Our search yielded 65 hits on PubMed, 1 study from AJO, and 14 on Google Scholar. Following an elaborate assessment, approximately 9 relevant articles were identified, and only 2 studies carried out on 453 (n = 102 IPAQ-SF Hausa & n = 351 for Yoruba) apparently healthy adult subjects who met the criteria for inclusion were documented (Fig. 1). Additionally, these studies were conducted in Africa (Nigeria).
The study by Awotidebe et al. [2021], conducted in Obafemi Awolowo University (OAU) Ile-Ife, Osun state, Southwest, Nigeria, purposively sampled a cohort of undergraduate students (age 22.4 ± 3.01 years, mostly male and singles) culturally adapted the IPAQ-SF into the Yoruba language. The prefinal version was tested on 15 subjects (mean age = 21.3 ± 4.7 years; female: n = 6; male: n = 9) who were recruited for the pilot study, after which the final version (Yoruba) was developed and used to collect data in the main study. Additionally, based on the outcome of the pilot, some modifications were made; for example, the word “digging” translated to “gbígbé ̣kòtò”, “vigorous aerobics and PA was replaced with “hard running and PA”, and “serious farm work” and “cutting grass around the house and fetching of water” were also included in the moderate and hard PAs. In contrast, the study by Oyeyemi et al. [2011], conducted at the University of Maiduguri, Borno state, northeastern Nigeria, sampled subjects (age = 36.2 ± 9.5 years, rage = 20–65 years, primarily males, and married) and had higher total PA than females__ 4383.7 vs 3058.1 MET-Min week-1 respectively, and mean rate pressure product, RPP = 9694.3 ± 1195.94) from a different background__ workplace (e.g., private establishments and university teaching hospital) and in Maiduguri city neighborhoods, culturally adapted IPAQ-SF into the Hausa language. A pilot study was conducted on 7 (out of the 12 recruited) subjects who could speak/write in the Hausa and English languages using the prefinal final version, after which the final version was developed for collecting data. Additionally, similar to the study by Awotidebe et al. [ 2021], some modifications were made after the pilot study of the Hausa version. For example, the phrase ‘aikin karfi’__meaning PA was misunderstood as being restless and replaced with, “motsa jiki”__physical exercise, “tafiya daga wannan wuri zuwa wancan”__traveling from place to place by walking was replaced with “tafiya (tattaki) daga wannan wuri zuwa wancan”__ “walking to move from place to place either for sports, leisure recreation or exercise”.
The Yoruba IPAQ-SF by Awotidebe et al. [2021] had good concurrent validity. For example, a significant correlation was found between vigorous PA in both the Yoruba and original English versions of the IPAQ-SF. Similar results were reported for the energy consumed in time spent sitting and total PA. However, there was a moderate correlation for energy consumed in walking and moderate activities between the original English and Yoruba versions. Additionally, in the female subjects, correlations were significantly greater than the in males for energy used in sitting times and other activities for both versions of the IPAQ-SF, which was inconsistent with the report of Oyeyemi et al. [2011], which found no meaningful socioeconomic and gender variance. Similarly, the Hausa IPAQ-SF by Oyeyemi et al. [2011] also had good concurrent validity ranging from moderate to high. For example, total PA from the original English version was significant/highly correlated with total PA in the Hausa version. Additionally, the time used in vigorous, walking, and moderate PA was significantly high and positively correlated with both versions of the IPAQ-SF. The time spent sitting on both versions of the IPAQ-SF (Hausa and English) was also significantly and positively linked. Additionally, in the study by Oyeyemi et al. [2011], the “Bland‒Altman plot” for the English and IPAQ-SF Hausa revealed a small mean variance that was not significant and with a wide range of 95% limits agreements. This finding was attributed to few subjects who reported moderate PA of more than 50 min/week. In contrast, the study of Awotidebe et al. [2021] did not carry out this plot.
Additionally, while the study of Awotidebe et al. [2021] carried out a discriminant and convergent validity (construct validity) and found a highly significant positive correlation between total PA and vigorous PA scores (good convergent validity) for the Yoruba version, the sitting time and total PA were not correlated (good discriminant validity). In contrast, the study of Oyeyemi et al. [2011] assessed construct validity by correlating the scores accrued from RPP with the time spent sitting from the IPAQ-SF Hausa and found a weak positive significant correlation. A similar result was reported between time used in moderate PA and BMI however, time used in walking, vigorous PA, and total PA showed no relationship with BMI and RPP.
The Hausa IPAQ-SF by Oyeyemi et al. [2011] showed an intraclass correlation (ICC) ranging from fair to good (substantial), with vigorous PA having the highest value and moderate PA having the lowest value. Additionally, the ICC scores for all items were significant and higher in males than in females (the ICC in women was lowest in moderate and highest in sitting activity). In contrast, the reliability (ICC value) of the Yoruba IPAQ by Awotidebe et al. [2021] ranges from poor to modest__ lowest for sitting and highest for vigorous PA. Additionally, the ICC value was fairly significant for all items in the Yoruba version; however, in the Hausa version, it was mostly substantial and modest. The findings for gender variance in the ICC value in the Yoruba version were consistent with those of Hausa__ higher in males than females.
The Syntheses of the data obtained from the Hausa and Yoruba IPAQ-SF
We conducted a synthesis of the physical characteristics, sociodemographic variables and psychometric properties of the data obtained from the recruited studies. The mean age and BMI of the subjects (n = 453) were 29.3 ± 6.255 years and 24.1 ± 4.335 (kg/m2) respectively. There were more male subjects (n = 293) than female and singles (n = 375) than married (Table 1 and Fig. 2).
Table 1
Physical and sociodemographic characteristics of the subjects for the Yoruba/Hausa IPAQ-SF
Variables | Yoruba (n = 351) | Hausa (n = 102) | Total (n = 453) |
Gender Female Male | 114 | 46 | 160 |
237 | 56 | 293 |
Marital status Married Single | 5 | 69 | 74 |
346 | 29 | 375 |
Variable | Yoruba | Hausa | Combine mean |
Age (years) | 22.4 ± 3.01 (18–35) | 36.2 ± 9.5 (20–65) | 29.3 ± 6.255 |
Body mass index (kg/m2) | 24.3 ± 4.37 | 23.9 ± 4.3 | 24.1 ± 4.335 |
Rat pressure product (RPP) | | 9694.3 ± 1195.94 | |
kg/m2 = kilogram per meter squared |
The correlation coefficient ranges from modest (r = 0.7185) to high (r = 0.8455), demonstrating good concurrent validity for the IPAQ-SF Yoruba/Hausa. These findings indicated that the total energy used in the total PA (MET-Min/week) in the IPAQ-SF Yoruba/Hausa was significant and highly (r = 0.839, p < 0.001) correlated with that of the original English version. Similar results were found in the moderate, vigorous, walking activities, and the time spent sitting. However, no meaningful gender variance was found, as presented in Table 2.
Table 2
Reliability (ICC) and concurrent validity of the Hausa/Yoruba IPAQ-SF
Hausa_ H (n = 102) plus Yoruba_ Y (n = 351) IPAQ-SF H + Y/2 | Total (n = 453) R | Female (n = 160) r | Male (n = 293) r |
VPA (Min week− 1) | 0.8105** | 0.8875** | 0.788** |
MPA (Min week− 1) | 0.7185** | 0.832** | 0.705** |
Walking (Min week− 1) | 0.7335** | 0.8885** | 0.6945** |
TPA (MET-Min week− 1) | 0.8455** | 0.9595** | 0.791** |
Time spent sitting (Min week− 1) | 0.839** | 0.899** | 0.7325** |
Hausa_ H (n = 102) plus Yoruba- Y (n = 240) IPAQ-SF H + Y/2 | Total (n = 342) ICC (95% CI) | Female (n = 102) ICC (95% CI) | Male (n = 240) ICC (95% CI) |
VPA (Min week-1) | 0.5815 (0.4375–0.6855) | 0.27 (-0.178-0.5805) | 0.6205 (0.4825–0.7215) |
MPA (Min week-1) | 0.328 (0.164–0.472) | 0.2275 (-0.055-0.473) | 0.3185 (0.1005–0.505) |
Walking (Min week-1) | 0.4355 (0.291–0.5555) | 0.2615 (-0.036-0.508) | 0.469 (0.301–0.5715) |
TPA (MET-Min week-1) | 0.492 (0.365–0.599) | 0.3645 (0.117–0.571) | 0.481 (0.319–0.6115) |
Time spent sitting (Min week-1) | 0.3575 (0.2295–0.478) | 0.568 (0.357–0.729) | 0.242 (0.0445–0.413) |
PA = Physical Activity, Min = Minute, MET = Metabolic Equivalent, r = correlation coefficient, **, p < 0.001, ICC = intraclass correlation coefficient, VPA = vigorous physical activity, TPA = total physical activity |
Additionally, Table 2 presents the results of the 1-week test-retest reliability (ICC), which generally ranged from 0.328–0.5815, highest for energy expended in vigorous PA and lowest for moderate PA. The coefficient of reliability for walking ICC = 0.469 (95% CI = 0.301–0.5715), vigorous ICC = 0.6205 (95% CI = 0.4825–0.7215), and total PA ICC = 0.481 (95% CI = 0.319–0.6115) were higher in males than in females. The time spent sitting showed the highest ICC in females (0.568, 95% CI = 0.357–0.729), while moderate activity showed the lowest score for reliability (ICC = 0.2275, 95% CI = -0.055-0.473).
In Table 3, the concurrent validity Male (p = 0.012), reliability total (p = 0.026), reliability female (p = 0.018) was all significantly different in the Yoruba, Hausa, and combined data, while the other variables were not significant (p > 0.05). The concurrent validity (r = 0.844) of subjects who were male in the Hausa was significantly (p = 0.022) higher than that in the Yoruba (r = 0.640) IPAQ-SF. Similar results were found in the reliability total and female (Table 4). However, the concurrent validity total and female and the reliability were comparable in both Hausa and Yoruba IPAQ-SF Table 4.
Table 3
Comparison of concurrent validity and reliability of the Yoruba, Hausa and combined data of the IPAQ-SF
Variables | F | p - value |
Concurrent validity total | 3.544 | 0.062NS |
Concurrent validity female | 3.243 | 0.075NS |
Concurrent validity male | 6.576 | 0.012* |
Reliability total | 5.024 | 0.026* |
Reliability female | 5.739 | 0.018* |
Reliability male | 3.199 | 0.077NS |
Analysis of variance__ ANOVA: * p < 0.05 = significant, NS = not significant |
Table 4
Comparison of concurrent validity and reliability of the Yoruba and Hausa IPAQ-SF
Variables | Yoruba | Hausa | t | P - value |
CV total | 0.723 ± 0.111 | 0.856 ± 0.054 | -2.414 | 0.054NS |
CV Female | 0.937 ± 0.053 | 0.850 ± 0 .062 | 2.381 | 0.045NS |
CV male | 0.640 ± 0.133 | 0.844 ± 0.064 | -3.099 | 0.022* |
Reliability total | 0.320 ± 0.104 | 0.558 ± 0.145 | -2.986 | 0.020* |
Reliability female | 0.187 ± 0.159 | 0.490 ± 0.126 | -3.348 | 0.011* |
Reliability male | 0.296 ± 0.126 | 0.556 ± 0.203 | -2.432 | 0.047NS |
CV = concurrent validity, * p < 0.05 = significant NS = not significant, |