1. IntroductionNon-communicable diseases (NCDs) are responsible for 41 million annual deaths globally [1]. Approximately 77% of NCD deaths are in low-and middle-income countries (LMICs) in South East Asia and Africa [1]. In LMICs like Kenya, NCDs are responsible for over 50% of all reported adult hospital admissions and 55% of adult mortality [2]. In these regions, the burden of morbidity and mortality due to NCDs is often overshadowed by the infectious diseases [3]. For instance, NCDs resulted in 67% of disability-adjusted life years (DALYs) and accounted for 74% of deaths globally [1,4]. However, In LMICs, NCDs receive 2% of global funds allocated from governments, the private sector, or donors in comparison to the response given to the treatments of infectious diseases such as HIV/AIDS that accounts for 3% of DALYs yet receives 30% of global funds [5,6,7,8].About three quarters of the global burden of type 2 diabetes (T2D) occur in LMICs [9]. In Kenya, the prevalence of diabetes is 12.2% in urban areas, higher than the global prevalence of 9.3% [10]. Kenya, like other LMICs, is undergoing a nutritional and epidemiological transition as the population increases the consumption of ultra-processed, high-caloric diets and adopts a sedentary lifestyle due to social-economic development and urbanization [11]. Dietary preferences are shifting from indigenous healthier food options, rich in micro-nutrients, high fiber and diverse, including a variety of fruits, vegetables, nuts, legumes, and root tubers to western-style diets that are energy-dense and ultra-processed, high in saturated fats, added sugar and sodium [9,10]. Additionally, countries in Sub-Saharan Africa, including Kenya, now have the highest smoking and alcohol consumption rates globally [11,12].Besides the dietary shift, changes have also occurred in physical activity patterns. For instance, there is a shift away from the high-energy expenditure activities such as farming, mining, and forestry towards the service sector and white-collar jobs, with reduced energy expenditures [13]. Similarly, economic growth, urbanization, and technological changes have also influenced and simplified how people move, with more individuals now owning cars or using train systems to conveniently get from one place to another, rather than walking [14], and this has contributed to a reduction in overall energy expenditure. In addition, unlike in the recent past, where leisure time would be spent outdoors, it is now more common that leisure time will include sedentary activities such as watching TV or smart devices [14]. Diet and physical activity transitions have influenced the demographic characteristics of those affected with T2D. For example, T2D was often associated with older age (65 years or older), but research now shows that globally, the greatest number of people with diabetes are between 40–59 years of age, with increasingly more younger people diagnosed with pre-diabetes [15].Another reason for the increase of T2D in Kenya is a lack of knowledge about diabetes and poor health attitudes [11]. Some studies report that low awareness of risk and preventive factors for T2D significantly contribute to the increasing cases of T2D in Kenya [10,16]. People are vulnerable to misinformation, unhealthy behavior, and poor health outcomes without proper health information [17]. Furthermore, the few existing T2D programs are primarily led by the government of Kenya through the Ministry of Health (MOH) [18]. However, according to the Ministry of Health, they face several challenges that prevent successful implementation of these programs, including a lack of capacity of the health workforce in terms of numbers, equipment, and skills and poor availability and affordability of quality, safe and efficient technologies and medications for screening, diagnosis and treatment [19]. As such, many of the available documented programs on T2D education are outdated and are not supported by behavior theories that offer a framework for understanding and predicting human behavior [18,19]. Additionally, many of the existing programs lack cultural tailoring, which is proven to be important in enhancing receptivity of health programs by adapting intervention materials to fit the needs, preferences, and norms of the population more appropriately [20].Due to the increasing prevalence of T2D enabled by the above-mentioned factors, there is a need to provide culturally sensitive diabetes education, guided by theory-based behavioral oriented methods that are effective in motivation and behavior change [21]. Consequently, this study aimed to examine the effectiveness of a theory-based and culturally sensitive educational intervention on diabetes prevention knowledge, health beliefs, dietary intake, physical activity, and weight status among adults in peri-urban communities in Nairobi, Kenya. 4. Discussion

Findings from this study showed that a theory-based intervention was effective in increasing diabetes knowledge and improving health beliefs (perceived susceptibility to diabetes, perceived benefits, and self-efficacy) in adopting a healthier lifestyle. The intervention also improved dietary intake with reduced intake of refined grains and oils and increased intake of fruits. However, the intervention did not result in a significant improvement in the other food groups, perceived seriousness, perceived barriers, physical activity and weight status.

The findings of this study showed an improvement in knowledge of diabetes risk factors, symptoms, role of nutrition and physical activity, alcohol and tobacco and other self-management skills. These findings are similar to other studies conducted in LMICs, including by Muchiri and colleagues, where participants in the treatment group who received a diabetes education intervention had higher knowledge scores than the control group [23]. Similarly, Chawla and colleagues found that participants that received education showed a significant increase in knowledge from baseline to endpoint compared to the control group [35]. Diabetes education interventions can lead to an increase in knowledge, as demonstrated by a systematic review that looked at over 19 heterogeneous trials and found that education interventions led to a significant increase in knowledge of T2D [35].Perceived susceptibility to diabetes, perceived benefits and self-efficacy improved in this study. However, perceived seriousness and barriers did not improve. Like these findings, several studies have found that while perceived susceptibility, benefits, and self–efficacy improve, there may be an inverse relationship with perceived seriousness. For instance, in a similar study where participants received an education intervention on calcium intake, the perceived seriousness of osteoporosis declined with increased knowledge, improved benefits and self-efficacy [36]. Likewise, Suratman and colleagues, in an educational intervention to improve perceptions for reducing exposure to pesticides in farmers, found a decrease in perceived seriousness with improved perceived susceptibility [37]. It is possible that as participants increase their knowledge and perceived benefits of adopting healthier behaviors and improve their self-efficacy, these improvements contribute to a decrease in the perceived seriousness of disease as participants feel empowered by the knowledge and skills to change their behavior to a healthier lifestyle.This study looked at barriers at the inter-personal level in the socio-ecological context, such as taste, time management, cost of food, social influences from family and friends, motivation, and stress [38,39]. These perceived barriers did not significantly improve. A possible reason for the lack of change is that the study was conducted during the COVID-19 pandemic and effects of the lockdowns could have exacerbated health barriers. Furthermore, barriers addressed were at the individual level, and yet there may be others at the macro level, such as lack of employment opportunities, gender inequalities, inhibitive policies, poverty, and insecurity [40]. Therefore, future studies can incorporate a broader scope while addressing perceived barriers to diabetes health.The intake of whole grains was low at 2.4 cups compared to the recommended 2.5–4.5 cup equivalents per day. Refined grain intake was high, with average intake of 6.03 cups compared to the recommended intake of 3.5–4.5 cup eq/day. These findings are consistent with other studies investigating dietary patterns that have found that the average diet in Nairobi, Kenya is heavily reliant on carbohydrates as a source of energy intake with primary staples being maize products, wheat, rice and cooking bananas [41]. A time series study assessing staple food consumption patterns in households in Nairobi, and its environs, found that there was a significant increase in consumption of milled and refined grains overtime at the expense of whole grains [41]. This could explain the high intake of refined carbohydrates versus whole grains in this study. The intake of refined grains significantly reduced after the diabetes intervention, and this was ultimately a positive effect as there’s higher risks of obesity and associated chronic diseases, with increased intake of refined grains [42]. However, while refined grains reduced, the intake of starchy vegetables increased. This may signify a dietary compensation mechanism that has been observed in some studies, and that needs further investigation, where in response to adjustments in energy intake, individuals compensate by increasing intake of certain foods to maintain satiety and regardless of portion control there’s constant energy intake [43]. The high mean intake of dark green vegetables (4.95 vs. 2.0–2.5 cups/week), red and orange vegetables (6.15 vs. 6.0 cups/week), and fruits (3.68 vs. 2.0 cups/day) at baseline could explain the lack of significant change in intake after the intervention [44]. Nyanchoka and colleagues had similar findings in a study in Kenya where 78% of the respondents met fruits, vegetables and beans/peas/lentils recommendations [45].Regarding the lack of change in physical activity, participants in this study already had a high physical activity of 3000 MET minutes per week with a combination of walking and moderate and vigorous activity at baseline assessment. It is possible that this contributed to the lack of observable change in the amount of physical activity at post-test and follow-up assessment. Weight status did not change significantly after the intervention. Notedly, there was a significant portion of the sample in the underweight and normal weight categories, and this could have statistically affected the ability to observe significant change in weight. Additionally, weight loss is extremely challenging due to interactions between our biology, behavior, and obesogenic environments [46]. Change in body weight also requires consistent behavior change and long-term observation, which could explain the lack of significant change in this study [47]. It is recommended that future interventions that hope to observe a significant change in weight status have a longer post-intervention period, preferably longer than the four weeks applied in this study. Implications and Recommendations

The findings of this study have several implications and recommendations for policy and future research. The increase in diabetes knowledge, perceived susceptibility, benefits, and self-efficacy after the intervention, amplifies the importance of educating people to prevent and manage T2D. Education interventions should target improving awareness of T2D and increasing perceived threats to diabetes, and the benefits of adopting healthier behavior while addressing barriers and building self-efficacy.

The CHWs were instrumental in implementing the intervention, since they are well known to the community, they can be tapped into to educate the community on T2D as a sustainable measure and can facilitate the scaling up of existing or future T2D programs. Lastly, the cultural and theory-based aspects of the intervention were a crucial part of ensuring that there was increased knowledge, perceived susceptibility, perceived benefits, self-efficacy and improving dietary intake. These have important implications for the designing of future community-based health interventions focusing on T2D prevention and management. It is recommended that similar health education interventions be culturally tailored, theory-driven and should also apply low health literacy strategies. They should also extend their duration to measure sustained behavioral changes.

This study had some limitations. For instance, some data were self-reported and may be subject to bias resulting from recall or social desirability, especially in reporting dietary intake data and health beliefs. Additionally, while the diabetes knowledge questionnaire was pilot tested and underwent face and content validation by experts, it may benefit from further confirmatory analyses to establish reliability before use in other settings. Again, the study was not powered to detect non-primary outcomes, which may have affected our ability to observe changes in physical activity and body weight. Lastly, the short study period may have limited the ability to detect body weight changes and hindered the evaluation of the long-term effects of the intervention effects.