There is growing concern about the rising global prevalence of overweight and obesity, as well as their associated adverse health implications [1–3]. One feature of this global problem is the fast pace at which developing countries are being affected compared to the developed countries. For example, while the number of people affected by overweight or obesity increased 1.7 times between 1980 and 2008 in developed countries, those affected in developing countries more than tripled from around 250 million people to 904 million over the same period [4].

As in many developing countries, the prevalence of overweight and obesity in Ghana has consistently increased over the last few decades, disproportionately affecting women than men. For instance, the Ghana Demographic and Health Survey (GDHS) estimated the prevalence of overweight or obesity among women (40%) aged 15–49 years to be more than twice that of men (16%) [5]. Specifically, the prevalence among women was estimated to have increased by 27% between 1993 and 2014 [5]. Further, the prevalence of overweight or obesity in Ghana affects more urban than rural residents, and increases with household wealth and level of education [5].

Although overweight and obesity result from energy imbalance (consuming more calories than are equivalently expended in physical activity), studies have found a variety of factors to be associated with the phenomenon. Various associations, in terms of magnitude and direction have been reported between overweight and obesity, and a number of factors including age [6, 7], gender [8, 9], socioeconomic status [10, 11], marital status [12, 13], education [14, 15], occupation [6, 15], ethnicity [6, 14], genetics [16], dietary [17, 18] and physical activity patterns [19]. For instance, contrary to developed countries, the wealthier, more educated and urban populations are more at risk of overweight and obesity in developing countries [10, 11, 20]. Prior studies [6, 21, 22] in Ghana have found overweight and obesity to be positively associated with  age, household wealth, education, being married, and parity among others.

Engaging in sedentary behaviours, such as television (TV) viewing has been implicated among the multiplicity of the factors underlying the increasing prevalence of overweight and obesity observed in many populations around the world [23, 24]. A number of theoretical propositions about the possible mechanisms through which TV viewing affects overweight and obesity have been advanced. It is hypothesized that, TV viewing displaces participation in high-intensity discretionary physical activity, reduces resting energy expenditure compared to other activities, and increases sleep deprivation [19, 25]. Nonetheless, studies exploring this hypothesis have found a weak relationship between TV viewing and physical activity [19, 25].

An alternatively hypothesis posited is that TV viewing leads to an increase in overall energy intake either indirectly through exposure to advertising and consequent intake of foods commonly advertised on TV, or directly through the consumption of high calorie foods and beverages while viewing TV [17, 18]. Indeed studies involving both children and adults have found a positive association between TV viewing and intake of high calorie foods such as soda, pizza, and high-energy snacks [17, 18, 23]. In an experimental study involving children, Blass et al. [26] found children’s energy intake to be higher when viewing TV than in control conditions when the TV is switched off.

Beyond exploring the possible mechanisms between TV viewing and physical activity or calorie intake, positive associations between TV viewing and obesity and/or related indicators have consistently been observed across space and time. Studies in the United States (US) have found that men and women viewing the most TV have an increased risk of obesity compared with those viewing the least TV [27–29]. In Australia, women who viewed TV for more than three hours a day were found to have a higher prevalence of severe abdominal obesity; while men who viewed TV for the same amount of time had a higher prevalence of moderate abdominal obesity [23]. More recently, a worldwide cross-sectional study across low, middle, and high income countries involving 207,672 adolescents from 37 countries and 77,003 children from 18 countries found increased TV viewing hours to be positively associated with body mass index (BMI) [30].

Although the public health importance of the association between TV viewing and obesity or related outcomes have been demonstrated in both cross-sectional and prospective studies involving children, adolescent and adult samples in other regions of the world, similar studies are rare within the African region. Nonetheless, while examining the relationship between ownership of different types of household assets and BMI among Ghanaian women, Dake and Fuseini [31] tangentially observed that women who reported viewing TV almost every day were likely to have obesity compared to those who did not. Taking cue from Dake and Fuseini [31], this study explored the association between TV exposure and overweight/obesity in Ghana with the view to contributing to the discourse on TV viewing and obesity, particularly in the African region.

In Ghana, TV ownership has been increasing since the inception of Television broadcasting in 1965. The sixth Ghana Living Standard Survey (GLSS) [32] estimated TV ownership to have increased between 1998/99 and 2012/13 from 40 to 75% and from 12 to 34% in urban and rural households, respectively. Indeed, hardly would one pass by a street, shop, bar, restaurant, bus station or even an office in most cities, towns and villages in Ghana without seeing a TV streaming. The screening of foreign telenovelas (soap operas) during primetime has become a means for competing stations to attract large audience, particularly women [33]. Telenovelas seem to have gone “viral” to the extent that TV stations screen them almost simultaneously with people – mostly women – seated and viewing for considerable hours each day. Some TV stations translate telenovelas into the local language (Twi) which makes it the more attractive for those who would otherwise have been deterred by language barriers.

With TV viewing considerably a part of life in Ghana today, this study explored the relationship between TV viewing and overweight/obesity among Ghanaian women using data from the 2014 GDHS. The study focused on women mainly because of the disproportionately high and rapid rates of overweight/obesity among Ghanaian women, compared to men [5]. In addition, Ghanaian women have also been found to report low levels of vigorous physical activity, which could increase their risk of developing overweight/obesity [34, 35]. This study would be useful in broadening our understanding of some of the drivers of obesity in Ghana or Africa, as well as at-risk groups for designing interventions.

Table 1 shows the distribution of background characteristics of women in relation to TV exposure and prevalence of overweight/obesity. In general, TV exposure ranged from about 4% among those with TV in their households, but with no TV exposure to 42% among those with TV in their households, and with high TV exposure. This is further illustrated in Fig. 2. TV exposure varied across the various age cohorts considered in this study. For example, while slightly more than a quarter (26%) of those with no TV in their households, and no TV exposure were 45 years or older, 47% of those with TV in their households, and with high TV exposure were within the 25–35 year cohort.

TV exposure also varied by educational status. The proportion of those with no TV in their households, and with no TV exposure decreased as level of education increased (from about 44% among those with no education to approximately 5% among those with secondary/higher education). In contrast, the proportion of those with TV in their households, and with high TV exposure seemed to increase with each higher level of educational. This ranged from about 21% among those without education to about 65% among those with secondary/higher education. The results on marital status show that 15% of women without TV in their households, but with high TV exposure were in the widowed/divorced/separated category; while approximately 45% of their counterparts with TV in their households and with high TV exposure were cohabiting.

More than two-thirds of those in the poorest wealth quintile had no TV in their households, and no TV exposure. However, the proportion of women with TV in their households, and with high TV exposure increased along with wealth quintile from about 8% in the poorest quintile to about 73% in the richest quintile. With regards to occupation, a greater proportion of those with TV in their households, and with moderate TV exposure (23%), and those with TV in their households, and with high TV exposure (63%) were engaged in professional/managerial/clerical occupations. The results also show differences in TV exposure by type of locality. For instance, while more rural (9%) than urban (7%) women had no TV in their households, but had high TV exposure; more urban (58%) than rural (22%) women had TV in their households, and had high TV exposure.

The prevalence of overweight/obesity among the women in this study was 40%, with some variations observed in relation to their TV exposure and background characteristics (Table 1). Overweight/obesity ranged from about 21% among those with no TV in their households, and no TV exposure to about 51% among those with TV in their households, and with high TV exposure. Regarding the other characteristics of women, the prevalence of overweight/obesity increased with age, educational level, and wealth quintile. Overweight/obesity was highest among widowed/divorced/separated (57%) women, those in professional/managerial/clerical (59%) occupations, urban (49%) dwellers, and those with three children (55%).

From the multivariate results (Model II), however, it was observed that only those with TV in their households, and with high TV exposure were significantly more likely (OR = 1.39, 95% CI = 1.002, 1.923) to have overweight/obesity, compared to those with no TV in their households, and no TV exposure. With the exception of marital status and parity, significant associations were observed between overweight/obesity and all the other background characteristics included in the multivariate model.

This study sought to explore the association between TV exposure and overweight/obesity among Ghanaian women. Despite the attenuation of the association between TV exposure and overweight/obesity after adjusting for other factors, the results show that women who had TV in their households, and had high TV exposure were significantly more likely to have developed overweight/obesity than their counterparts with no TV in their households, and no TV exposure. This finding resonates with previous studies [23, 30, 31, 40] on the association between TV viewing and obesity, irrespective of the fact that such studies have reported varied magnitudes of association, depending on the covariates studied and how TV viewing and obesity were measured and categorized.

The cross-sectional nature of this study limits the possibility of drawing direct cause-and-effect conclusions based on the results. Nonetheless, a number of plausible explanations for the findings could be surmised, particularly in the light of existing body of knowledge on the subject. First, it is possible that women who have TV in their households, inherently spend a lot of time viewing TV in place of engaging in exercise or other forms of rigorous physical activities (for example, traditional dancing or wrestling competitions among others), resulting in an overall decrease in energy expenditure and the subsequent development of overweight/obesity [17, 18, 23, 40].

There is evidence of reduced physical activity globally, which is attributable to gradual shifts from outdoor leisure-time physical activities to sedentary types of activity such as viewing television [41]. In the developing world for instance, Ng and Popkin [42] report high declines in physical activity and a shift towards sedentary activity in domestic domains aided by greater access to home appliances including TV and computers. Hence, it might as well be the case that women with overweight/obesity resort to TV viewing as their main form of recreation rather than engaging in more physically demanding recreational activities such as sports.

An alternative explanation is that the more time such women spend viewing TV, the more they are likely to be inundated with the dozens of commercials and programmes which promote the consumption of high calorie foods [17, 18, 43]. Indeed, commercials on food and beverages (both alcoholic and non-alcoholic) seem to dominate in all the TV stations available in Ghana. Besides the potential of TV commercials to influence dietary patterns and risk of overweight/obesity among Ghanaian women, there is equally the tendency that such women have the habit of consuming high calorie food or drinks simultaneously while viewing TV without realising the quantities being consumed [17, 18].

Overall, a cycle of mutual interaction and reinforcement between TV viewing, physical activity, and dietary patterns could be at play [41], leading to an increased risk of overweight/obesity among Ghanaian women. Thus, as more time is spent viewing TV, less time is spent engaging in physical activities, while the consumption of high caloric food increases, leading to the development of overweight/obesity. And with the onset of overweight/obesity, the drive to substitute more active pastime activities for passive ones such as TV viewing increases, and the cycles continues on.

The positive effect of TV exposure on overweight/obesity could equally be a function of other confounders as prior studies [17, 18] have noted. Indeed the strong association found between the control variables in the present study and overweight/obesity, and their effect on the link between TV exposure and overweight/obesity demonstrate this possibility. In particular, the strong positive effect of education and wealth on overweight/obesity in this study highlight the consistent effect of socioeconomic status on the risk of developing overweight/obesity typically observed in developing countries such as Ghana [6, 22], in contrast to developed countries [10, 11]. In developing countries, including Ghana, socioeconomic development has been found to be associated with changes in dietary and physical activity patterns such that Western type foods and energy saving devices replace traditional (healthier) staple foods and ways of doing things [31, 44].

Apart from the socioeconomic dimensions which seem to be playing out in this study, socio-cultural interpretations are equally plausible. Some Ghanaian women might still hold on to the historic Ghanaian perceptions that large-bodied women are beautiful, healthier and more prestigious, and may therefore be engaging in fattening dietary and physical activity patterns [45, 46]. Contrary to the reported protection that residing in a rural locality provides against overweight/obesity, this study found women in rural localities to be at risk of developing overweight/obesity. As Tuoyire et al. [6] noted, this could be the result of the gradual exposure of rural residents to obesogenic risks commonly seen in urban areas.

Much as the study provides some useful insights for interventions to remedy the problem of overweight/obesity in Ghana, the results need to be interpreted with caution in the light of some limitations. As already alluded to in the preceding section, the study cannot directly establish cause-and-effect relationship between TV exposure and overweight/obesity due to the fact that the data was collected based on a cross-sectional design. Nonetheless, the associations established in this study are to a large extent consistent with prior prospective [27] and cross-sectional studies [23, 30] that have demonstrated connections between TV exposure/viewing/watching and the development of overweight/obesity. Secondly, due to lack of data, the study could not control for other potential mediators of the relationship between TV exposure and overweight/obesity, most importantly dietary behaviour, physical activity and genetic factors. In a related development, the measurement of TV exposure as used in this study could not account for specific duration of TV viewing; neither could it account for the content of the messages to which study participants were exposed. Perhaps, the temporal amount and content of exposure to TV messages is one research area that needs to be explored further.

These limitations notwithstanding, the current study is about the first to specifically examine the relationship between TV exposure and overweight/obesity in Ghana and Africa where overweight/obesity and associated non-communicable diseases are increasing rather rapidly. An additional strength of this study is that the findings can be generalised to a large extent, since national level GDHS data with a high response rate (98.5%) was used. Further, the study contributes to the TV-obesity discourse and broadens our knowledge on some of the potential drivers of overweight/obesity in Ghana and perhaps the rest of the African region.

The study demonstrates that increased TV exposure is significantly associated with overweight/obesity among women in Ghana even after adjusting for other factors. Interventions aimed at tackling obesity in Ghana should focus on encouraging the uptake of more physically demanding pastime activities (such as brisk-walking and dancing) in place of TV “sit time”. More importantly, intervention planners should consider using the TV as an important means to channel tailored behavioural change communication messages that seek to promote healthy dietary and physical activity behaviours.