SENECA_PA_SPORT

Creation rule: Assign median value of range in each category of time to all subjects belonging to this category (for last category, assign maximum of [median value of those in EPIC Elderly, Esther and Zutphen with >=4 hrs/week] 4.5). Estimate for each subject the total hrs/week in all sports. The original variables were based on intensity, time (hrs per week) and proportion (months per year).

Contact round 1:

PA_SPORT is based on SENECA variables HR1SPORT (“Hours of first sport”), HR2SPORT (“Hours of second sport”), and ACT1SPRT (“First physical active sport”)

IF HR1SPORT IN(88,99,.) THEN HR1SPORT=. ;

IF HR2SPORT IN(88,99,.) THEN HR2SPORT=. ;

IF ACT1SPRT IN(8,9,.) THEN HRSPORT=.;

IF ACT1SPRT = 0 THEN HRSPORT = 0;

IF ACT1SPRT = 1 THEN HRSPORT = SUM(HR1SPORT, HR2SPORT);

IF HRSPORT =. THEN PA_SPORT=.;

ELSE PA_SPORT=(0.1 * HRSPORT);

Contact round 2:

PA_SPORT is based on SENECA variables FHR1SPOR (“Hours of first sport”), FHR2SPOR (“Hours of second sport”), and FACT1SPR (“First physical active sport”)

IF FHR1SPOR IN(8.8,9.9,.) THEN FHR1SPOR=. ;

IF FHR2SPOR IN(8.8,9.9,.) THEN FHR2SPOR=. ;

IF FACT1SPR IN(8,9,.) THEN HRSPORT=.;

IF FACT1SPR = 0 THEN HRSPORT = 0;

IF FACT1SPR = 1 THEN HRSPORT = SUM(FHR1SPOR, FHR2SPOR);

IF HRSPORT =. THEN PA_SPORT=.;

ELSE PA_SPORT=(HRSPORT);

Contact round 3:

PA_SPORT is based on SENECA variables SHR1SPOR (“Hours of first sport”), SHR2SPOR (“Hours of second sport”), and SACT1SPR (“First physical active sport”)

IF SACT1SPR = 0 THEN HRSPORT = 0;

IF SACT1SPR = 1 THEN HRSPORT = SUM(SHR1SPOR, SHR2SPOR);

IF HRSPORT =. THEN PA_SPORT=.;

ELSE PA_SPORT=(HRSPORT);

CR_MORGAM_PRIME Belfast_PA_SPORT

This variable is not harmonized in MORGAM, but some cohorts can convert it from their local data.

PRIME Belfast (UNK-BEL)

The PRIME Belfast variable “SPORTHOR: When you did this sport or exercise, how much time did you usually spend on each session?” was converted from minutes to hours by dividing by 60. Then in conjunction with the categorical variable “SPORTWEK: In the months when you did this sport or exercise, how many times per week did you usually do it?” we calculated hours/week spent playing this sport in total. Those who replied to SPORTWEK with “0=less than once per week” had their hours/week spent playing sport in total halved.

Similar conversion rules were applied to “OTHSPHOR” and “OSPORTWEK” which are coded similar to above but covers time spent doing other sport.

PA_SPORT was coded by adding these two converted variables together: hours/week spent playing sport and other sport.

ESTHER_PA_SPORT

pa_sport = L47BTNUS;
if pa_sport = . then pa_sport = 9999;

Source of L47BTNUS is participant's questionnaire at baseline
Includes not only sport but also other work that caused sweating (also at the workplace)

EPIC_PA_SPORT

EPIC variable M_SPORT was used for the production of PA_SPORT.

EPIC var definition(s):
M_SPORT = Epic elderly Greece variable showing MET hours spent in sports (both Winter and Summer)

Coding rule: 
Variable PA_SPORT is produced using the following formula: M_SPORT/6

Coding rule in STATA:
-gen PA_SPORT = M_SPORT/6

Comment: Variable M_SPORT was constructed as the mean of MS_SPORT and MW_SPORT (MS for summer sports and MW for winter sports and correspond the to METS for the hours per week spent in each sport reported). The initial EPIC variables (HS_SPORT and HW_SPORT) corresponding to hours per week spent in sports were then multiplied by 6 (taking into account the average met hours value for all sports).