Reliability and validity of a Global Physical Activity Questionnaire adapted for use among pregnant women in Nepal

Study design

This prospective cohort study recruited 101 pregnant women attending the Obstetric Outpatient Department (OPD) at Dhulikhel Hospital in DhulikhelNepal, for antenatal care (ANC) between January 2019 and January 2020. Women were enrolled during their 1st trimester ANC visit (5-14 weeks of gestation) and followed through the 2nd and 3rd trimester, up until six weeks postpartum. The primary aim of the study was to adapt and validate dietary and PA assessment tools for use among Nepalese pregnant women. With respect to PA, the study aimed to evaluate the reliability and validity of a culturally adapted version of the 16-item GPAQ [10]. Because PA may vary dramatically during pregnancy due to physiological changes, we assessed its reliability at two-time points in the 2nd and 3rd trimester of pregnancy. Accelerometer data from the 2nd trimester was used for validation as pregnant women are more likely to be active during this time [9].

To participate, women had to be 18 years or older, ≤14 weeks of gestation at enrollment, and carrying a single fetus. Additionally, to participate in the validation study with the accelerometer in the 2nd trimester, the participant had to agree to wear the accelerometer on their wrist for seven consecutive days; women with contraindications to exercise were not eligible for the validation study. At enrollment and subsequent ANC visits, a trained research assistant collected data from each participant via medical record review and patient interview. Data was collected on socio-demographic, lifestyle, and clinical characteristics. All participants provided written informed consent. The Kathmandu University Ethical Review Board (102/18) and the Rutgers Newark Health Sciences Institutional Review Board (Pro2018001976) approved the study protocol.

GPAQ adaptation for the target population

The GPAQ includes 16 questions which gather information about SB as well as levels of PA across five settings. These include PA related to work/domestic/occupational tasks (vigorous and moderate), travel/transportation such as walking or biking, and leisure/recreational activities (vigorous and moderate) [9, 10, 14,15,16]. Vigorous activity is defined as an activity that causes large increases in breathing or heart rate while moderate activity causes small increases in heart rate.

For this study, a modified GPAQ was developed by translating the questionnaire into Nepali and creating a PA chart that included examples specific to the target demographic of Nepali pregnant women. For PA related to work/domestic tasks, examples for vigorous activity included construction work and carrying/stacking heavy loads (i.e., bricks), whereas examples for moderate activity included farm work, drawing water from well, and carrying light loads. For leisure/recreational PA, examples for vigorous activity included running and hiking/trekking, while examples for moderate activity included swimming or yoga. For each of the five PA domains (work-related vigorous, work-related moderate, travel, recreational vigorous, and recreational moderate), the GPAQ asks participants to report the time duration and frequency (number of days per week) of these activities during a typical week. To ascertain SB, participants are asked to report the amount of time they spent sitting per typical day (i.e., watching television/lounging on the couch/sitting at work). To adapt the questionnaire for use in pregnancy, we further specified the timeframe so that it refers to the PA during a typical week in the respective trimester (2nd or 3rd). The questionnaire was back translated in order to ensure fidelity between the English and Nepali versions. The Nepali version of the questionnaire was also cognitively tested among 8 pregnant volunteers to ensure its suitability in capturing the usual types and duration of physical activity among our target population.

GPAQ administration

To determine the trimester-specific reliability of the GPAQ, the questionnaire was administered to participants at two time-points, one month apart, in the 2nd and 3rd trimesters. Despite being administered a month apart, at both time points, the questions on the GPAQ were directed to ascertain PA and SB during a typical week in the respective trimester (2nd or 3rd trimester). Responses on the GPAQ were used to determine the duration of SB (min/day), as well as PA (minutes/week) for each of the five domains (work-related vigorous, work-related moderate, travel, recreational vigorous, and recreational moderate). Total moderate or vigorous PA (MVPA in min/week) was calculated by adding time spent in minutes/week across all five domains.

The GPAQ also allows for the computation of energy expenditure, recorded in metabolic equivalent tasks (METs). A standard value of 4 METs was assigned for moderate-intensity activities related to work/recreation/travel; 8 METs was assigned for work-or leisure-related vigorous activity [21]. According to the GPAQ responses with respect to duration, intensity, and frequency of activity, total MVPA was also calculated in MET-min/week. Total MVPA level was also classified as low, moderate, and high based on WHO guidelines (Fig. 1) [21].

Fig. 1figure 1

Classification of PA Level based on WHO Guidelines. Shows the specific criteria for low, moderate, and high physical activity classifications based on the World Health Organization standards. Abbreviations: PA; physical activity, MET; metabolic equivalent

Accelerometer

Data collected from the GPAQ was validated against PA data collected by a triaxial accelerometer (Axivity AX3; UK). The device captured activity based on changes in acceleration using a three-dimensional (3D) acceleration sensor, providing a more accurate and objective calculation of energy expenditure as compared to uniaxial devices and self-reported PA [22]. A subset of the subjects (n=34) who agreed to take part in the validation study wore the Axivity accelerometer on their left wrist for seven consecutive days during their 2nd trimester.

The accelerometer was set to record 3D acceleration data at a sample rate of 50Hz with a dynamic range of ±8g. The activity classification and non-wear time for each participant was extracted using the UK Biobank accelerometer analysis tool [23]. Non-wear time was defined as at least 60 consecutive minutes of stationary episodes with a standard deviation of less than 13.0mg and was imputed with data from similar time-of-day vector magnitude. For this analysis, MVPA was defined using a 100mg cut-off, as is done traditionally; and average daily time in MVPA (min/day) and SB (min/day) were calculated. All accelerometer data analysis was performed in Python. A minimum of 72 hours of continuous and valid wear time was necessary for inclusion in the analysis; 21 of 34 participants were included in the final analyses [23].

Statistical analysis

Descriptive statistics are presented as mean ±SD or median (25th-75th percentiles) for continuous variables, and frequencies (n, %) for categorical variables. The reliability of the GPAQ between the two visits within each trimester was tested by intra-class correlation coefficients (ICC) and Spearman's correlation coefficients (rho) for continuous variables including PA measures in each of the five domains (min/week), total MVPA (min/week), and SB (min/day). With respect to the level of PA (i.e. high, moderate, or low), the degree of consistency between the two GPAQ administrations in each trimester was evaluated using the Gwet's agreement coefficient (AC1). Gwet's AC1 was used because a skewed distribution was observed and the AC1 statistic overcomes the noted limitations of the kappa statistic and provides a more precise, stable agreement statistic [24]. For validation analyses comparing the average GPAQ measures in the 2nd trimester with those from the accelerometer, we used Spearman's correlation coefficients (rho) for continuous variables including total MVPA and SB.

ICC and Gwet's AC1 analyses were interpreted based the Cicchetti and Sparrow (1981) criteria [25]: <0.40 poor, 0.41-0.59 fair/moderate, 0.60-0.74 good and >0.75 excellent reliability. To interpret Spearman’s rho, the following criteria were used: 0-0.20 poor, 0.21-0.40 fair, 0.41-0.60 moderate/acceptable, 0.61-0.80 good, and 0.81-1.00 strong correlation [26].

留言 (0)

沒有登入
gif