AbstractBackgroundSubjective memory complaint (SMC) is defined as the perception of memory deficits despite a normal cognitive assessment, it is related to long‐term cognitive decline risk. The guidelines recommend medical advice about risk factor control, a healthy diet, and regular exercise for lower risk. However, there is no evidence that would guarantee good adherence. Social networks are conduits of support, information, and health behavior flows, they may have an impact on SMC adherence to the suggested changes. Our aim is a) to assess long‐term adherence of SMC to lifestyle changes b) to evaluate social networks as a predictor of adherence.MethodsProspective cohort study recruiting SCM participants with a normal UDS3 neuropsychological battery. Participants were evaluated on individual risk based on CAIDE score and received guideline‐based recommendations about lifestyle changes. Subjects were re‐evaluated 6 months after recommendations to assess their adherence to a healthy diet (MIND score), physical activity (IPAQ), risk factor control, and their social network (Figure 1). We fitted a generalized linear model to predict IPAQscore and MINDscore with social network features.ResultsWe recruited 200 participants (age 68.1 ± 6.5; 62% female; 57% first‐degree family history of dementia, and 15.91±2.28 years of education). Participants presented a 7.55±2.04 CAIDE score (Table 1): 37% high blood pressure, 51% obesity, dyslipidemia 43%, Social isolation 25%, 51% sedentarism, 9% smoking, and 22% heavy drinking. Six months after medical advice, 15% stopped drinking, 5.1% quit smoking, 37% started physical activity, and 66% a healthy diet. Regarding social networks, the median size of the network was 8 IQR[5‐12], 50% were kin, gender diversity: IQV = 0.77; 50% IQR[20%‐60%] sedentary, 20% IQR[0%‐44%] unhealthy diet. The percentage of sedentary social network members was a predictor of the intensity of the weekly physical activity of participants (in METS/minutes/week)(β ‐2.82 95%CI[‐4.7,‐9.27] p = 0.005) (Table 2).ConclusionParticipants showed good adherence to the proposed intervention. Social networks have an impact on adherence and it is a potential target to prevent dementia.