Comparison of the Danish Physical Activity Questionnaire with a validated position and motion instrument

Abstract

Objective To compare the Danish Physical Activity Questionnaire (DPAQ) estimating physical activity energy expenditure (PAEE) and physical activity level (PAL) and the pattern of physical activity (including health-related physical activity) with measurements from a validated position and motion instrument (ActiReg^®). Methods One hundred and thirty-eight healthy Danish volunteers aged 20–59 years participated. Participants filled in the DPAQ at the end of the day every day for seven consecutive days and carried the ActiReg^® on the same days as the DPAQ were filled in. Results A small, but statistically significant difference was seen between the DPAQ and the ActiReg^® for PAEE (mean: −0.32 MJ·d⁻¹; 95% limits of agreement: ((−2.88)–2.24 MJ·d⁻¹); P = 0.003) and PAL (−0.03; ((−0.37)–0.31); P = 0.02) for the whole group and for women (P < 0.008 for PAEE and PAL), but not for men. The correlation between methods was good for PAEE (r = 0.71, P < 0.001) and PAL (r = 0.64, P < 0.001). No difference was observed for time spent in moderate plus vigorous physical activity (MVPA) (P = 0.40). Time reported in MVPA with the DPAQ was positively correlated with time spent in MVPA measured by the ActiReg^® (ρ = 0.53, P < 0.001). Conclusions Although the volunteer sample may influence the representativeness of the results, the DPAQ provided a close estimate of PAEE, PAL (2–5% underestimation) and health-related physical activity (MVPA) in healthy adults at group level. The results indicate that the questionnaire can be used to rank individuals according to energy expenditure and level of total physical activity and to provide information on health-related physical activity.

Appendix I

Self-registered time in the DPAQ

Mean registered time per day in the DPAQ was 22.8 h (SD 2.1 h) and 67% of all reporting days ranged between 22 and 26 h per day. No difference in mean registered time per day was seen between genders (P = 0.87). Fifty-five percent of all the participants reported being physical active as usual during the data collection period, while 32% reported being a little less active than normal.

Procedure for correcting self-registered time to 24 h per day

If self-registered time including sleep did not add up to 24 h per day, the time was corrected to 24 h per day by adding missing time or subtracting over-reported time. When self-registered time was higher than 24 h per day, potential double registrations were controlled for relevant activities in the context for which they may be undertaken, e.g. walking as transport or as leisure-time activity. Over-reported time was corrected for double registered time, and then for time spent sleeping, as this was an average value. If this was not sufficient to reach 24 h per day, a down adjustment of all the registered time was made except for occupational activity as total working time was assumed to be the most reliable estimate of all in the DPAQ.

MET values assigned to each self-reported physical activity category in the DPAQ. METs were obtained from international tables [7] and from a compendium of physical activities [10, 11]

Activity domain	Activity type/level	Self-reported pace	MET
Occupational	Vigorous		8
	Moderate		4
	Light		2.3
	Sitting		1.3
	Walking/cycling	Vigorous	5
		Moderate	3.3
		Slow	2.5
Transport	Sitting		1.3
	Cycling	Vigorous	8
		Moderate	6
		Slow	4
	Walking	Vigorous	5
		Moderate	3.3
		Slow	2.5
Yard/garden	Vigorous		5.5
	Moderate		4.0
	Light		2.5
Household	Vigorous		4.0
	Moderate		3.0
	Light		2.3
Leisure	Walking	Vigorous	5
		Moderate	3.3
		Slow	2.5
	Vigorous		8.0
	Moderate		4.0
	Light		2.5
Sitting leisure	Home		1.3
	Outside home		1.3
Sleep			1.0
Missing timea			1.6

1. ^aRepresents a time-weighted average of typical everyday life activities like sitting, light household chores, self-care, and social and cultural activities