of first child	Under 20	15	NS	518	360	49	***	431	298
	20 to 29	15		1504	1496	31		1251	1242
	30 to 39	14		885	1099	22		728	903
	40 or over	16		27	35	30		21	27
Mother's mental health	Poor	16	NS	1001	987	35	*	842	823
	Average	14		944	997	29		769	815
	Good	13		979	1000	30		812	827
Mother's physical health	Poor	15	NS	995	980	35	*	814	797
	Average	15		1003	1022	29		839	860
	Good	13		927	982	30		771	808
Mother's smoking in pregnancy	No	14	NS	2221	2395	26	***	1834	1968
	Yes	15		702	585	47		588	494

Notes: p=probability based on adjusted F where NS=not significant, † p<0.1, * p<0.05, ** p<0.01, *** p<0.001. Note that associations are measured separately for each factor and do not control for other influences.

4.7.4 Proximal family constraints

We explored two family level constraints that might affect how mothers managed their child's physical activity and sedentary behaviour. These were mothers' awareness of current guidelines ("target levels") for physical activity in children, and mothers' concerns about their child's weight.

Mothers' views on desirable amount of physical activity in children

At age 6, mothers were asked two questions about the desirable amount of physical activity for children:

• "How many days a week do you think children of [childname]'s age should do physical activity? Please include all moderate physical activity, including physical activity at school."
• "On each of the days someone of [childname]'s age does moderate physical activity, how many minutes a day should they do it for it to be good for their health?"

These questions were asked after questions on the child's actual physical activity had been completed. This helped to avoid the possibility of bias in the mothers' responses to the physical activity measures, after raising the issue of desirable levels of physical activity.

Although most (84%) mothers thought their children should take moderate to vigorous physical activity every day, many mothers (57%) did not think children needed to exercise for at least 60 minutes or more each day. ¹⁶

Views on desirable levels of physical activity were not associated with mothers' education or occupational class. Mothers who thought children should exercise for at least 60 minutes daily were likely to exercise more themselves, as measured at sweep 5. Mothers with an overweight child and who did not recognise the child's overweight were less likely to think at least an hour of daily physical activity was desirable.

Mothers' views were associated with whether their child met the physical activity target at age 6, but not with their child's sedentary behaviour (screen time):

• 9% of children failed the target when their mother thought at least 60 minutes daily was desirable, compared to 18% when mothers thought less activity was sufficient
• levels of screen time of 3 hours or more on a week day were almost identical whether mothers gave a 60-minute target level for physical activity (32%) or not (31%)

Mothers' concerns about their child's weight

At sweeps 4 and 6, mothers were asked: "How concerned are you about [childname]'s weight at the moment?" (with answers on a four-point scale of "not at all", "a little concerned", "quite concerned" and "very concerned"). Mothers were divided into those who were not at all concerned, and those who expressed some concern.

The GUS data do not appear to support the idea that mothers who are more concerned about a child's weight will encourage their child to exercise more and limit sedentary time watching television or using a computer.

There was no clear association between mothers' concern for their child's weight (measured at age 4 or age 6) and children's physical activity level at age 6. Mothers' concern was associated with greater sedentary behaviour: 42% of children with concerned mothers used screens for 3 or more hours, compared to 30% of children with unconcerned mothers ( p<0.001). A similar effect was seen with mother's concern measured at age 4. These findings suggest that the child's weight and/or sedentary behaviour may have led to mothers' concern, rather than mothers' concern succeeding in limiting the child's sedentary behaviour.

4.7.5 Distal neighbourhood constraints on children's physical activity and sedentary behaviour

Children had higher levels of screen use if they lived in an urban or small town setting rather than a rural setting, and if they lived in a deprived neighbourhood. However, meeting recommended physical activity levels was not clearly associated with these neighbourhood factors (Table 4.9).

Table 4.9 Distal neighbourhood constraints associated with children's physical activity and sedentary behaviour at age 6

		Not active for 60 mins/day (including school activity)				Screen time 3+ hours weekday
		%	p	Bases		%	p	Bases
				Wtd	Uwtd			Wtd	Uwtd
Urban-rural	Urban	15	NS	2027	1937	33	*	1693	1612
	Small town	12		370	407	33		307	339
	Rural	15		536	645	25		430	518
Area deprivation quintile	1 - least deprived	12	NS	576	697	20	***	482	578
	2	16		615	706	26		496	571
	3	14		565	617	29		461	506
	4	12		551	510	38		468	430
	5 - most deprived	17		627	459	43		524	384

Notes: p=probability based on adjusted F where NS=not significant, † p<0.1, * p<0.05, ** p<0.01, *** p<0.001. Note that associations are measured separately for each factor and do not control for other influences.

4.7.6 Proximal neighbourhood constraints on children's physical activity and sedentary behaviour

Social and physical aspects of the neighbourhood environment were explored in relation to children's physical activity and screen use at age 6.

Lower "child-friendliness", in terms of mothers' ratings at sweep 3 of the local social and physical environment for children, was associated with failing to meet the physical activity target and with greater screen time at age 6. However, child friendliness measured at age 6 did not show a clear asssociation with meeting the physical activity target at this age.

Some associations between access to local facilities and activity levels were found:

• access to a playground or park (sweep 2) and access to a pool at age 6 (but not sweep 2) were associated with age 6 physical activity
• access to a garden ¹⁷ was associated with lower screen time, but was not clearly associated with physical activity

Walking time to nearest green space (measured at age 6) was not associated either with physical activity or sedentary behaviour (not shown).

Greater screen time (but not physical activity) was also associated with a lower quality neighbourhood environment, in terms of:

• poorer quality of local green places
• concerns over safety of neighbourhood and local play areas
• greater levels of antisocial behaviour

Table 4.10 Proximal neighbourhood constraints associated with children's physical activity and sedentary behaviour at age 6

		Not active for 60 mins/day (including school activity)				Screen time 3+ hours weekday
		%	p	Bases		%	p	Bases
				Wtd	Uwtd			Wtd	Uwtd
Garden (sweep 1)	Yes	14	NS	2653	2767	30	***	2194	2283
	No	18		280	223	44		237	187
Playground/park (sweep 2)	Not in area	22	*	191	180	33	NS	155	148
	In area	14		2732	2802	31		2267	2316
Swimming pool locally (sweep 6)	Not in area	18	**	740	750	34	NS	617	625
	In area	13		2191	2238	31		1812	1844
Rating of local green places to walk/sit (sweep 6)	1 - highest	13	NS	1007	1139	23	***	809	912
	2	14		635	672	29		540	568
	3	17		596	582	37		508	496
	4	14		306	258	37		250	211
	5	13		161	150	45		131	122
	6	12		95	80	46		86	73
	7 - lowest	22		121	97	42		99	80
Rating of safe places for children to play locally (sweep 6)	1 - highest	12	NS	715	808	23	***	566	636
	2	16		641	695	26		534	579
	3	15		674	684	33		576	581
	4	14		379	344	40		333	301
	5	11		210	182	39		169	150
	6	16		136	121	43		120	105
	7 - lowest	21		166	144	40		126	111
Neighbourhood safety (sweep 3)	High	14	NS	1885	2020	29	***	1561	1663
	Low	16		1048	969	36		869	806
Neighbourhood safety (sweep 6)	High	14	?	1470	1604	26	***	1209	1315
	Medium	13		692	697	35		564	569
	Low	17		772	689	38		657	586
Safe public play areas locally (sweep 3)	Yes	14	NS	2020	2180	27	***	1655	1780
	No	16		885	783	40		751	667
Neighbourhood antisocial behaviour problems (age 6)	None	15	?	1501	1636	28	***	1233	1343
	1	11		625	640	29		518	531
	2	17		295	296	37		245	244
	3	19		197	183	43		174	159
	4 or more	12		312	233	41		258	191
Child friendly neighbourhood (sweep 3)	Good	13	*	1100	1236	27	***	906	1014
	Average	12		760	795	30		641	667
	Poor	17		1073	958	37		883	788
Child friendly neighbourhood (age 6)	Good	14	NS	965	1082	26	***	784	875
	Average	14		989	1042	31		846	888
	Poor	15		979	866	37		800	707

Notes: p=probability based on adjusted F where NS=not significant, † p<0.1, * p<0.05, ** p<0.01, *** p<0.001. Note that associations are measured separately for each factor and do not control for other influences. Bases: WTD = weighted, UWTD = unweighted.

4.8 Multivariate modelling of children's physical activity and sedentary behaviour at age 6

This section reports the main findings of multivariate analyses of children's low physical activity (below 60 minutes daily) and high sedentary behaviour (3 hours or more screen time on a weekday) at age 6.

Logistic regression models first investigated selected blocks of variables, before combining all factors in the final model. All models controlled for the child's gender, ethnic group, the number of children in the family, the child's general health and the season of the year when age 6 interviews took place. Detailed modelling results are provided in the Technical Appendix.

In the final model of physical activity, the following factors were significantly associated with the child exercising for less than 60 minutes daily:

• ethnic minority group
• being an only child
• winter season (October-March)
• no swimming pool in area
• mother's view that less than 60 minutes' daily physical activity was sufficient for children
• mother's lower level of physical activity
• less warmth in the mother-child relationship

In the final model of screen time, the following factors were significantly associated with the child using screens for three hours or more on a weekday:

• male gender
• low maternal education
• mother smoking in pregnancy
• poor quality local green spaces
• mother's greater screen time
• TV in the child's bedroom
• Fewer joint mother-child activities
• Fewer rules for the child's behaviour

These findings are summarised in Figure 4.6.

Figure 4.6 Summary of risk factors for children's low physical activity and high screen time at age 6: results of final multivariate modelling

Note: Dark shaded areas indicate associations with low physical activity and high screen time. Model adjusted for all factors together, see Technical Appendix for further details.

Parents who do not recognise that their child is overweight or obese are unlikely to help the child to lose weight. Research elsewhere has found that few parents seem to recognise overweight in their child (Parry et al. 2008; Jones et al. 2011a).

For the most part, this section focuses on the subgroup of GUS children who were defined as overweight or obese on the basis of their BMI at age 6. It explores the extent to which mothers recognised these children as overweight, and investigates factors associated with poor recognition and low levels of concern about children's overweight.