Undernutrition

As a general rule, the functionality of crucial organs such as the kidney and gastrointestinal tract deteriorates with increase in age and may have an impact on nutrient absorption. In addition, food intake in older persons can be affected by issues related to muscular and skeletal changes that may impact on ability to prepare food, appetite, taste acuity and dental use. These as well as social issues including loneliness and bereavement can lead to older people becoming undernourished. Older persons are also at a higher risk of impaired immunity and susceptibility to infection which can be exacerbated by undernutrition^{( 14 )}.

Undernutrition in many cases is classified as having a BMI <18·5. In a summary of research findings by HelpAge International in 2004, the undernutrition prevalence across several countries in SSA was observed to range from 62·2 and 44·6 % in Ghana for men and women, respectively, to 7·6 % in Tanzanian men and 2·2 % in South African women^{( 14 )}. When faced with emergency situations, the prevalence may be considerably higher, as exemplified by Sierra Leone, where 75 % of older persons were classified as underweight.

Older populations in rural areas were observed to suffer from poor nutritional status, particularly underweight⁽ Reference Blankson and Hall ^{15 )}. A small number of studies report this issue. In a cross-sectional survey conducted in rural Ghana, 41 % of fifty-nine women aged 60–92 years were categorised as underweight (BMI <18·5 kg/m²), whereas 16·9 % were categorised as overweight or obese⁽ Reference Blankson and Hall ^{15 )}. A cross-sectional study in Burkina Faso indicated that age was a significant factor for undernutrition, noting that 50 % of participants over 70 years were undernourished compared with 31 % between the ages of 60 and 69 years. In addition, approximately 65 % of participants, all of whom were over 60 years were suffering from multi-morbidity, most commonly being hypertension (82 %), malnutrition (39 %) visual impairments (28 %) and diabetes (27 %)⁽ Reference Hien, Berthé and Drabo ^{16 )}.

It was noted that obesity was more prevalent in urban areas, whereas in contrast undernutrition was more prevalent in rural areas. This was suggested to be due to the impact of the nutrition transition in urban areas where a higher intake of sugar, saturated fat and low nutrient density foods may encourage obesity. Another possible reason given was due to the high energy expenditure of rural inhabitants who were more likely to source their food directly from the land and have a more physical lifestyle compared with their urban counterparts who have become accustomed to a more modern sedentary way of life. This observation presents a potential research opportunity to establish whether undernutrition in rural areas and/or low BMI is due to low energy intake or high energy expenditure.

Undernutrition is also present in older persons residing in urban areas. Underweight prevalence in the urban areas of Lake Victoria Basin in East Africa was 24·1 % among older men and 12·3 % among older women⁽ Reference Cheserek, Waudo and Tuitoek ^{17 )}. Only 22 % of older men and 28 % of older women met their daily energy requirements; This was attributed to inadequate access to food, improper eating patterns and poor health and living arrangements. Similar observations of high prevalence of undernutrition were made in a study conducted to access nutritional status of older persons in urban setting in Central Africa (The EDAC study). The study concluded that eating only one meal was the sole factor associated with older persons⁽ Reference De Rouvray, Jesus and Guerchet ^{18 )}.

Obesity

Obesity is currently a worldwide epidemic that has almost doubled between 1980 and 2008, and now attributes to 2·8 million deaths each year. In SSA, women have been observed to be almost twice as likely to be obese than men^{( 10 )}. Studies by Kimokoti and Hamer⁽ Reference Kimokoti and Hamer ^{19 )} show that up to half (6–48 %) of older people aged 60 years and over in SSA were underweight and nearly a quarter (2·5–21 %) were overweight; while 56 % of older South Africans alone were obese⁽ Reference De Rouvray, Jesus and Guerchet ^{18 )}.

Nutrient deficiencies

To date, very little published data is available on micronutrient deficiencies in older persons. In a related study conducted in Sharpeville, South Africa, deficiencies in calcium, magnesium, zinc, copper, selenium, iodine, vitamins A, B₁, B₂, B₃, B₁₂, C, D, E, folate and biotin were recorded, with women having higher nutrient deficiencies than men. A total of 58·6 % of the adults older than 60 years consumed three meals daily, while 28·9 % consumed two meals daily⁽ Reference Oldewage-Theron and Kruger ^{20 )}. The dietary diversity scores in the study were as follows: 55·1 % had low dietary diversity (0–3 food groups), 37·6 % had medium dietary diversity (4–5 food groups) and 7·4 % had high dietary diversity (6–9 food groups). The vast majority of foods consumed were carbohydrates. Nutrient adequacy increased in line with the dietary diversity. It could be hypothesised that the prioritising of children's needs before their own was responsible for their reduced nutritional status, as 43·8 % of grandmothers were responsible for feeding the family children.

A follow up study in the same area recorded a positive increase in dietary diversity (98·1 %) subsequent to intervention. Further research could be looked at in different areas in SSA.

Hypertension

Hypertension is a precursor for a number of NCD including stroke, dementia and CVD. Its prevalence is believed to be on the rise as a result of increased urbanisation and the associated nutrition transition. The prevalence of hypertension is also known to increase with age, as a disproportionate number of older persons are affected by hypertension⁽ Reference Peltzer and Phaswana-Mafuya ^{21 )}. In a study conducted among SSA adults, where close to 25 % of the study population were over 50 years of age, it was observed that hypertension prevalence was significantly higher among older persons than adults from other groups, particularly adults aged 18–29 years (prevalence ratio: 2·20 (95 % CI 1·49, 3·25)). This concluded that age was strongly associated with increase prevalence of hypertension, alongside BMI, level of education and tobacco use⁽ Reference Guwatudde, Nankya-Mutyoba and Kalyesubula ^{22 )}. In a South African study, a cross-sectional survey among 3840 adults aged 50 years and over revealed a prevalence rate of 77·3 %, with 38·1 % aware of their condition⁽ Reference Peltzer and Phaswana-Mafuya ^{21 )}.

Using household survey data (2007–2010) from the WHO study on Global Ageing and Adult Health; researchers revealed a hypertension prevalence among adults aged 50 years and over of 57·1 and 77·9 % in SSA countries Ghana and South Africa, respectively⁽ Reference Lloyd-Sherlock, Beard and Minicuci ^{23 )}. In addition, it was shown that hypertension was associated with overweight/obesity, women and those in the lowest wealth quintile. Among those who were hypertensive, only 38 % in South Africa and 23·3 % in Ghana were aware of their condition⁽ Reference Lloyd-Sherlock, Beard and Minicuci ^{23 )}.

Despite urbanisation being considered a risk factor, hypertension may also be an issue in rural settings. In rural Tanzania, blood pressure measurements were taken from 2223 adults aged 70 years and over. It was observed that 69·9 % of participants were hypertensive; with only 37·7 % aware of their condition and 6·1 % receiving treatment⁽ Reference Dewhurst, Dewhurst and Gray ^{24 )}.

It has been predicted that an increased NCD burden among older persons may pay a significant toll on the health services in SSA. NCD were responsible for 81·0 % (708/874) of hospital admissions among older patients in Nigeria, Sudan and Tanzania, while tuberculosis, malaria and HIV accounted for only 4·6 % (40/874)⁽ Reference Akinyemi, Izzeldin and Dotchin ^{25 )}. Interestingly, it was observed that hypertension rates were similar in these three countries to the rates in the UK (40·2 % compared with 45·8 % in the UK). In a national cross-sectional survey of 3840 participants aged 50 years or older in South Africa, it was shown that 19·7 % consumed tobacco daily, while 68·5 % did not consume an adequate amount of fruit and vegetables. In addition, 60·5 % did little physical activity, which went up to 63·1 % in women and 71·2 % in those over the age of 70 years. Unsurprisingly, 68·2 % of participants were overweight or obese (71·9 % in women) and 75·3 % were suffering from hypertension⁽ Reference Phaswana-Mafuya, Peltzer and Chirinda ^{26 )}.

In addition to obesity, it is being theorised that undernutrition may increase the risk of some NCD such as stroke in older persons. As well as having the highest stroke prevalence, older persons in Tanzania also have the lowest mean BMI compared with the rest of the SSA⁽ Reference Kimokoti and Hamer ^{19 )}. Given this combination of factors it could be theorised that undernutrition or low BMI may be an influence of stroke, either solely or combined with a third feature. A further study supports this theory by showing the risk of haemorrhagic stroke in individuals with BMI <18 kg/m² is equal to those with BMI 26 kg/m²⁽ Reference Song, Sung and Davey Smith ^{27 )}. It was noted that due to the high rate of mortality in SSA, the occurrence of stroke among older persons may be less common; however the prevalence in persons over 70 years in Tanzania was relatively high (23/1000) compared with rest of SSA. In fact, stroke incidence could be as high, if not higher than in high-income countries, with risk increasing with age. To date there has been an absence of adequate community-based stroke incidence studies. Hospital-based incidence was observed to be lower than in high-income countries, but higher in young people, possibly due to hospital admission bias. There has been no community-based data on case fatality⁽ Reference Connor ^{28 )}.

Little is known of the direct mechanism behind the link between undernutrition and NCD. However, it is proposed that poor prenatal nutrition and resulting low birth weight can genetically predispose individuals to NCD risk in later life; in particular CVD. Based primarily on animal model data, the theory postulates that fetal exposure to maternal glucocorticoids, as well as stimulation of the fetal renin-angiotensin system may lead to vascular resistance and hypertension, which are precursors for NCD⁽ Reference Langley-Evans ^{29 )}.

During the period of 1999–2003, 7074/100 000 deaths in persons over 65 years were recorded in Burkina Faso as a result of CVD, which was the fourth leading cause of death in adults over 40 years. In terms of uncommon NCD such as peripheral artery disease, the stark reality is that many older persons go largely undiagnosed, and in cases where they are, patients rarely have access to affordable, quality health care⁽ Reference Guerchet, Aboyans and Mbelesso ^{30 )} A systematic door-to-door survey among 976 persons over 65 years in the Central African Republic and Congo Brazzaville measured the prevalence of peripheral artery disease and observed a 15 % prevalence in Central African Republic and a 32·4 % in Brazzaville, which increased with age. Researchers also noted that this prevalence was higher than in some North American studies. In addition, hypertension was associated with peripheral artery disease in women, while diabetes was associated with men⁽ Reference Guerchet, Aboyans and Mbelesso ^{30 )}.

Diabetes

Globally, diabetes mellitus is the fourth most common chronic disease within multi-morbidity at 27 %, with the prevalence of multi-morbidity among older persons at 65 %⁽ Reference Hein ^{31 )}. In addition to environmental factors such as obesity and sedentary behaviour, a growing ageing population has been associated with the increase in prevalence of diabetes in SSA⁽ Reference Werfalli, Musekiwa and Engel ^{32 )} Currently, the highest global age-specific mortality rate for diabetes is in Africa⁽ Reference Werfalli, Musekiwa and Engel ^{32 )}. Diabetes appears to have been well researched in SSA, either solely or as a factor of multi-morbidity; however, little has been done with regard in terms of research related to older persons.

Dementia

In many SSA countries dementia has become a significant economic burden. Yet in spite of this, the condition is largely perceived to be a normal part of ageing, leaving many patients to suffer undiagnosed⁽ Reference Olayinka and Mbuyi ^{33 )}. As a general observation, epidemiology data on dementia in SSA has been limited; however, it is suggested that prevalence rate may in fact be parallel with that of high-income countries⁽ Reference Olayinka and Mbuyi ^{33 )}. In a systematic review investigating dementia and cognitive impairment in older persons in SSA, eleven identified studies highlighted a wide variation in prevalence across different countries. A lower overall prevalence compared with global reports was observed, possibly due to the fact that there are a higher number of persons aged 66–80 years and a smaller number aged 81 years and older⁽ Reference Mavrodaris, Powell and Thorogood ^{12 )}. However, similar to global reports, older age and female sex was associated with dementia⁽ Reference Mavrodaris, Powell and Thorogood ^{12 )}. Other studies indicate the highest prevalence of dementia was found in ages 60–69 in women and in men over the age of 70 years⁽ Reference Olayinka and Mbuyi ^{33 )}. In 2010, dementia prevalence in adults aged 50 years and older was estimated at about 2·1 million people (2·4 %), which was highest in women aged 80 and over (19·7 %). Alzheimer's disease was found to be responsible for 57·1 % of all dementia cases, followed by vascular dementia at 26·9 %⁽ Reference Olayinka and Mbuyi ^{33 )}. Apart from increasing age and sex, the main risk factor appeared to be CVD, which in itself is associated with overweight and obesity⁽ Reference Olayinka and Mbuyi ^{33 )}.

Dementia epidemiology is believed to be influenced by several factors including both communicable and NCD such as HIV/AIDS, CVD, hypertension and diabetes, as well as illiteracy. The traditional underreporting of dementia and other age-related conditions such as stroke and Parkinson's disease in SSA may have been as a result of the low life expectancy at birth and low utilisation of healthcare facilities⁽ Reference Callixte, Clet and Jacques ^{34 )}. The prevalence of dementia among rural populations in SSA has been observed to be double that of urban populations (40·7 v. 21·9/1000), which has been attributed to lower socioeconomic status⁽ Reference Dewhurst, Dewhurst and Gray ^{35 )}.

In a cross-sectional epidemiological survey in rural Tanzania, the age-adjusted prevalence of neurological disorders among the 2232 participants was calculated at 154·1/000, which included tremors, headaches, stroke, peripheral polyneuropathy, upper limb mononeuropathy and Parkinson's disease⁽ Reference Dewhurst, Dewhurst and Gray ^{35 )}. The authors concluded that neurological disorders does and will continue to contribute significantly to the NCD epidemic⁽ Reference Dewhurst, Dewhurst and Gray ^{35 )}. In addition to socioeconomic status, nutritional status is also believed to play a role in susceptibility to neurological disorder, with individuals with BMI <18·5 kg/m² are considered to be at greater risk of developing dementia. Individuals aged 65 years and older in Central African Republic reported that 70 % of dementia sufferers were undernourished with BMI ≤18·5. Also, 67·5 % of dementia patients were observed to have hypertension⁽ Reference Mbelesso, Tabo and Guerchet ^{36 )}.

In a study conducted in the Central African Republic and the Democratic Republic of Congo, dementia patients over the age of 65 years had a higher prevalence of undernutrition compared with healthy counterparts (32·0 v. 17·7 %, respectively). Such patients were less likely to eat with their family compared with healthy people (66·7 v. 90·6 %)⁽ Reference De Rouvray, Jesus and Guerchet ^{18 )}.Within the sample population of 1016 participants, dementia prevalence was 7·4 %, with an undernutrition prevalence of 19·2 %. Indeed dementia patients were more likely to be undernourished compared with healthy participants (32·0 v. 17·7 %, respectively)⁽ Reference De Rouvray, Jesus and Guerchet ^{18 )}.