The nutrition transition in Morocco
I will first discuss the prevalence of female obesity in Morocco in the context of the nutrition transition theory. The nutrition transition is distinguished by a shift from traditional diets high in grains, fruits and vegetables and low in fat to diets high in sugar, refined grains and fat (Drewnowski and Specter 2004, Monteiro et al. 2002). In the Moroccan context, the diet has changed considerably as a result of the socioeconomic transition: the intake of animal products has increased while that of cereals and sugar remained relatively high, reflecting the specificity of Moroccan dietary habits. In addition, the rise in the consumption of meats and vegetables has been accompanied by a steady consumption of bread, used to eat the sauce in which the meat and vegetables are cooked. This is all occurring in combination with the excessive all-day consumption of the national drink, mint tea – whose main ingredient is large amounts of sugar – as well as increased access to imported processed foods associated with high social status (i.e. McDonalds, Coca Cola).
According to the nutrition transition theory, increases in urbanisation promote a significantly more sedentary lifestyle, as both leisure and labour activities are replaced by more mechanised activities, such as television watching and non-agricultural labour (Brown 2001, King et al. 2001, Pasquet et al. 2003, Pearte, Gary and Brancati 2004, Sobngwi et al. 2004). In Morocco, the urban population has increased from 29 per cent in 1960 to 53 per cent in 1997 and, on average, 145 people out of 1000 own at least one television set. Both of these numbers are increasing rapidly (UNDP 2005).
Rapid urbanisation has also been observed to produce the coexistence of malnutrition and obesity based on studies in Brazil, Russia, China and the US (Doak et al. 2000, Wang et al. 2002, Monteiro et al. 2002, 2004). As such, food insecurity still exists in many countries, such as Morocco, experiencing rapid increases in obesity. Contrary to more developed countries, food choice in developing countries may remain limited, either because of market limitations or cost (Aguirre 1994, Gulliford et al. 2003). Relative to the rest of the African continent, however, Morocco has a low proportion of under-nourishment, a relatively high dietary energy supply (per capita kcals/day) and a ‘sufficient’ national food energy supply (FAO 2006). Consequently, according to the World Bank (1998), Morocco is considered among other countries not to be ‘highly food insecure’. However, this was not always the case. Historically, Morocco has experienced times of significant famine and starvation and this still occurs seasonally, as witnessed recently by droughts in Southern Morocco (Meyers 1981, Swearingen 1992). Thus, there might be an important generation effect on the relationship between food insecurity and obesity risk in Morocco. There is increasing evidence that malnutrition early in life is one additional risk factor for obesity and other chronic diseases in adulthood, as well as resulting in an increased risk of stunting (Barker 1992, Law et al. 1992, Osmani and Sen 2003, Phillips et al. 1994).
Nutrition transition theory does not consider the role that gender and other social, economic and cultural factors may play in influencing diet and physical activity patterns. That is, these macro-level processes may have differential effects on the diet and physical activity patterns of men and women. Moreover, some of the aforementioned unique aspects of the experience of less developed countries may also affect women differently from men. For example, girls may have been more likely to experience stunting due to son preference beliefs and practices. During periods of food shortage, girls and women may have been the first to experience the effects of malnutrition and undernutrition. The differential effects of such macro-level processes have yet to be incorporated into a framework of social, epidemiological, demographic, and economic change.
Nutrition transition theory, however, is inadequate in explaining the persistent gender differentials in obesity risk in Morocco. It does not suggest a clear mechanism through which this relationship may operate nor does it reflect the particular socio-cultural contexts in which these inequalities occur. I suggest, therefore, an alternative theoretical framework to explain the gender differential prevalence rates of obesity in Morocco that combines both role theory and empirical data from Morocco.
My starting point is that lifestyle is a key factor in accounting for the rise of obesity rates, and gender differences in lifestyle provide one possible explanation for this obesity risk. Since women and men are exposed to different lifestyles, this differential ‘exposure’ affects their health. In this way, in a Durkheimian sense, obesity rates can be conceived as social facts that can be explained in terms of lifestyle changes. Yet, it is not enough to say that obesity is caused by changes in eating habits. Rather, the way people eat is interconnected to other aspects of how they conduct their lives and the related ‘power balances’ at play (Elias 1978, Crossley 2004). I examine these power balances by focusing on how men's and women's roles in the household determine their specific and different lifestyles. The primary mechanism through which the relationship between stratification and obesity operates is through the particular household roles men and women occupy in Morocco. Role theory helps us to understand the relationship between men's and women's household roles and their health.
Role theory suggests that social institutions are made up of roles into which individuals fit. This framework recognises that these roles are pre-existing rather than negotiated by the individual. However, it incorporates a functionalist approach to society when it describes the roles as being made up of institutionalised norms and values (Collins 1994). That is, individuals are assigned roles (role allocation) and socially expected behaviour patterns (socialisation) by others, and act in response to their perceptions of the expectations that others, especially significant others, hold for them (Biddle and Thomas 1966).
Aspects of Nathanson and Schoen's (1993) theory of bargaining can be applied to the exchanges that occur across different Moroccan households. Their theory argues that women seek economic security and social status. In Morocco, this is most often achieved when a woman exchanges her domestic services and childbearing ability for her husband's financial support. According to this theory, other strategies may be employed based on the normative setting in which this exchange is occurring. In the Moroccan context, there is a clear set of rewards and sanctions for the particular strategy a woman selects in order to secure financial security. One common strategy for women in Morocco is to accept normative roles and their corresponding activities, such as those associated with religious rituals. Women are often motivated to conform to these roles because of both the set of rewards associated with the fulfilment of expected roles and the consequences of rejecting these roles.
Role theory has been extensively criticised for its overly functionalist basis and its lack of considering agency. In a recent study of women's health in mid-life McMunn, Bartley and Kuh (2006) draw on Giddens's (1984) theory of structuration and Doyal and Gough's (1991) theory of human needs to develop a theory of role quality based on the concept of agency. Giddens suggests that human agency and social structure are in a relationship with each other, and it is from the repetition of the acts of individuals that the structure is reproduced. According to Giddens, however, social structures such as traditions, institutions, moral codes, and established ways of doing things can be renegotiated when individuals replace them, ignore them or reproduce them in different ways. Yet, if the individual does not have what Doyal and Gough term ‘the autonomy of agency’ then these structures will be maintained. Following their work, I suggest that in the context of Morocco it is also important to consider the relationship between social roles and women's agency. That is, patriarchal structures (i.e. norms, traditions) limit the autonomy women have to choose their social roles and it might be this lack of agency that also contributes to the relationship between social roles and health, and not solely the deleterious health effect of the activities associated with the particular role. For example, it is well established that women in certain societies are often not in a position to choose or even prioritise work over family (Merton 1957, Moore and Gobi 1995, Izraeli 1993). Indeed, what follows, is that in most societies the work of caring for the family and doing domestic work falls to women – or is at least considered a woman's domain – while work outside the home is usually considered the domain of men (Lehrer and Stokes 1985, Marini and Brinton 1984).
It is well established that the social roles men and women occupy may account for gender differences in mental health (Sachs-Ericsson and Ciarlo 2000, Gore and Mangione 1983, Simon 1995), but obesity has only recently begun to be examined in this light (McMunn et al. 2006). Several theoretical frameworks describe the particular relationship between roles and health: women are thought to have poorer experiences within any given role (role strain theory), have more conflicts among their different roles (role-configuration theory), or benefit from multiple roles (health enhancement hypothesis) compared to men (Gove 1972, 1978).
The importance of household labour for women's health has already been noted in several studies (Bartley et al. 1992, Annandale and Hunt 1993). No study, however, has tried to explain gender differences in health through gender differences in household labour (Lennon and Rosenfeld 1992, Artazcoz et al. 2001, Borrell et al. 2004). In the Moroccan context, women's household roles, including childrearing and domestic chores (cleaning and food preparation), do not leave women with much leisure time to participate in exercise. Men, however, are more likely to participate in exercise than women because their household roles do not limit their leisure time. In addition, because food preparation is an integral part of women's household roles, women often eat snacks between meals thus increasing their overall daily caloric intake. Furthermore, during meals women tend to eat less nutritious, higher caloric foods than men because they tend to eat last. This is because of logistical reasons (i.e. they are busy serving the men) and ideological reasons (i.e. men are supposed to eat first because they are more socially and economically valuable).
Determinants of women's household roles in Morocco
Based on role theory and qualitative research in Morocco, I conceptualise marital status, age and socioeconomic status to be determinants of women's household roles in the Moroccan context.
It is important to recognise that women's roles in Morocco have changed over time both inside and outside the household. For example, recent reforms to the Moroccan family code (Mudawana) in 2004, which aimed to empower women, suggest interesting questions about how changes in women's status may affect their role in society (Sadiqi and Ennaji 2006). Moroccan women are increasingly participating in the paid labour market, delaying marriage and having fewer children (Assaad and Zouari 2003). The nature of their household roles has become somewhat less energy exerting with greater access to modern conveniences and technology, such as running water and refrigerators. However, it is difficult to quantify the effects of these changes on women's health because not all Moroccan women have equal access both to employment and technology. Similarly, there is tremendous diversity among Moroccan women in terms of their role preferences (Davis 1983).
It is generally accepted that, overall, the married are in better health than the unmarried, because marriage has beneficial effects on health (marriage protection effects) and/or because healthier individuals are more likely to marry and to stay married (marriage selection effects) (Ross, Mirowsky and Goldsteen, 1990, Waldron, Hughes and Brooks 1996). However, the relationship between marital status and obesity is not well established (Sobal, Rauschenbach and Frongillo 2003). For example, a study examining the effect of ethnicity on weight gain in the United States found that married people weighed more than the unmarried (Sobal and Devine 1997). More specifically, they found that women tended to change weight more in that first year of marriage than men did. They point out, however, that other studies have shown that during the first two years of marriage, husbands and wives tended to exercise less and eat more but only husbands gained weight. Married men in these studies were found to have a higher BMI and to be more likely to be obese than never married or previously married men. These differences, however, do not emerge during the first year of marriage.
Similarly, another study in Poland found that in general married men and women were more likely to be overweight and obese than never married individuals (Lipowicz, Gronkiewicz and Malina 2002). The results indicated a significant association between marital status and BMI in both sexes. In fact, after age, marital status was the most important predictor of overweight/obesity among men in this study (Lipowicz, Gronkiewicz and Malina 2002). A Bahraini study on the prevalence of obesity and the demographic factors associated with it found that marriage was positively associated with obesity among women (al-Mannai et al. 1996).
Marriage is a particularly salient institution in Moroccan society (Kapchan 1996, Mernissi 1975). Families often spend a significant portion of their savings and monthly income on a wedding and may even invest in their daughter so as to improve her marriage potential. For many a woman, becoming a wife is the most important role that she and her family strive for throughout a woman's life in Morocco (Mernissi 1975). Even starting at an early age, parents take into consideration the consequences that certain decisions may have on their daughter's potential for marriage, such as her education and employment opportunities. For example, working as a maid in a household may decrease her potential for marriage.
Following marriage and becoming a wife, the next role most women work towards is becoming a mother as soon as possible after marriage. This role can be thought of as the ‘gateway’ to all other marital roles. While some women are able to fulfill this marital role rapidly and to acquire their corresponding rewards, others are not as successful and suffer the consequences, which may include divorce and tension in the household. In addition, there are certain activities associated with the role of daughter-in-law, such as performing the heavy household labour and any other chores requested by one's mother-in-law, such as preparing food involved with religious rituals. Although not as significant within the early years of marriage, the lifestyle associated with marriage may contribute to a woman's increased obesity risk over the lifecourse.
The accumulation of traditional roles, the most significant being the role of wife, contributes to a woman's obesity risk for several reasons. First, the expected activities associated with the role of a wife are located exclusively in the household and are often centred on food preparation. Secondly, prior to marriage, a woman may have more leisure time to participate in physical activity or other beneficial health activities. Thirdly, a single woman is not likely to have given birth which removes the possibility of weight gain associated with pregnancy. Fourthly, marriage and obesity risk are both associated with increased age; therefore, single women are often younger and are less likely to be obese.
In the Moroccan context, age generally determines one's status within the household and how tasks are allocated (Moghadam 1995). Younger household members, particularly women, often perform the most energy exerting domestic chores. For example, a pattern can be observed in which daughters-in-law, along with the other younger female household members, perform the majority of heavy domestic tasks and older women have more sedentary responsibilities such as child care. In addition, women gain household status through their increased age. As a result, older women usually eat at the same time as the men. They thus consume higher quality food (i.e. more protein).
As discussed above, women accumulate a variety of roles over the lifecourse. Later in life, women are faced with another status-changing role; mother-in-law. Becoming a mother-in-law has important consequences and possible challenges. For example, when a mother becomes a mother-in-law her sphere of authority widens significantly, and her household roles change and become less energy exerting. For example in terms of religious ritual, the mother-in-law may be given the honour of pouring the tea for guests, but she would no longer be responsible for its preparation. In this same fashion, her contribution to domestic work becomes primarily supervisory. For example, a mother-in-law is likely to delegate some or all of the heavy workload to her daughter-in-law, such as cleaning the home. In exchange for her daughter-in-law's domestic labour, a mother-in-law may assume some of the responsibility for taking care of the children in the household (Lane and Meleis 1991).
The ordering of household roles by age and marital status may be moderated according to a household's particular economic situation. Morocco can be characterised as having a sizeable unequal distribution of wealth such that there is a large (and, indeed, growing) gap between the rich and poor. Although a significant middle class is beginning to emerge in the major cities (i.e. Casablanca and Rabat), where industry and employment are most abundant, there remains a sharp and significant economic disparity between the highest and lowest economic classes (White 2001).
I suggest that cultural beliefs (i.e. certain ideas of beauty and space) and cultural practices (i.e. non-egalitarian divisions of household labour and particular rituals of hospitality) may foster certain household roles associated with increased obesity risk. The effects of these cultural beliefs and practices may be moderated by several socioeconomic co-variates, such as education, income and material lifestyle.
First, education may interact with the effects of cultural beliefs and practices on health. More specifically, high levels of education may reduce the effect of these beliefs and practices on a woman's likelihood of participating in obesity risk behaviour. Women with high levels of education may have access to information regarding the negative effects of obesity risk behaviours. In addition, education may allow women to participate in the labour force and thus be exposed to alternative health knowledge.
Secondly, income may interact with the effects of cultural beliefs and practices on health. Most importantly, women from high income households may have the financial ability to choose the foods they consume. Depending on income, households may be forced to allocate food resources based on a household member's economic contribution and thus may or may not practise gender segregated eating patterns. The Nutrition Efficiency Wage Hypothesis suggests that there is a nonlinear link between health and productivity and that the function is convex at very low levels of health (Strauss and Thomas 1998). That is, there may be some people who are so poor and so unhealthy that they are too costly to be employed. People in better health (higher BMI) are more likely to undertake strenuous tasks (Pitt, Rosenzweig and Hassan 1990). Similarly, this model predicts that poor households will allocate health resources unequally among members to ensure that at least one member is fit to work. This is less likely to occur in higher-income households.
Thirdly, in the Moroccan context, different social classes have particular material lifestyles associated with obesity risk behaviour. This assertion reflects Bourdieu (1984) who suggests that different classes think differently in relation to food, activity and the body, such that some groups are more prone to obesity than others. In fact, Williams (1995) has argued that we can observe Bourdieu's theoretical ideas on the class-based nature of ‘relations to the body’ in empirical data on health-related matters.
For example, higher social classes in Morocco almost always have domestic help. The presence of maids may change the nature of a woman's household role, as her responsibilities may no longer consist of domestic chores, such as food preparation and cleaning. More directly, domestic help physically reduces female household members’ burden of household chores, as well as the physical exertion accompanying these chores. In addition, higher income households may also own appliances, such as washing machines and dishwashers, resulting in the reduction of the physical exertion required for household activities. Living in a home with running water (thus avoiding the physical labour and time associated with fetching water from a well) and owning a refrigerator (thus making food preparation more efficient because food can be saved for future meals and shopping can be done less frequently) will similarly reduce the physical labour associated with domestic chores.
Women with access to these material goods might have increased leisure time. The activities that occupy this ‘leisure time’ are also dependent on social class. For example, women of the highest social classes may have been exposed to gyms and Western ideas of beauty whereas women of a lower social class may associate obesity as a symbol of affluence and may occupy their often-limited leisure time with what they conceive as high-status behaviour, such as watching television. As a symbol of status, as well as through the ability to afford the financial costs associated with exercise, high social class women may be more able to participate in exercise. In addition, because of their higher levels of education and exposure to preventive medicine these women may hold a certain knowledge regarding the etiology of obesity. These women may abstain and/or participate in certain adverse obesity risk behaviours, such as particular diets and physical activity patterns.
Women of high social classes may also hold body image ideals more congruent with ideas of beauty in the West that overwhelmingly value slender women. Following these ideas of beauty, women may alter their diet, physical activity and even dress patterns (i.e. whether to wear the traditional djelaba, a loose, non-form-fitting dress). Body image has been identified as closely related to social class. Higher social classes have been found to be associated with greater body dissatisfaction and slimmer body image ideals across diverse cultural settings, including Canada, the US, India and Egypt (McLaren and Kuh 2004, Jackson, Rashid and Saad-Eldin 2003, McLaren and Gauvin 2002, Chugh and Puri 2001). As noted, body image is not a static construct: it is dynamic and changes as women encounter changes in socioeconomic status at both the individual and household levels. In combination with other measures of material lifestyle, body image helps to elucidate how socioeconomic class affects obesity risk through the household roles it creates for men and women.
Middle class women, therefore, were hypothesised to be at the greatest risk of obesity because they neither participate in energy-exerting household labour nor do they access gyms and other alternative sources of exposure (i.e. particular media images) that might encourage a slim body image ideal. In sum, cultural contexts and structural variables do not exist in separate environments in Morocco. I therefore suggest that it is the particular combination of these contextual factors that explains the growing prevalence of obesity among women and the persistence of gender difference in prevalence rates.